Social Corporatism: A Superior Economic System (Full Book

This is a rekeyed full text file of the volume
Social Corporatism: A Superior Economic System?

Edited by Jukka Pekkarinen, Matti Pohjola, and Bob Rowthorn
Originally published in 1992 by Oxford University Press

the book is now out of print and the copyright has reverted
to UNU-WIDER.
File produced by Macmillan Publishing Solutions, Chennai

WIDER
Studies in Development Economics embody the output of the

research programmes of the World Institute for
Development Economics Research (WIDER), which was
established by the United Nations University as its first
research and training centre in 1984 and started work in
Helsinki in 1985. The principal purpose of the Institute is
to help identify and meet the need for policy-oriented
socio-economic research on pressing global and
development problems, as well as common domestic
problems and their inter-relationships.
Social Corporatism:
A Superior Economic System?
Edited by
Jukka Pekkarinen, Matti Pohjola,

and Bob Rowthorn
A Study prepared for the World Institute
for Development Economics Research (WIDER)
of the United Nations University
CLARENDON PRESS • OXFORD

1992
Oxford University Press, Walton Street, Oxford 0x2 6DP
Oxford New York Toronto
Delhi Bombay Calcutta Madras Karachi
Petaling Jaya Singapore Hong Kong Tokyo
Nairobi Dar es Salaam Cape Town
Melbourne Auckland
and associated companies in
Berlin Ibadan
Oxford is a trade mark of Oxford University Press
Published in the United States

by Oxford University Press, New York
© The United Nations University 1992
World Institute for Development Economics Research (WIDER)—

The United Nations University 42C, 00100 Helsinki, Finland
All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise, without
the prior permission of Oxford University Press
British Library Cataloguing in Publication Data

Data available
Library of Congress Cataloging in Publication Data

Social corporatism: a superior economic system? / edited by Jukka
Pekkarinen, Matti Pohjola, and Bob Rowthom.
p. cm.—(Studies in development economics)
“A Study prepared for the World Institute for Development
Economics Research (WIDER) of the United Nations University.”
Includes bibliographical references and index.
1. Sweden—Economic policy. 2. Corporate state—Sweden.
3. Sweden—Politics and government—1905– 4. Scandinavia—Economic
policy. 5. Corporate state—Scandinavia. 6. Scandinavia—Politics
and government—1945– 7. Economic policy. 8. Corporate state.
I. Pekkarinen, Jukka, 1947– . II. Pohjola, Matti. III. Rowthom,
Bob. IV. World Institute for Development Economic Research.
V. Series.
HC375.S6855 1992 338.9485—dc20 91–5027
ISBN 0–19–828380–6
Typeset by Alliance Phototypesetters, Pondicherry

Printed and bound in
Great Britain by Bookcraft (Bath) Ltd.
Midsomer Norton, Avon
FOREWORD
Post-war macro-economic management in OECD has long pursued what

seemed an elusive ‘Holy Grail’. This was a combination of high growth
with low inflation and high levels of employment. For many OECD
countries the reduction of inflation has meant the acceptance of socially
and politically problematic levels of unemployment. Attention therefore has
increasingly centred on a group of economies described as corporatist which
had succeeded in the 1970s in reconciling these apparently conflicting
objectives. What the so-called corporatist economies have in common is an
annual central wage bargain involving an annual centralized process of wage
bargaining with the employers as a group meeting labour unions as a group,
with the whole process being mediated by the political system usually
involving a coalition. Since growth in the developing world is still largely a
function of successful recipes for growth in the developed world the cor-
poratist experiment was a natural focus for WIDER research. Indeed the
wage-bargaining process was a particular focus of the report of WIDER’s
World Economy Group.1 The group asserts that: ‘If EEC countries wish
to return to full employment and stay there, they could do worse than look
at Sweden.’
The present volume, taking the common feature of what may be de-
scribed as social corporatism as a working definition, namely the centralized
wage bargain, seeks to probe the conditions under which corporatism
succeeds in actually finding the Holy Grail. There are corporatist success
stories, notably the Scandinavian countries—Sweden, Finland, Norway,
and Denmark—and there are the failures such as The Netherlands. The
working definition that has been adopted for the purpose of this research
permits the corporatist label to be extended to countries as diverse as
Austria and Australia, and the extent to which these experiments have suc-
ceeded is also reviewed. The broad conclusion of the research is that the
centralized wage bargain is a necessary but by no means sufficient condition
for the successful reconciliation of otherwise incompatible objectives.
The Swedish experience is particularly instructive in that unemployment
has never exceeded 3.5 per cent and now stands at under 1.5 per cent. In
addition to centralized wage determination there are two other features
common to the social corporatist success stories, namely unemployment
benefits that run out after a year or so, and major training programmes for
the unemployed. The Swedish labour market policy specifically has four
main ingredients. First, a placement and counselling service operated by
1
World Imbalances, WIDER World Economy Group Report (Helsinki, 1989), ch. 3:
‘European unemployment’.
vi FOREWORD
employment exchanges where each staff member has a case load of 15

unemployed compared with 375 for example in Britain. Secondly, there is
provision for retraining of workers even before they become unemployed
and about 1 per cent of the labour force are on training courses of this kind.
Third, recruitment subsidies are given to employers for recruiting workers
who have not been placed within six months. The subsidy is 50 per cent
of the wage and lasts for six months. Fourthly, there is provision for
temporary public employment and a right to work. The public sector, in
effect chiefly local authorities, becomes the employer of last resort and work
is available to anyone whose unemployment benefit entitlement has run
out. While such policies are expensive and Sweden spends nearly 1 per cent
of national income on them, there are substantial savings resulting from the
resulting reduction in unemployment benefit. While unemployment benefit
takes up 1.5 per cent of GNP in the EEC, it costs 0.7 per cent in Sweden.
Overall, therefore, while Sweden may be spending 1.7 per cent of GNP on
her full-employment programmes, or about as much as the EEC overall,
the output gains of substantially lower unemployment mean that the pro-
gramme is largely self-financing.
The WIDER study suggests that developing countries could do well to
emulate the example of the successful corporatist countries in building the
necessary institutions at an early stage of their development. Indeed as the
pivotal centrally planned command economies of Eastern Europe move
increasingly over to market solutions they will have the opportunity to profit
from the lessons of the corporatist experience by incorporating the key ele-
ments in rebuilding structures that are now very much in the melting-pot.
Lal Jayawardena
Director, WIDER
CONTENTS
LIST OF CONTRIBUTORS ix
LIST OF FIGURES xi
LIST OF TABLES xiii
1. Social Corporatism and Economic Performance: Introduction

and Conclusions 1
Jukka Pekkarinen, Matti Pohjola, and Bob Rowthorn
2. Lessons from ‘Corporatist’ Theorizations 24
Göran Therborn
3. Corporatism and Wage Bargaining 44
Matti Pohjola
4. Corporatism and Labour Market Performance 82
Bob Rowthorn
5. Corporatism, Patterns of Employment, and Access
to Consumption 132
Andrew Glyn
6. Saving and Economic Growth from a Nordic Perspective 178
Katri Kosonen
7. Social Corporatism and Long-Term Economic Performance 210
Michael Landesmann and Juhana Vartiainen
8. Industrial Policies and Social Corporatism 242
Michael Landesmann
9. Exchange Rate Policy and Employment in Small Open
Economies 280
Sixten Korkman
10. Corporatism and Economic Performance in Sweden,
Norway, and Finland 298
Jukka Pekkarinen
11. Corporatism: Success or Failure? Austrian Experiences 338
Alois Guger
12. The Swiss Model: Corporatism or Liberal Capitalism? 363
Wolfgang Blaas
13. The Unexpected Emergence of Australian Corporatism 377
Robin Archer
INDEX 419
Acknowledgements
The papers published in this volume are the outcome of a research project
on social corporatism carried out in 1988–1989 with financial contribution
provided by the Ministry for Foreign Affairs of the government of Finland.
The project was organized under the joint auspices of WIDER and the
Labour Institute for Economic Research in Helsinki. The authors would
like to thank the staff of these institutes for their help and encouragement.
We should also like to thank the participants in the various seminars and
meetings organized in connection with this project, in particular, Tariq
Banuri, Lal Jayawardena, Pekka Korpinen, Stephen Marglin, Juliet Schor,
and of course all the contributors to this volume.
J.P.
M.P.
B.R.
LIST OF CONTRIBUTORS
ROBIN ARCHER, Balliol College, Oxford, UK.

WOLFGANG BLAAS, Associate Professor of Economics, University of Technology,
Vienna, Austria.
ANDREW GLYN, Fellow, Corpus Christi College, Oxford, UK.
ALOIS GUGER, Senior Research Fellow, Austrian Institute of Economic Research,
Vienna, Austria.
SIXTEN KORKMAN, Head of Economic Department, Ministry of Finance, Helsinki,
Finland.
KATRI KOSONEN, Senior Research Officer, Housing Department, Ministry of En-
vironment, Helsinki, Finland.
MICHAEL LANDESMANN, Senior Research Officer, Department of Applied Eco-
nomics, University of Cambridge, and Fellow, Jesus College, Cambridge, UK.
JUKKA PEKKARINEN, Director, Labour Institute for Economic Research, Helsinki,
Finland.
MATTI POHJOLA, Professor of Economics, Helsinki School of Economics and Bu-
siness Administration, Helsinki, Finland.
BOB ROWTHORN, Professor of Economics, University of Cambridge, Cambridge,
UK.
GÖRAN THERBORN, Professor of Sociology, University of Gothenburg, Gothenburg,
Sweden.
JUHANA VARTIAINEN, Senior Research Fellow, Labour Institute for Economic Re-
search, Helsinki, Finland.
LIST OF FIGURES
3.1 Average employment and unemployment in 17 OECD

countries in 1980–1985 48
3.2 Changes in average employment and unemployment rates
between the periods 1968–1973 and 1980–1985 48
3.3 Centralization of the bargaining structure and average
employment in 1980–1985 51
3.4 Centralization of the bargaining structure and changes in
employment between the periods 1968–1973 and 1980–1985 51
3.5 Centralization of the bargaining structure and average
unemployment in 1980–1985 52
3.6 Centralization of the bargaining structure and changes in
unemployment between the periods 1968–1973 and 1980–1985 52
3.7 Centralization of the bargaining structure and the growth of
real GDP in 1973–1984 54
3.8 Centralization of the bargaining structure and real wage rigidity 55
3.9 Corporatism and average employment in 1980–1985 60
3.10 Corporatism and real wage rigidity 60
3.11 Co-operative and non-co-operative solutions to union rivalry 63
4.1 Bargaining structure and wage dispersion 96
4.2 Bargaining structure and employment 98
4.3 Bargaining structure and labour market performance 103
4.4 Inter-firm productivity and wages in Swedish industry, 1985 115
5.1 Demarketization and employment 141
5.2 Employment and market hours, 1965–1973 143
5.3 Employment and market hours, 1979–1986 144
5.4 Market work and market output, 1979–1986 145
5.5 Employment spreading, 1979–1986 147
5.6 Consumption from earnings and productivity, 1973–1979 153
5.7 Sweden: consumption and employment 155
5.8 Austria: consumption and employment 156
5.9 Denmark: consumption and employment 157
5.10 Finland: consumption and employment 158
5.11 Norway: consumption and employment 159
5.12 The Netherlands: consumption and employment 160
5.13 Belgium: consumption and employment 161
6.1 Household sector: net saving divided by disposable income 181
6.2 Net household saving divided by net saving of the nation 181
6.3 Household sector, general government, and corporate saving
divided by GDP 183
6.4 Financial saving by sector 184
6.5 Total social expenditure divided by GDP, 1971–1981 191
xii LIST OF FIGURES
7.1 Productivity in manufacturing 230

7.2 Real earnings and productivity in manufacturing 231
7.3 Real earnings in manufacturing 232
7.4 Investment shares and profit ratios in OECD economies,
1960–1985 234
7.5 Investment ratios in three groups of countries 235
8.1 Employment in Austria 256
8.2 Growth of manufacturing output 257
8.3 Investment rate in manufacturing, 1970–1986 258
8.4 R. & D. expenditure and fixed investment in Finnish
manufacturing constant prices 259
8.5 Swedish industrial subsidies, 1970–1979 264
8.6 Structure of industrial support in Sweden 265
8.7 International competitiveness in manufacturing in Norway 269
8.8 The degree of subsidization: an international comparison
(Norway, Sweden, the FRG, and the United States) 270
9.1 Industrial production in Finland, Austria, Denmark, Norway,
and Sweden 289
9.2 Finland: international competitiveness 290
9.3 Sweden: international competitiveness 291
9.4 Austria: international competitiveness 292
9.5 Norway: international competitiveness 293
9.6 Denmark: international competitiveness 294
11.1 Foreign workers and unemployed in Austria 342
11.2 Employment in nationalized industries in Austria 344
11.3 Growth of nominal and real wages per capita in Austria 351
12.1 Unemployment rates 365
12.2 Total employment 365
12.3 Foreign labour in Switzerland 372
13.1 The six dimensions of an industrial relations system 378
13.2 Union centralization 384
13.3 Terms of trade, Australia 394
13.4 Net external debt, Australia 395
13.5 Balance on current account, Australia 395
13.6 Inflation rates, Australia: quarterly data, percentage changes on
a year earlier 397
13.7 Labour market aggregates, Australia 398
13.8 Real commodity prices, Australia 404
LIST OF TABLES
3.1 Classification of countries according to their degree of

centralization 50
3.2 Classification of countries according to their degree of
corporatism 59
4.1 Dispersion of hourly earnings in the late 1980s (mostly manual
workers) 91
4.2 Decomposing changes in wage dispersion, 1973–1985 92
4.3 Regression analysis: employment rate and centralization 97
4.4 Aspects of labour market performance, 1985 102
4.5 Dispersion of hourly earnings in three countries in the late
1980s 105
4.6 INDEED indices, 1973–1985 106
4.7 Regression analysis: employment rate, centralization, and wage
dispersion 112
4.8 Labour market performance and welfare benefits in the 1980s
in 17 OECD countries 121
4.9 Labour market performance and welfare benefits in 17 OECD
countries 123
5.1 Demarketization and employment, 1965–1986 136
5.2 Employment and consumption spreading, 1965–1986 149
6.1 The average rates of saving, investment, and growth in various
OECD countries, 1960–1984 180
6.2 The rate and the structure of taxation in various OECD
countries in 1985 197
6.3 Average and marginal income tax rates in various OECD
countries in 1986 198
7.1 GDP (growth per annum) 229
7.2 Real GDP per capita (growth) 229
7.3 Real value-added in manufacturing (growth) 229
7.4 Real value-added in manufacturing per person employed 230
7.5 Real gross fixed capital formation (growth) 233
7.6 Gross fixed capital formation as a percentage of GDP 233
7.7 Rates of return in the business sector 233
7.8 Employment in manufacturing, services, and government 236
7.9 Real value-added in services (growth), and per person employed 237
7.10 Trade balance of a percentage of GDP (average) 237
7.11 General government claim on resources 238
8.1 Percentage of the labour force employed in agriculture,
industry, and services in Austria, Finland, Norway, and Sweden 250
8.2 Developments in Austria, Finland, Norway, and Sweden,
1973–1987 251
xiv LIST OF TABLES
8.3 Share of high-technology products in exports of manufacturing

in Austria, Germany, and Sweden 253
8.4 R. & D. expenditure as percentage of GDP in Austria, Finland,
and Norway 253
8.5 Total patent applications in Austria, Finland, Norway, and
Sweden 254
8.6 Unemployment in Austria 255
8.7 Constant market share analysis of Finnish exports 259
8.8 Firm-specific subsidies by receiving sector in Sweden,
1977–1979 264
8.9 The structure of Swedish exports 266
8.10 State ownership in industry in Sweden in 1976 267
8.11 Sectoral employment shares in Norway 268
8.12 Industrial support to shipbuilding and state companies in
Norway 271
8.13 Employment trends in three ailing sectors (Norway, Sweden,
Germany, France, United Kingdom) 272
10.1 Economic indicators on Sweden, 1973–1988 307
10.2 Economic indicators on Norway, 1973–1988 318
10.3 Economic indicators on Finland, 1973–1988 324
11.1 Growth of real GDP per capita 341
11.2 Employment, unemployment, and labour market participation
in Austria 343
11.3 Labour costs and productivity in industry, Austria 351
11.4 Demand management and macro-economic performance after
OPEC I and OPEC II in Austria 353
11.5 Wage differentials in Austrian industry 357
12.1 Macro-economic performance: Switzerland in international
comparison 364
13.1 Key events in national wage fixation in Australia, 1904–1983 389
13.2 Wage-fixing systems under stagflation in Australia 396
13.3 Real GNP in Australia 399
13.4 International employment and unemployment 400
13.5 Wages and household income in Australia 402
1
Social Corporatism and Economic Performance:
Introduction and Conclusions
Jukka Pekkarinen, Matti Pohjola, and Bob Rowthorn
1.1 THE CONCEPT OF SOCIAL CORPORATISM
The notion of corporatism as an economic system distinct from both

classical capitalism and socialism has experienced a revival of intellectual
interest since the mid-1970s. This was prompted by the diversification of the
economic performance of different countries which called for an explana-
tion. In the writings of modern social scientists various prefixes or adjectives,
such as neo, liberal, social, or societal, are attached to the term to make a
clear distinction between its modern and old meanings (for reviews see, for
example, Pryor 1988 and Williamson 1989).
The old strand had its roots in nineteenth-century Roman Catholic social
thought as well as in German romanticism and idealism. It first appeared as
a response to the social problems arising from industrialization and main-
tained that class conflict is not an inherent feature of the capitalistic way of
organizing ownership and production. Great emphasis was placed on the
roles of semi-autonomous bodies representing all occupations in the or-
ganization of economic activities and the making of parliamentary policy.
During the first decades of the current century various political movements—
Fascists on the far right and Social Democrats and guild socialists on the
left—developed their own variants of corporatism as an alternative economic
system to either capitalism or socialism.
Leaders of the former movement saw corporatism as a means to support
nationalism whereas those of the latter regarded it as a way of avoiding the
excessive centralism of the Soviet type of socialism. After the Second World
War, however, the papal interest in corporatism waned and the state corpor-
atism of the extreme right was politically discredited.
What remains of this older tradition in the modern strand of corporatist
theorization is the view that corporatism is a variant of co-ordinated capitalism
in which interest organizations, especially those representing the interests of
labour, play important roles in the organization of economic activities and,
consequently, in the economic performance of a nation. The links with Social
Democracy have also remained strong. But just like capitalism and social-
ism, the precise meaning of corporatism is hard to define. In Chapter 2 of
2 J. PEKKARINEN, M. POHJOLA, B. ROWTHORN
the present volume Göran Therborn writes under the title ‘Lessons from
“corporatist” theorizations’ that the concept has become a bit like God: many
people believe that it is an important phenomenon but nobody really knows
what it looks like. This does not, however, mean that studying it is useless.
In fact, his conclusion is that corporatist theorists have hit upon something
which is very important but difficult to analyse because corporatism remains
a moving target, being subject to various definitions and specifications.
Our project was set up to hit this target by defining corporatism in an
explicit way and by exploring the explanatory power of the chosen definition.
In doing this we deliberately put substantial emphasis on the recent contribu-
tions of those economists who have studied the operation of organized labour
markets and their importance to macro-economic development (for a survey,
see Matti Pohjola’s essay in Chapter 3). Economists’ interest was drawn to
corporatism by Bruno and Sachs (1985), but they had a notable predeces-
sor—John Maynard Keynes (1927), who advocated something similar to it
but did not succeed in persuading his colleagues. The sociological and polit-
ical aspects received less attention in our plans—not because we regarded
them as unimportant but because we wanted to concentrate on the economic
consequences of the operation of organized interest groups.
We started by defining corporatism in the narrow sense as meaning
centralized wage bargaining at the national level between strongly organized
employee and employer associations. But the inadequacies of this concept
became evident in the course of the work of the project. It turned out that the
degree of centralization cannot in itself explain the diversity of economic
performance between countries. This is reflected in many essays in the
volume. Corporatism has to be specified in greater detail. In summarizing the
contributions, we now find it most helpful to use the more specific concept
of social corporatism by which we mean an economic system whose labour
market is characterized by two basic features. The first one is centralized
wage bargaining. Besides wages, bargaining may also include government
economic and social policies in which case the state is either formally or
informally involved in die process.
Corporatism thus derives from associative action, in contrast to individual
or commanded action, and, more specifically, from a particular kind of
associative action: class organization by both labour and capital. As
demonstrated by Therborn in Chapter 2, differences between countries in
their historical experiences and legal traditions make some societies more
likely than others to exhibit this type of associative action. In particular, the
Germanic type of labour movement and the Nordic type of employer
organizations are conducive to politico-economic deals. The former is
characterized by either close co-operation with or guidance of political
parties to make a common effort at class organization, while the latter has
also transferred experiences from parliamentary politics and the civil service
CORPORATISM AND ECONOMIC PERFORMANCE 3
in mobilizing the employer side. The Nordic countries and Austria come at
the top of the list in various rankings of the countries according to the degree
of centralization of the bargaining structure, while the United States and
Canada are at the bottom (see the essays by Pohjola and Bob Rowthorn in
the present volume).
The second central feature of social corporatism is its non-exclusive and
egalitarian nature. By this we refer, first, to the non-exclusion of any social
group from the labour market and, second, to a high degree of equality in
sharing both the benefits of increasing economic welfare and the miseries of
recession or the burden of adjustment. This feature can be seen as an
extension of citizens’ political rights to cover economic activities in order to
achieve social order in industrial and labour relations. Solidarity is its natural
concomitant.
A perspective on our refinement of the definition of corporatism is given
by Therborn. In his essay he argues that peaceful industrial order can be
achieved in two different ways: either through the institutionalization of
partnership or consensus, as for example in Austria and The Netherlands,
or through the institutionalization of conflict as in the Nordic countries. In
the latter model labour and capital are organized from below according to
the principles of democracy and are regarded as labour market parties whose
autonomy and conflict rights are respected and not violated. Using econom-
ists’ terminology, this means that their property rights are well defined. Using
our terminology introduced above, organizing labour market relations on
conflict bases is a way to institutionalize the non-exclusive and egalitarian
feature of corporatism.
The institutionalization of partnership or consensus is different in the
sense that industrial peace may be achieved at the expense of groups which
are either economically or politically too weak to have any considerable
bargaining power, that is by excluding them from the bargaining table. From
the essays by Therborn, Rowthorn, and Alois Guger the conclusion can be
drawn that Austria and the Nordic countries are different in the respect that
only the latter, and even then in a varying degree, satisfy our second criterion
of social corporatism.
1.2 LABOUR MARKET PERFORMANCE
The received view in the economic literature is that perfect competition

between atomistic economic agents is a socially efficient way to organize the
operation of product and factor markets. Given well-known qualifications
regarding externalities and scale economies, private non-co-operative
decision-making results in an efficient and effectively co-ordinated solution
for wages and prices in the economy as a whole. According to this view,
which we may here call market liberalism, the lack of competition in output
and factor markets is the reason for the poor economic performance of a
nation. The enhancement of competition is seen as the desirable solution to
unemployment, inflation, growth, and productivity problems. As regards the
labour market and the view that unemployment results from excessively high
wages, this policy prescription means that the power of organized labour
should be reduced in one way or another. By such power is generally meant
the ability of trade unions to push wages above the levels which would
equalize the demand for and the supply of labour. As their paradigm the neo-
liberals take the deregulated markets of countries like the United States and
Canada in which, they argue, the labour market exhibits flexibility and ability
to adapt to changing economic circumstances.
Paradoxically, however, the Walrasian ideal of full employment with
approximately equal wages seems to have been best achieved in the social
corporatist countries. This fact is considered in Chapters 3 and 4. First,
Pohjola’s survey ‘Corporatism and Wage Bargaining’ shows that full employ-
ment may be best achieved by countries having either completely centralized
or extremely decentralized bargaining structures, while the intermediate
economies are likely to do much worse. The basic explanation is the following.
If labour markets are dominated by a moderate number of medium-sized
organizations, each big enough to enjoy considerable market power but small
enough to pass the costs of their actions on to the others, then they all face
a kind of prisoner’s dilemma. Although all employees would benefit from
wage restraint in some circumstances, any individual group might be sig-
nificantly worse off if it were to accept wage moderation while other groups
were able to obtain large pay increases. Aspects of the public goods problem
are also present: benefits of nominal wage restraint do not flow only to the
employees bearing its costs but also to the whole society in the form of a
reduction in inflation and an increase in output and employment. There are
two basic ways to deal with these kinds of externalities and, consequently,
to improve the employment performance: either internalize them by central-
izing the bargaining structure or remove them altogether by decentralizing
wage fixing. A number of reasons are given why the first alternative might
be socially more desirable than the second.
Pohjola also shows that this analysis has to be qualified in one important
respect. Even the countries with intermediate bargaining structures are likely
to achieve good employment performance under the normal circumstances
of stable economic growth. This is because the unions and employer federa-
tions are ongoing organizations: they confront each other from year to year
and are thus induced to behave strategically. They can then achieve tacit or
implicit agreement on wage formation without any formal arrangement if
they understand that cheating and free riding will be punished by the others
by their not adhering to the implicit agreement in the future. The conclusion
is that the bargaining structure may matter only in turbulent circumstances.

Various essays in the volume support the view that the oil crises are con-
comitant with the divergent economic performance in the 1970s and 1980s
of the countries with centralized and intermediate bargaining structures: the
first group did much better.
Rowthorn’s essay ‘Corporatism and Labour Market Performance’ adds
a new and important aspect to the analysis, namely the role of wage disper-
sion. He shows that one of the most striking features of decentralized labour
markets is the extremely large dispersion of earnings. Some workers earn
very high wages and enjoy a high degree of job security, whilst others—
mostly female workers—are crowded into low-paid, low-productivity, and
often insecure forms of employment. This phenomenon is clearly visible in
Japan and North America whose labour markets are segmented, having a
highly paid formal sector and a badly paid informal sector. The resulting
allocation of labour is inefficient and much of the employment in low-paid
jobs should properly be described as underemployment or disguised unem-
ployment. The employment performance of these countries, and to a certain
extent of Switzerland as well, is also influenced by the fact that unem-
ployment benefits are low, of short duration, and limited in coverage. In
these circumstances measured unemployment is made lower as job seekers
are compelled either to take low-paid jobs or to withdraw from the labour
market altogether. This makes it preferable to use the employment rate (that
is, employment as a percentage of working-age population) rather than the
unemployment rate as an indicator of labour market performance.
Under centralized wage bargaining this inefficiency and injustice could be
avoided by establishing broadly uniform wages across the whole economy
with some allowance for skills, unpleasant working conditions, and the like.
But as is shown by Rowthorn in his essay only the Nordic countries have
been able to achieve both high employment and low wage dispersion. Among
the corporatist countries, Austria’s labour market performance is much
worse: her employment rate has not been outstandingly high while wage
dispersion has been rather large. The important conclusion is then that while
centralized wage bargaining may be a necessary condition for good per-
formance it is not sufficient. This brings our attention to the importance of
the second criterion of social corporatism: to be successful corporatism has
to be of the non-exclusive and egalitarian type.
Pohjola’s and Rowthorn’s essays can be seen as a commentary on recent
developments in labour economics. Labour markets in real life are not
perfectly competitive. There are many imperfections, like the externalities
described above, which make the Walrasian theory inapplicable. Modern
labour economics recognizes this fact and there nowadays exist a number of
theories, such as implicit contract, insider-outsider, and efficiency wage
theories, which explain why the labour markets even in the absence of any
trade unions may deviate from full employment with equal wages. Pohjola’s
and Rowthorn’s essays show that if unions are brought into the picture in
the social corporatist framework, then the economy is in fact likely to move
towards the Walrasian ideal.
Not unexpectedly, the above general themes connected with corporatism
have to be qualified on the level of the experiences of individual countries.
This conclusion comes out in a clear way from the country studies ‘Cor-
poratism and Economic Performance in Sweden, Norway, and Finland’ by
Jukka Pekkarinen in Chapter 10 and ‘Corporatism: Success or Failure?
Austrian Experiences’ by Alois Guger in Chapter 11. In fact, as far as
economic performance since the mid-1970s is concerned, the corporatist
countries display considerable variety. There are countries like Sweden,
which over the period as a whole are quite generally considered as success
stories. Some corporatist countries, like Austria and, in some respects,
Norway, which often used to be praised as success stories have met difficult-
ies in the course of the 1980s. As is shown by Rowthorn’s labour market
analysis, Austria never did particularly well as far as employment, wage
dispersion, and the relative pay between men and women are concerned.
Moreover, the recent difficulties of the Austrian, as well as of the Norwegian,
economy partly reflect the fact that these countries did not manage very well
in the restructuring of their economies after the oil crises (see Michael
Landesmann’s essay in Chapter 8).
On the other hand, Denmark, which is often regarded as a failure on the
basis of its high rate of open unemployment alone, actually displays strong
labour market solidarity: her employment rate has been among the highest
and wage dispersion among the lowest. Finally, the Finnish economy has
succeeded quite well after the deep recession in the late 1970s, and her
corporatist structures have become stronger at the same time. But one may
wonder whether enhanced economic performance has been due to favour-
able structural background factors or to corporatism—in fact, as is suggested
by Pekkarinen in his essay, the strengthening of corporatism in Finland may
well have been a consequence of a strong economy rather than the other way
round. This reversal of causation and endogeneity of corporatism has also
been discussed by Pohjola in his essay in the volume.
Wolfgang Blaas’s country study on Switzerland in Chapter 12 entitled
‘The Swiss Model: Corporatism or Liberal Capitalism?’ demonstrates clear-
ly how mistaken it is to regard it as a corporatist country. The Swiss model
resembles more the paternalistic-liberal capitalism of the Japanese type.
Although the business community is strong and concentrated, labour is
rather weak. Wage bargaining does not take place at the national level, but
instead at the enterprise or plant level. The state plays only a minor role in
organizing social security which has traditionally been a matter of private
initiative. As measured by unemployment Switzerland’s labour market
performance has been excellent, and also rather good in terms of employ-
ment and wage dispersion. However, the Swiss system is of the exclusivist
type. It has been able to maintain full employment and a considerable degree
of wage equality among Swiss males through the exclusion of women from
the labour force and by increasing the flexibility of labour supply by adjusting
the pool of foreign workers. Blaas calculates that had the overall labour force
participation developed in line with other European OECD countries and
had the number of foreign workers remained at its 1973 level, Switzerland
would have had an unemployment rate of 9.8 per cent in 1986 instead of the
actual 0.8 per cent.
Corporatism has been in retreat in a number of OECD countries in the
1980s. There are various possible explanations for this fact, such as increas-
ing economic integration, the introduction of new flexible production tech-
nologies, and the increasing importance of public and service sector
employees. These emerging features tend to weaken the working relationship
between labour and capital. Although there are differences between coun-
tries, described in various essays in the volume, the trend has been almost
universal. There are, however, certain exceptions. Robin Archer describes
in Chapter 13, entitled ‘The Unexpected Emergence of Australian Corpor-
atism’, how Australia has recently established a Nordic-type centralized
wage-bargaining system in an environment being shaped by industrial rela-
tions of the British type.
The essay is important in that it addresses the question whether different
societies are free to accept or reject a corporatist system or whether they are
bound by tradition to their respective industrial relations systems. The
Australian experience has so far been promising and Archer is optimistic in
his account of the future of corporatism. The new system has already been
able to cope with the shift in policy-making from tackling stagflation by wage
restraint to industrial restructuring. Ireland is another economy which has
recently moved in a corporatist direction, following the establishment of a
‘National Wages Agreement’ between the government and trade unions as
part of an austerity package for dealing with the country’s gigantic foreign
debt problem. However, the Irish case is not discussed in this book.
1.3 DISTRIBUTION OF CONSUMPTION AND SAVING
In Chapter 5 Andrew Glyn approaches the question of labour market

performance from a somewhat different perspective. He analyses the process
of ‘consumption spreading’, whereby a given volume of market sector
output is shared out amongst the population. Some mechanisms for spread-
ing consumption achieve their result by increasing the total number of
people employed in the economy, which Glyn calls ‘employment spreading’.
One method for spreading employment is simply to hold down labour

productivity in the market sector, thereby increasing the total amount of
work available in this sector. Another is to create additional work in the
non-market (i.e. government) sector. A third method is to reduce the hours
worked by the average employee, so that any paid work available in the
economy is shared out more widely. All of these mechanisms spread con-
sumption by increasing the total number of people in paid employment.
Their effect is complemented by the impact of transfer payments, which
weaken the links between consumption and paid employment.
By comparing the corporatist countries with the EEC, USA, and Japan,
Glyn shows that in the years of the golden age before 1973 the first group
most clearly exemplified the process of consumption spreading through
increased government employment, leisure, and transfers. These factors
contributed to consumption spreading throughout the OECD, but their
combined effect was greatest in the corporatist countries prior to 1973. Since
1973 world economic growth has slowed down and the relative importance
of the various mechanisms for spreading consumption has altered somewhat.
Government employment, leisure, and transfers are still important, and in
most OECD countries growth in these items has continued to play a major
role in spreading consumption. However, these are not the items which
distinguish the corporatist countries since 1973. On average, the combined
contribution to consumption spreading of increased government employ-
ment, leisure, and transfers has been about the same in the corporatist
countries since 1973 as in the rest of the OECD. What distinguishes the
corporatist countries over this period is a quite different item, namely the
response of market sector productivity to slower output growth. All OECD
countries have experienced a marked slow-down in the growth rate of market
sector output since 1973. In many countries, labour productivity in the
market sector has continued to rise quite rapidly, with the result that the total
amount of work available in this sector has fallen dramatically. Labour
shedding has also occurred in the corporatist countries, but to a lesser
degree, primarily because of the slower growth of market sector labour
productivity in these countries. Comparing Belgium and Sweden, for ex-
ample, market sector output per head of population grew at an average rate
of 1.2 per cent and 1.3 per cent p.a. respectively over the period 1973–86.
The corresponding figures for hourly labour productivity were 3.5 per cent
and 2.4 per cent. As a result of this differential growth in productivity, the
total hours of work performed in the market sector (per head of population)
fell by 2.3 per cent p.a. in Belgium as compared to 1.1 per cent p.a. in
Sweden. Cumulated over such a long period, a difference of this magnitude
has enormous social implications. In Sweden, the slow growth of labour
productivity has helped to maintain the demand for labour in the market
sector and thereby keep down unemployment and spread consumption more
widely in the population. In Belgium, by contrast, relatively fast productivity

growth has meant a dramatic reduction of employment in the market sector,
a massive rise in unemployment, and a marked contraction in the spread of
consuming power. The slow growth of labour productivity in Sweden and
certain other corporatist countries since 1973 is in part a deliberate choice.,
reflecting a conscious desire to maintain employment in the market sector
as a way of preventing unemployment. It is a solidaristic mechanism for
spreading employment and consumption, which is unnecessary in good
times, but can play a vital role in bad times. Interestingly, though for
somewhat different reasons, productivity has also grown very slowly in the
market sector in the USA since 1973, which has been a major factor in
that country’s ability to generate so much employment in modern times. For
slow productivity growth to be a viable means of maintaining employment
and spreading consumption, it must, of course, be accompanied by a
correspondingly slow growth in real wages, which has been the case in both
Sweden and the USA.
An important aspect of labour market solidarity is the social security
system. Katri Kosonen draws in her essay ‘Saving and Economic Growth
from a Nordic Perspective’ in Chapter 6 a distinction between the Nordic
type and the Continental or, as she calls it, a ‘Bismarckian’ type of a welfare
state. She considers Austria as an example of the latter type. Historically, the
Nordic welfare states have opted for a wide coverage and egalitarian thrust
of the social security system. The public sectors in the Nordic countries
provide more services in kind than the Continental countries which rely
more heavily on transfer payments. The Nordic countries consequently have
a bigger share of the labour force in the public sector, which in turn is partly
reflected in their high participation rates for women.
A further distinctive characteristic of the Nordic countries, partly related
to the financing of social security, is the role historically played by public
sector saving in the ‘Nordic model of accumulation’ described by Kosonen
in her essay. After plummeting into deficit during the ‘bridge over’ period in
the late 1970s and early 1980s, the public sectors in the Nordic countries
now each run a convenient surplus. On the other hand the contribution of
the private sector to national saving in the Nordic countries is much lower
than in Continental Europe, Austria included, and its share has greatly de-
creased recently. The possible reflections of the recent decrease in household
saving on the current account deficits notwithstanding, the role of the public
sector as a financial surplus sector is intimately connected with the tradi-
tional growth-orientated profile of state intervention in the Nordic countries.
The egalitarian thrust of the low degree of wage dispersion in the Nordic
countries is further increased by the redistributive tax and social security
systems. On the other hand, and this is emphasized by Glyn and Pekkarinen
in their essays, the flexibility of the average real wage was rather high in the
Nordic countries in the late 1970s and real take-home pay was further
decreased by the increasing burden of taxation. This is a further aspect of
labour market solidarity characteristic to these countries. By reducing their
standard of living those employed shared the burden of adjustment and
made possible the creation of new jobs for those entering the labour force or
losing their jobs in the sectors most badly hit by the crisis. In many countries
with strong trade unions which failed to avoid the trap of high unemploy-
ment in the late 1970s and early 1980s, inflexibility of real wages and the
consequent profit squeeze contributed to the increase of unemployment.
In an exclusive system, certain marginal groups—women, foreign workers,
small enterprises, etc.—are discriminated against. They are rewarded more
meagrely than the privileged groups while they often have to bear the heaviest
burden of adjustment in the form of lower pay or exclusion from the labour
force. It is our interpretation of the views expressed in various essays of the
current volume that among the corporatist countries Austria displays certain
features of an exclusive system to a considerable degree.
1.4 LONG-RUN PERFORMANCE AND INDUSTRIAL POLICIES
It is implied by several essays in this volume that exclusive attention to the

relationship between corporatist institutions, the general level of wage settle-
ments, and demand for labour, notwithstanding the fact that it has largely
characterized economists’ recent interest in the topic, misses an essential
point. In the long run it is the capability of the corporatist countries to
restructure their economies and to foster technical progress which is decisive
for their abilities to combine full employment and equality with increase in
the standard of living. Two of the essays of the book, i.e. ‘Social Corporatism
and Long-Term Economic Performance’ by Michael Landesmann and
Juhana Vartiainen (Chapter 7) and ‘Industrial Policies and Social Corpor-
atism’ by Landesmann (Chapter 8), directly address this issue.
In these two highly complementary essays the connections between corpor-
atism, industrial policies, and long-term economic performance are scrutin-
ized at conceptual, empirical, and policy level. It is argued that due to a
smooth process of ‘natural’ catching up and relatively fast economic growth,
the awareness of the need for explicit industrial policies was weak in Austria
until the early 1980s. But difficulties piled up gradually in the late 1970s, as
heavy subsidies were used to maintain employment in existing industries and
restructuring was at the same time largely neglected. Landesmann regards
this as an important example of failure of corporatist arrangements. But most
recently the awareness of the importance of industrial policy has greatly
increased in Austria, the results of which are yet to be seen. Guger agrees
with this interpretation in his analysis of Austrian developments.
The papers by Landesmann and Vartiainen and Landesmann and

Pekkarinen amount broadly to the following interpretation of structural
change and industrial policies in the three Nordic countries. Like Austria,
Norway was also characterized by a defensive industrial strategy in the late
1970s. But a crucial difference was that Norway was able to rely on North
Sea oil. It was not until the mid-1980s that the fall in oil prices and a dramatic
weakening of the trade balance led Norway to move towards a more offensive
industrial policy to secure competitive foundations in its non-oil tradable
sector. As in the Austrian case, this turn has recently been favoured by the
international upswing but its final results remain to be seen.
The open sector of the Swedish economy was very badly hit by the
recessionary period of the late 1970s, but what we are witnessing now is a
relatively successful survival of the Swedish model. The essays offer two
explanations for this. First, for historical reasons, Sweden had a highly
concentrated and internationally orientated corporate sector which was able
to modernize its operations. Second, the state involvement in the restructur-
ing process in the late 1970s, while defensive in character, facilitated the
preconditions of restructuring through extensive labour market programmes
etc.; moreover, various selective support schemes and labour market
measures were promptly dismantled when the open sector started to recover
in the early 1980s.
Finland’s response to the international economic developments of the
1970s and 1980s represented a continuation of its attempts to promote
industrial development and to broaden her export base. In this process, the
role of state as producer and provider of educational, social, and technical
infrastructure has been profound, although it has not performed as such a
guarantee of full employment as in Austria, Norway, or Sweden.
A long-run point of view is also strongly present in Sixten Korkman’s essay
‘Exchange Rate Policy and Employment in Small Open Economies’ in
Chapter 9. A main thrust of his essay is an analysis of the relation of exchange
rate policy to inflation and employment. The discussion is motivated by the
policies of frequent devaluations the Nordic countries pursued in the late
1970s and early 1980s (and Finland still earlier). After surveying traditional
views in macro-economics according to which exchange rate changes will
have real effects only in the short or medium run, he gives an interpretation
of the accommodative exchange rate policies of these countries in terms of
their long-run growth policies. Here exchange rate accommodation is seen
to act as an insurance system for the profitability of the open sector and to
reduce Lancaster’s famous dynamic externality resulting in excessive real
wages.
Korkman assigns the accommodative exchange rate policies a pivotal role
in the corporatist policies of the Nordic countries. He also gives an account
of developments in individual countries. Towards the end of his essay he
draws attention to financial integration which tends to render these tradi-

tional policies unviable—a challenge the corporatist countries now face.
1.5 FURTHER REFLECTIONS ON THE CONCEPT OF

SOCIAL CORPORATISM
As was indicated at the beginning of this introductory chapter, in the course

of the work on this book the relevance of the distinction between exclusive
and inclusive types of corporatism emerged as a central lesson of the project.
This classification was developed at such a late stage that it is not reflected
in all the essays. Many authors use their own definitions. Consequently some
general concluding comments on the conceptual background of this distinc-
tion are in order.
Our usage of the term ‘social corporatism’ as referring to a centralized and
non-exclusive system of industrial relations is more specific than, for ex-
ample, the meaning originally given to it by Katzenstein (1985: 129) in his
classification of corporatist countries into social and liberal ones according
to the degree of centralization, the dominant social coalitions, and the
political tendencies of the government. In fact, quite in contrast to the kind
of approach adopted in various essays of this book, Katzenstein (1985: 127)
regards the search for consensus as characteristic of corporatism, both social
and liberal in his use of the term.
It seems that the distinction between conflict- and consensus-based sys-
tems of industrial orders has not been widely noticed in the corporatism
literature. In generalizations on the basis of the Continental experiences,
corporatism has rather been interpreted as a system based on consensus
prevailing between the leaders at the top of the interest organizations and
state bureaucracy while the internal conflicts among organizations are sup-
pressed through a strictly hierarchical order (Schmitter 1982; Lehmbruch
1982). On the other hand scholars like Walter Korpi (1983) who have found
this description unfitting to the Scandinavian, particularly Swedish, ex-
perience have rejected the notion of corporatism and instead referred to
Sweden as a case of a democratic class struggle. It seems that this controversy
is unnecessary and that the interpretation implied by this project, i.e. that
these two industrial orders represent varieties of corporatism, opens new
perspectives on the differences in the experience of different corporatist
countries.
The fact that various articles in this volume emphasize differences in per-
formance with respect to employment and equality between different cor-
poratist countries raises the question whether it is the institutional structures
or the goals adopted by interest organizations and economic policy makers
which matter for differences in performance. The distinction between the
institutionalization of conflict and that of consensus as the basis for industrial

order seems to imply that these two varieties of social corporatism are
historically wedded to different kinds of goals, i.e. that the former gives more
priority to the non-exclusive concept of equality than the latter. In fact the
emphasis on the influence of corporatism on the setting of goals is strongly
present in several articles of this volume. Therborn goes as far as to criticize
the corporatism literature for the implicit assumption that objectives do not
count. Certainly this criticism is not valid for this volume.
Rowthorn in turn suggests at the end of his essay that the institutionaliza-
tion of conflict reflects a way to accommodate the struggle by Labour for
democracy and equality. The acknowledgement of conflict as the basis for
compromise has fostered the mobilization of Labour and its emphasis on
inclusive and egalitarian ideology. This in turn has created, as Rowthorn
notes, an inherent dynamic towards equality. On the other hand industrial
orders based on consensus are more conservative and tend to preserve
inequalities which were in existence when these orders were established.
Managing of adjustment and sharing of burdens and rewards is controlled
by elites at the top of the hierarchies of interest organizations with an
exclusive thrust as a natural result. The conflicts are suppressed rather than
acknowledged.
But this shift of emphasis from the institutional structures to goals in turn
raises the question which is also dealt with in several of the articles of the
volume: is centralized wage bargaining then obsolete as a characteristic of
corporatism? It seems to be in the spirit of most of the essays that centralized
institutions do matter in various ways and that they in a considerable degree
account for the similarities between the countries of social corporatism.
Centralization facilitates the achievement of at least certain types of object-
ives. In general a high degree of centralization of interest representation
makes possible trade-offs and deals that may not be feasible in its absence.
Thus even in the game-theoretical perspective centralization of bargaining
may offer an institutional advantage in a turbulent situation where new,
previously unknown changes take place and where there is little use for
learning from experience. This may explain why the countries of social
corporatism did not deviate much from other European economies until the
mid-1970s when the economic turbulence started.
Apart from its appropriateness for co-operation, the centralized structure
of interest representation also facilitates state intervention in the form of
incomes policy and the like. For example, the notion of a general wage level
would be unthinkable as a policy instrument in a completely decentralized
system.
But it may be suggested that the importance and the rationale of
centralization may be different in the inclusive and exclusive, or the institu-
tionalization of conflict and consensus, types of corporatism. The former
seems to be historically associated with a strong thrust by strongly organized

labour for democracy and equality. Centralization may not be as essential
for the rise of consensus. Blaas gives in this volume an account of how a
consensus-based, highly exclusive Swiss system has worked under condi-
tions of highly decentralized bargaining. In fact it may well be that a high
degree of unionization and centralization tends to undermine the working of
a highly exclusive system in the long run. Temporarily at least, centralization
may perform as a tool to enforce exclusion. But in the course of time the
suppressive nature of such a system becomes more and more obvious. It
causes dissatisfaction among those who feel themselves to be discriminated
against. The result is either alienation with, for example, a sharp decrease in
union enrolment and the consequent weakening of the institutions of social
corporatism or a change in the union objectives to more egalitarian direc-
tion—that is to say, a shift from exclusive to inclusive corporatism. It is this
suppressive aspect of (the exclusive type of) corporatism some social scient-
ists, like Leo Panitch (1980), seem to have referred to when they analysed
the social instability of corporatism (see also the discussion in Rowthorn and
Glyn 1990).
Apart from the eventual social instability, there is another aspect through
which the distinction between exclusive and non-exclusive systems may be
relevant for the dynamic of corporatist societies. This concerns their capacity
to adjust the production structure to changing circumstances. This con-
sideration became especially important in the aftermath of the oil shocks
which rendered much of the production capacities of the open sectors of the
industrialized countries obsolete. It is implied by several essays in this
volume that some of the corporatist countries were clearly more successful
in restructuring than others. In this respect too the two extremes seem to be
formed by Sweden and Austria: while the former had largely succeeded in
solving by the mid-1980s the difficult structural problems that had met its
manufacturing industries in the late 1970s, Austria had rather pushed the
restructuring forward in time with a deepening structural crisis of many of
the branches in manufacturing as a consequence by the early 1980s. Norway
and Finland can be considered as special cases as far as restructuring is
concerned: in the former case it was closely linked with the rise of the oil
sector whereas the structural problems of Finnish manufacturing remained
minor, partly due to its relative backwardness—or the gains from the
catching up—partly due to the special link in the Finnish case between the
price of oil and the Soviet trade. But both these countries face a more difficult
challenge of restructuring in the years to come.
The fact that the ranking of countries with respect to their restructuring
capacity correlates with the grouping of countries into exclusive and inclusive
types of corporatism raises the question whether there is a causal relation
between the two. In particular, is the inclusive and egalitarian nature of the
industrial orders based on institutionalized conflict conducive to a strategic,

forward-looking type of industrial adjustment where problems are reacted to
before they become really severe? And does the consensus based on the
maintenance of the status quo in turn increase the likelihood of defensive
strategies, where old structures are maintained by subsidies and preventive
measures until the structural problems become unbearable with an acute
crisis as a result? The inherent logics of the two systems may indeed give rise
to such a difference. The implicit premiss of a high employment rate with
fairly equal pay characteristic of the inclusive type of corporatism may cause
a pressure to strategic, forward-looking behaviour in industrial adjustment,
whereas the groups privileged in the decision-making in a corporatist
economy of the exclusive type may well tend to preserve their own status and
shift the burden of adjustment to others. Apart from differences in the
capacity to adjust to new circumstances, this difference in the fundamental
constraints of industrial adjustment may be reflected, for example, in the
greater emphasis on active manpower policies prevalent in the inclusive
corporatism of the Swedish type. This, together with a high level of general
education, in turn reinforces the low degree of wage dispersion in an
inclusive corporatism, whereas the tendency to an unequal sharing of the
burden of adjustment inherent in an exclusive type of corporatism is con-
comitant with big wage differentials. At least during deep structural crises,
an exclusive type of corporatism is more prone to adopt dualistic solutions
ruled out by the solidarity principle adhered to in an inclusive system.
1.6 THE FUTURE OF CORPORATISM
The corporatist economies are in the process of change. Their relationship

to the outside world is altering and their internal modes of operation are
being transformed. These developments have already created considerable
stress in some corporatist economies and will continue to do so in the future.
Opinions differ about the long-term viability of the corporatist model, but
all agree that it will face a serious challenge in the coming decades. The
difficulties emanate from a number of sources. Externally, the barriers
separating the corporatist countries from the rest of the capitalist world are
breaking down, and their scope for economic autonomy is being reduced.
Internally, some of these countries face difficulties because the trade union
movement, which has traditionally been a force for coherence and order, has
become more fragmented or weakened by the centrifugal tendencies of
modern economic development. In response to such difficulties, the state in
these countries is being drawn ever more actively into areas, like wage fixing,
which previously were often the preserve of private bodies such as employers’
and workers’ organizations.
The integration of the corporatist economies into the capitalist world

economy is nothing new. They have always had extensive foreign trade with
other countries and their societies have been profoundly shaped by this fact.
Indeed, corporatism itself can be seen as a particular kind of institutional
response by small open economies to the requirements of competition in
world markets (see Katzenstein 1985). What is new in the present situation
is the rapid extension of integration into spheres which were previously
rather insulated from the outside world. For example, in most corporatist
countries, capital markets were until recently highly regulated, but have now
been or soon will be almost entirely deregulated. There will be virtually no
restrictions on the movement of funds across the frontiers of these eco-
nomies, and the owners of funds, whether domestic or foreign, will be free
to transfer their money to wherever they can obtain the highest return.
As Korkman points out in his essay in Chapter 9, such a development
makes it very difficult to pursue the kind of low interest rate policy often
used in the past to promote industrial investment in the corporatist coun-
tries. Any attempt to reduce real interest rates substantially below those
available elsewhere will lead to an outflow of funds and to an eventual
collapse in the exchange rate. Moreover, the freeing of capital movements
greatly increases the danger of speculative panics, making economic policies
a hostage to the opinions of international finance. To reduce this danger,
most corporatist countries are now seriously considering joining the
European Monetary System (EMS). Such a step will provide valuable
external support against exchange rate speculation, but it will also mean a
further decline in policy autonomy as the price of remaining in the EMS. In
particular, membership of this system will make it more difficult to use
devaluation as a deliberate policy for promoting competitiveness, as some of
the corporatist economies have successfully done in the past. Thus, financial
integration will reduce the scope for active interest and exchange rate policies
designed to stimulate industrial investment and enhance international com-
petitiveness.
Another area where integration has potentially major implications for
national autonomy is the multinational corporation. In some corporatist
countries, such as Finland and above all Sweden, national firms are rapidly
extending their overseas operations and many of them will soon produce
merely a fraction of their global output at home. Meanwhile, outside firms
are investing in these countries. As a result, the importance of purely national
firms is diminishing, and production and foreign trade are becoming increas-
ingly dominated by multinational corporations. Such corporations differ
from purely national firms in several major respects, connected with their
greater freedom of action on a world scale and their different global priorities.
The fate of the purely national firm is closely linked to that of its domestic
economy. Since it produces only at home, any increase in its own production
contributes directly to national output, whilst an increase in its sales abroad

adds directly to national exports. This will not automatically be the case for
a multinational corporation, which may raise its global output or sales abroad
by increasing production overseas, neither of which is of immediate benefit
to the domestic economy. Moreover, a purely national firm will be propor-
tionally more affected by a slump in the domestic economy than will a
multinational corporation with extensive global interests to cushion it. For
a variety of reasons, therefore, the interests of purely national firms are more
closely tied to the fate of the domestic economy than are those of multina-
tional corporations. As a result, national firms may be more willing than
multinational corporations to enter institutional arrangements which involve
sacrificing part of their autonomy in return for a better growth performance
in the domestic economy. Multinational corporations can always ‘exit’ from
a situation they do not like, whereas national firms are trapped and have no
alternative but to seek a ‘voice’ in domestic affairs. The rise of multinational
corporations may also weaken corporatist structures by creating a greater
diversity of interest on the employers’ side, thereby increasing the difficulty
of forming stable coalitions amongst employers and undermining the
authority of their central organizations.
Where it occurs, disintegration on the employers’ side will make it increas-
ingly difficult to maintain the kind of long-term working relationships
between capital and labour which have typified the corporatist economies
until now. Moreover, these difficulties may be compounded by a parallel
disintegration on the workers’ side. In most corporatist countries, economic
developments are undermining the authority of the old centralized workers’
federation, either because these organizations are fragmenting or else be-
cause their coverage of the work-force is declining. There are various forces
making for fragmentation. The traditional leaders of the trade union move-
ment, which gave it coherence and direction in the past, were the pre-
dominantly male, manual industrial workers who originally created the
strong centralized trade union federations characteristic of the corporatist
economies. The influence of these workers has declined rapidly over the past
twenty years as the expansion of private and public services has given rise
to a whole variety of new jobs and raised dramatically the proportion of
women in the work-force. Workers in the private services are frequently not
unionized, and even where unionization of private service workers is the
norm, as in Sweden, there have often been serious problems in forging a
common policy with traditional industrial workers.
In some countries, there has also been considerable friction between
private and public sector unions, especially between private industrial
workers and public service workers. In part, this is a male-female conflict,
but it also reflects other factors. Public service jobs are usually more secure,
and often less onerous, than private sector industrial jobs, so workers in the
latter quite reasonably expect higher wages in compensation. However, their

case is frequently not accepted by public service workers, especially profes-
sionals, who have been traditionally privileged and expect to keep their
privileges. The result has been considerable friction and the emergence of
wage militancy by public sector professionals, fighting to preserve their
traditional status. Another source of friction between workers in the public
services and those in private industry has been the different basis of their
financing. Workers in private industry exist in a world market environment
and the ability of their firms to increase wages depends on the amount of
competition they face abroad in export markets or at home from competing
imports. Public service workers, however, are paid out of taxes and the ability
of their employers to raise wages is determined in the first instance by the
government’s tax and borrowing policies. The latter are, of course, affected
by general economic conditions, but only indirectly and often with a consid-
erable delay. The insulation of public service workers from immediate
economic conditions tends to make them less willing than those in private
industry to accept wage restraint as the price of better economic perform-
ance.
A final factor tending to fragment the trade union movement is the spread
of plant-level bargaining and wage incentive systems in the private sector. In
some cases, this is the result of new kinds of production process, as ex-
emplified by the shift from standardized, ‘Fordist’ mass production methods
to more flexible and heterogeneous ‘post-Fordist’ techniques in manufactur-
ing industry, which are less amenable to centralized wage fixing. In other
cases, it is an unexpected result of the industrial democracy movement of
the 1970s, which gave workers new powers at the plant or firm level, which
they have used to establish new forms of local wage determination, thereby
reducing the influence of centralized bargaining. Thus, on the workers’ side
there are a number of factors tending to increase fragmentation. The rise of
private and public services has reduced the influence of the traditional
manual industrial workers and created new frictions within the organized
labour movement, whilst the growth of new payments systems is undermin-
ing the unity of traditional industrial workers themselves. Thus, the tendency
towards fragmentation on the employers’ side is matched by an equivalent
tendency on the workers’ side.
One possible outcome of these tendencies towards fragmentation is that
corporatist arrangements will simply collapse and the corporatist countries
will become like countries in the European ‘core’, with stronger trade unions
maybe, but with a similar incoherence in capital-labour relationships,
especially wage determination. A more likely possibility is that the respons-
ibility for coherence will move increasingly away from employers’ and
workers’ organizations towards the state, which will replace, reinforce, or
supplement private, centralized wage bargaining with various kinds of official
intervention. Indeed, the state already performs this kind of role on an

extensive scale in some corporatist economies, such as Denmark and Nor-
way where government incomes policies have been of great importance in
recent years. Even where formal incomes policies are avoided, the state may
still play a vital role in setting the agenda and orchestrating agreement
between somewhat fragmented employers’ and workers’ organizations
which are unable to reach agreement on their own. This role is also likely to
be backed up with various sanctions to force dissidents to participate in the
process or else abide by the outcome. None of this means that centralized
workers’ and employers’ organizations will disappear altogether or become
irrelevant. What it means is that economic development is tending to weaken
or undermine such organizations, and that the state will find itself increas-
ingly involved in shoring them up and, where required, supplanting or
supplementing them. In some corporatist countries this involvement is
already well established; in others it is merely a tendency. How successful
such involvement will be is an open question. Whilst it may be effective in
the short run, action by the state to shape or steer employers’ and workers’
organizations may be self-defeating in the long run, since it may simply
undermine the legitimacy of these organizations and reduce their ability to
mobilize support for government policies. Moreover, attempts to bypass or
supplant these organizations may cause resentment and an eventual ex-
plosion of discontent, especially on the side of workers. Thus, the drift in
corporatist economies towards greater state intervention in areas like wage
bargaining faces certain inherent limits, and it may turn out that such
intervention cannot adequately offset the tendency towards fragmentation
in employers’ and workers’ organizations. Only time will provide an answer
to this question.
As mentioned above, the corporatist economies are being integrated into
the world capitalist economy through both financial markets and multi-
national corporations. This process is augmenting the threat potential of
both financial and industrial capital over individual corporatist countries,
and indirectly therefore over the workers in these countries. Since capital of
all kinds can now move across frontiers with relative ease, the individual
corporatist economies will be forced to provide conditions for investors
which are competitive with those elsewhere. At the margin, and allowing for
exchange rate and other kinds of risk, expected returns on both financial and
real investment will have to be similar to those obtainable elsewhere. Other-
wise capital will simply boycott the country concerned. The power of either
workers’ organizations or the state to reduce these rates of return through
purely national action will therefore be limited. The individual corporatist
country will have to pay capital roughly what is available elsewhere or face
the consequences. Any substantial reduction in the return to capital will have
to come through some kind of international action to reduce interest rates
or avoid competitive tax reductions. This raises more general questions

concerning international co-operation and the evolution of political rela-
tionships between the corporatist countries and the rest of Europe. Most
important here is, of course, the European Community. One corporatist
country, Denmark, is already a member of the Community and over the
coming decade certain others may join. Even those which do not formally
join will move economically and politically closer to it and bring their own
internal economic policies more closely into line with those of Community
countries. This in itself, in addition to some of the developments described
previously, will tend to reduce the differences between the corporatist and
other West European economies. However, the convergence will not be only
in one direction. Developments like the Social Charter, with its emphasis on
workers’ rights and the role of trade unions, suggest that neo-liberalism is on
the wane and that the European Community may be moving in a broadly
Social Democratic direction, taking on features which are traditionally
associated with the corporatist countries. It is difficult to predict what will
happen over the long term, but it seems likely that differences between the
various Western European economies will decline over the next twenty years.
Even so, the present-day corporatist countries may still have certain special
features which mark them off from the rest; in particular they will almost
certainly have much stronger trade unions than most of the rest of Europe,
and they will probably have a greater degree of centralized wage fixing than
elsewhere. Moreover, those corporatist economies in the Nordic group will
continue to be significantly more egalitarian than most other West European
countries.
1.7 ON THE TRANSFERABILITY OF THE CORPORATIST MODEL
A final intriguing question concerns the transferability of corporatist institu-

tions between countries. There are obviously reasons why certain countries
have developed such institutions, in particular centralized wage bargaining,
whilst others have not. It is also clear that corporatist institutions cannot be
arbitrarily transplanted between countries independently of local conditions.
However, as the recent example of Australia indicates, there are conditions
under which transplantation is possible. As Archer describes in his essay in
this volume, Australia has recently developed a system of centralized wage
bargaining which, although building on pre-existing institutions, has been
consciously shaped in the light of Nordic experience.
Nordic-style corporatism also has many admirers in the Eastern Bloc, not
least amongst Gorbachev’s advisers, some of whom have explicitly taken
Finland or Sweden as examples. The more optimistic, both in the Soviet
Union and Eastern Europe, believe it is feasible to refashion the Communist
countries along corporatist lines, with a vigorous market economy, an activist

state, a strong and independent trade union movement, together with a good
welfare state and safety-net for the casualties of perestroika. One major
difference perhaps would be in the realm of property rights, where some of
the Eastern admirers of the corporatist model envisage a mixed economy
where the mix has a greater public component than is normal in the West.
‘Privatization’ of industry, in their view, would mean transferring property
rights to a variety of autonomous public and semi-public bodies, leaving
individual shareholders with a less important role than they typically enjoy
in the West. However, the difference should not be exaggerated. In certain
corporatist countries, like Finland for example, the state and other public or
semi-public bodies, such as co-operative banks, trade unions, and local
authorities, are substantial shareholders in the nation’s industry, and in these
countries the mix already has a large public component and is not so very
different from what Eastern admirers of corporatism would like to create.
The importation of some variety of Nordic-style corporatism has obvious
attractions for the countries of Eastern Europe. It would be considerably
more dynamic than what they have now, and probably more egalitarian as
well. However, in most cases there are serious obstacles in the way of such
a transformation, which make it seem unlikely, if not impossible. On the plus
side, a number of countries in Eastern Europe have experienced political
revolutions which have mobilized enormous popular enthusiasm for change.
These revolutions, especially where they are allied with nationalist senti-
ment, have created a new mood of solidarity amongst the population and a
willingness to accept personal sacrifice for the greater good. This democratic
idealism provides just the kind of environment in which centralized workers’
federations of the Nordic variety could be created, and a co-operative
relationship established in each country between workers, the government,
and the employers (whoever the latter might be!). Indeed, Poland, where
Solidarity is actively collaborating with the government, already provides an
example of such a set-up.
The real problem, however, lies not in the creation of centralized workers’
federations, but in their effective long-term survival, and in the maintenance
of a co-operative relationship between these federations and other economic
actors. This is turn depends on what happens to the countries of Eastern
Europe over the next few years. These countries are facing severe economic
difficulties, their industries are antiquated, and most of them have gigantic
foreign debts. To overcome these difficulties in a reasonably egalitarian and
orderly fashion they require massive aid from the West. It seems improbable
that such aid will materialize on anything like the required scale. Moreover,
much of the limited aid they do receive will be conditional upon the
implementation of economic policies which are brutal in their impact and
profoundly divisive. For example, in return for IMF loans equivalent to less
than six months’ interest on its foreign debt, Poland was compelled to
implement a Latin American style ‘stabilization’ policy in late 1989, which
it was officially predicted would lead to an increase of around 1 million in
unemployment, together with a substantial reduction in real earnings.
One view is that these measures will quickly revitalize the East European
economies, sweeping away outdated industrial structures and production
methods in an orgy of creative destruction, clearing the decks for a dramatic
economic recovery in the near future. If the Latin American experience of
stabilization policies is anything to go by, this view is unduly optimistic. A
more credible scenario is that these countries will face some years of travail
and privation before they eventually emerge from the morass. Moreover, the
economic policies they follow during this period will be highly in egalitarian.
Some people will get rich, others will be impoverished, and the rest will
somehow get by. Such developments will have an extremely divisive effect
on the social fabric of Eastern Europe. It is difficult to see how the current
enthusiasm and idealism in Eastern Europe can survive years of privation
and growing inequality, and there is an obvious danger that the recent mood
of consensus and restraint will disintegrate into conflict and bitterness, as
the fruits of sacrifice do not materialize and disillusioned groups mobilize to
pursue their sectional interests. This outcome would be the very antithesis
of Nordic-style corporatism and the countries concerned would then have
more in common with Argentina than Sweden.
Even if things never quite reach this stage, the present mood of economic
quiescence in Eastern Europe is unlikely to last. The consensual arrange-
ments established during the democratic revolution will come under increas-
ing economic pressure in the coming years and a considerable amount of
conflict and fragmentation seems inevitable. For most East European
countries, the corporatist model, despite its obvious attractions, does not
seem a viable option at the present time. One possible exception is Czecho-
slovakia, whose political traditions and better average economic situation
may allow it to organize a relatively smooth and orderly transition to some
variant of Nordic style corporatism.
A country not considered so far is the Soviet Union itself. Despite
admiring remarks from some of Gorbachev’s advisers, the likelihood of
the Soviet Union transforming its economy into anything resembling
Nordic-style corporatism in the foreseeable future is remote. The country is
racked by ethnic and national conflicts, and the degree of social cohesion
required for the creation of viable, independent corporatist institutions does
not exist on a country-wide level, although it might conceivably exist in some
of the individual republics should they break away from the Union and set
up on their own.
1
The figure for unemployment is an IMF estimate as revealed to Polish parliament’s social
affairs committee by the Minister of Labour, Jacek Kurom (Guardian, 5 Jan. 1990).
The general conclusion from all this seems to be that corporatist institu-
tions can in principle be transplanted from one country to another, with
appropriate modifications, but conditions must be right. With one or two
possible exceptions, conditions do not appear to be right for such a trans-
plantation in Eastern Europe or the Soviet Union, where economic life is
likely to be relatively disorderly and conflictual in the foreseeable future.
REFERENCES
BRUNO, M., and SACHS, J. (1985), The Economics of Worldwide Stagflation, Oxford:
Basil Blackwell.
KATZENSTEIN, P. J. (1985), Small States in World Markets: Industrial Policy in Europe,
Ithaca, NY: Cornell University Press.
KEYNES, J. M. (1927), The End of’ Laissez-Faire, London: Hogarth.
KORPI, W. (1983), The Democratic Class Struggle, London: Routledge & Kegan Paul.
LEHMBRUCH, G. (1982), ‘Introduction: neo-corporatism in comparative
perspective’, in G. Lehmbruch and Ph. C. Schmitter (eds.), Patterns of Corporatist
Policy Making, London: Sage.
PANITCH, L. (1980), ‘Recent theorizations of corporatism: reflections on a growth
industry’, British Journal of Sociology, 31: 159–87.
PRYOR, F. L. (1988), ‘Corporatism as an economic system: a review essay’, Journal
of Comparative Economics, 12: 317–44.
ROWTHORN, B., and GLYN, A. (1990), ‘The diversity of unemployment experience
since 1973’, in S. A. Marglin and J. B. Schor (eds.), The Golden Age of Capitalism,
Oxford: Oxford University Press.
SCHMITTER, PH. C. (1982), ‘Reflections on where the theory of neo-corporatism has
gone and where the praxis of neo-corporatism may be going’, in G. Lehmbruch
and Ph. C. Schmitter (eds.), Patterns of Corporatist Policy Making, London: Sage.
WILLIAMSON, P. J. (1989), Corporatism in Perspective, London: Sage.
2
Lessons from ‘Corporatist’ Theorizations
Göran Therborn
2.1 THE ‘CORPORATIST’ PARADOX
‘Corporatism’ has become a bit like God. Many people believe it is an import-
ant phenomenon, crucially affecting social life. But nobody really knows
what it looks like, so disagreement persists, apparently for ever, about what
it is and about what it does. Everybody can lay claim to it for whatever pur-
pose. Between believers and non-believers no criteria seem capable of settling
the argument, which, however, does not render impossible a civilized and
interesting discussion about such things as the meanings, the history, and
the functions, and even the referent, of corporatism-belief.
The paradox of ‘corporatism’ is its wide attraction as a synthesizing desig-
nation and as an explanatory variable to scholars of several disciplines, polit-
ical science, sociology, contemporary historiography, and economics, while
at the same time remaining a moving target, accorded a number of different
definitions and specifications. That ‘corporatism’ and its effects are being
studied is clear, but what it is that is then being studied is not.
The two main subvariants of corporatism are still those drawn up by
Philippe Schmitter and Gerhardt Lehmbruch (Schmitter and Lehmbruch
1979; Lehmbruch and Schmitter 1982), as centralized and monopolistic
interest ‘intermediation’ and as public policy-making by concertation be-
tween the state and interest organizations, respectively. Each has a number
of sub-sub variants. Economists have been attracted primarily to the former,
political scientists and sociologists studying political economy to the latter
(see further Therborn 1988 with extensive references).
Elsewhere (Therborn 1986: 98 ff.; 1988) I have argued that corporatism,
in either main conceptualization, is not very fruitful to an explanation of
cross-national differences in unemployment in the 1974–85 crisis, both be-
cause of the limitations of union-wage explanations and because of the con-
ceptual unclarities of corporatism.
A recent addition to the attraction of corporatism has been the launching
of a hypothesis of a hump-shaped relationship between corporatism and
(un)employment in the latest crisis. Intermediately corporatist, in this case
intermediately centralized, systems of wage bargaining are argued to be the
worst performers in cross-country comparisons. A nag though, at least to
some empirically minded social scientists, might be that the authors do not
LESSONS FROM ‘CORPORATIST’ THEORIZATIONS 25
find any ‘clear evidence in favour of a non-monotonic relation between cen-

tralization and macroeconomic performance’ with the help of price and wage
equations, nor much relationship between their elaborated theoretical argu-
ment for a hump-shaped relationship and the actually existing world
(Calmfors and Driffill 1988: 30, 41). While they do find (pp. 20, 22) that
countries they classify as having an intermediately centralized system of
wage bargaining generally performed worst, their empirical evidence does not
include anything about wages. If it had, they could have found, for instance,
that the three ‘intermediate economies’ for which data are available (Belgium,
Germany, The Netherlands) in 1973–85 had after Japan, in the Belgian case
also after Austria, the lowest wage increases in relation to labour productivity
of the OECD area (OECD 1987: 48, 90).
However, it would be naive or presumptuous to assume that a paradox of
enduring conceptual confusion and widespread analytical attraction among
sophisticated social scientists could be dismissed as a deplorable error. More
likely is that the corporatist theorists have hit upon something important, the
full and precise implications of which have not yet been brought out, neither
by them nor by their more or less heretical followers. Without any pretension
to providing a solution to the paradox, this chapter intends to locate the
‘corporatist’ impetus into a broader context of social theory concerns and to
put forward some argued ideas about relevant empirical relationships and
processes of determination.
2.2 THREE ‘CORPORATIST’ CONTRIBUTIONS TO SOCIAL THEORY
The hard core of contemporary ‘corporatism’ can be summed up in: ‘interest

organizations are important, and they have a particular, variable relationship
to contemporary states.’ This simple sentence, seemingly banal to the point
of, perhaps, invoking controversy for lack of subtlety, is not meant as a
put-down, but as a baseline. The ‘corporatist’ contributions bring something
to bear upon three central areas of social theory, social action, social order,
and social steering.
‘Corporatism’ contributes to the theory of social action by asserting,
implicitly or explicitly, the strength and the long-run consequences of asso-
ciative action. Furthermore, ‘corporatism’ directs our attention to important
variations in the settings and the forms of associative action. Secondly
‘corporatism’ has always implied a certain kind of social order, although
originally more empirically delimited than theoretically elaborated. Recently,
however (Schmitter 1985), a specific ‘associative model of social order’ has
been spelled out, in contrast to market, state, and community.
‘Corporatism’ finally involves an argument about the steering of society,
the guidance of a social (dis)order to certain goal states. This problematic is
26 GÖRAN THERBORN
there from the outset in Lehmbruch’s subvariant, with its consistent focus
on concerted policy-making. But it has also a home base in Schmitter’s (1981)
thesis of corporatist interest intermediation as a mechanism of governability.
The relative value of these contributions is difficult to pin down in any
very precise manner. The corporatist theories of associative action assume
or imply that the dilemmas and the problems posited by what has become
known as the ‘theory of collective action’ (Olson 1965; Hardin 1982) are
solved or irrelevant. Or, ‘encompassing’ or ‘comprehensive’ associative action
is regarded as a solution to the disruptive effects of ‘free-riding’ less-than-
encompassing organizations. The implicit assumption, then, is that (relative)
organizational size is the main determinant of the forms of associative action,
which in cross-national research or in the history of industrial relations can
hardly be taken for granted. The relationship of the associative model of
social order to die ‘consensus’ invoked by many corporatist theorists and
analysts (Bruno and Sachs 1985; Katzenstein 1985; McCallum 1986; Schmidt
1982, 1986) remains unclear. Crude counterpositions with ‘pluralism’ apart,
‘corporatist’ steering has not been inserted into any explicit theory of steering
or guidance (see e.g. Lehmbruch 1984, 1985; as a contrast cf. the overview
in Kaufmann et al. 1986).
On the other hand, the ignorance of the ‘corporatist’ thrust among general
sociological theorists is amazing. For instance, neither in Giddens (1984),
in Giddens and Turner (1987), nor in Alexander et al. (1987) is there any
discussion of corporatist action, order, and social steering. Neglect does not
hold for the whole discipline, however. In Goldthorpe (1984), for example,
‘corporatism’ features highly when sociologists and their invitees try to come
to grips with order and conflict in contemporary society. But clearly, the
uncertainties of where ‘corporatism’ fits in and becomes important are not
just to be blamed on the ‘corporatist’ theorists, but also, and perhaps more,
on the non-responsiveness of general social theorists.
Anyway, our problem in this context is rather the continuing elusiveness
of corporatism as an explanatory variable in cross-country studies of eco-
nomic performance. The basic reason for that, this chapter will argue, is that
corporatism, however defined, mainly refers to a pattern of strategic action.
Only in limiting cases may the notion refer to an institution with certain fixed
objectives and behavioural rules. What is usually referred to under the rubric
of corporatism has a fluidity in time and social space, which makes for end-
less controversies about empirical indicators and about their values for any
but a few countries. The objectives of the actors in socio-economic systems,
corporatist or non-corporatist, are typically left implicit in the corporatist
literature, which makes the outcomes of corporatist patterns of action
largely indeterminate. In order to make any advances from the corporatist
debate so far, advances in the direction of finding socio-political factors con-
tributing to explaining variations of macro-economic outcomes, the notion
of corporatism should be unbundled and sorted out from the perspective of

strategic action.
2.3 ASSOCIATIVE ACTION AND TYPES OF CLASS FORMATION
Corporatism, of whatever definition, presupposes, derives from, a particular

kind of social action. That is from associative action, from people coming
together in enduring formal organizations made up (predominantly) of
members sharing a common purpose or interest. Associative action is a major
variant of a broader category of collective action, which also includes more
ephemeral or fleeting collectivities, such as informal insidership, with con-
comitant insider-outsider demarcations, crowds, and not formally bounded
movements. Collective action stands in contrast not only to individual
action, but also to commanded action determined by another, exterior actor,
a patron, a sovereign, or a boss of one sort or another.
The distinction between associative action and commanded action was
first systematically spelled out by Alexis de Tocqueville in 1848: ‘Every-
where, where at the head of a new enterprise in France you would find the
government and in England a grand seigneur, count on seeing an associa-
tion in the United States’ (Tocqueville 1986: ii. 155). Advanced countries
still differ considerably in the amount of associative action by their citizens.
A cross-national compilation of studies of associative membership shows,
for the late 1950s and early 1960s, the highest density of associative action
in Sweden, followed by the other Nordic countries, accompanied by the
United States with regard to non-economic association and followed also by
Austria as far as economic association is concerned. Low values of associa-
tion characterized France and Italy, and rather low West Germany (Pestoff
1977: 168).
Now, corporatism is interested mainly in one basis of eventual associative
action, by workers and, though sometimes more obliquely, employers. In
other words, sotto voce, corporatism is about class organization, by labour
and capital. This means that corporatist organization is hardly reducible to
the anodyne variables of size—‘encompassing’ or not—and power locus,
centralized or not, which economists have tended to focus on in their
attempts at incorporating corporatist theorization into economic analysis,
and at linking it up with Olsonian perspectives on collective action and/or
game-theoretical views of strategic action. Without denying the great intel-
lectual contributions of the two latter, the broader context of the referents
of corporatism has to be brought forward and kept in mind.
Firstly, current forms of action and organization by the actors on the
labour market are bound up with, crucially affected by, the historical
outcomes of vast processes of social transformation, by the formation of the
28 GÖRAN THERBORN
classes of industrial capitalism. Secondly, as class organizations labour

market organizations are intrinsically adversary, that is they were formed
with a view to dealing with a known counterpart. This means, in turn, that
the action of one (aggregate of) class (members) and its organization (s)
strongly affects and is affected by the actions of the other. Thirdly, through
their direct link to the always ongoing work and accumulation process, both
capital and labour have a permanent capacity of serious social disruption and
of financially disrupting the polity. Therefore, their mode of action has a
special relationship to the state, to the ultimate guardian against social
disruption and the sovereign claimant of the resources of the territory.
The connection of current labour market organizations to historical in-
dustrial class formation explains, for example, a large part of the affinity of
Swiss, Austrian, German, and Nordic trade unions in spite of vast differences
in size, associational monopoly, and organizational centralization. Or of the
difference between US and French unionism, the two least encompassing
among the rich OECD countries, both having unions competing with each
other and both having union confederations with little central bargaining
power.
The labour movements of the contemporary advanced capitalist countries
have been formed in one of three basic moulds, plus the late, post-1945
Japanese form, with its characteristic importance of company unionism,
among other things (cf. Shalev 1988). According to the locations of their
historical origins, in the late nineteenth-early twentieth century, we may
simply call them the Anglo-American, the Latin, and the Germanic kinds of
labour movement. They differ in the approach to working-class organization,
particularly to party-trade union relations, differences which in turn derive
from cross-country differences in the relations between industrialization and
state politics.
Anglo-American labour organization centred on trade unions, gradually
developing with capitalist relations of production and industrialization. A,
for its time, relatively open polity, and defeats for early, little-organized
working-class politics (such as Chartism), postponed the issue of political
organization till later, in the USA seemingly for ever. The trade union
structure was enduringly engraved by the enormous variety of sites of
immediate class confrontation, and union action has concentrated very
strongly on wages and work-place conditions. The later added parliamentary
representation, the Labour Party, has added a dimension of parliamentary
politics, without much altering the trade union movement, however. Neither
British-style parliamentary politics nor American presidentialism puts much
emphasis on enduring political associative action.
Latin working-class politics developed in the course of the Great Revolu-
tion of 1789–93 and of the ensuing nineteenth-century revolutions. That is,
well before any significant industrialization and in a context of trade union
prohibition. From this common revolutionary heritage, modern socialist

politics came to set itself off from bourgeois radicalism in terms of ideology
of ends but rather little in form of association, still much based on intellectual
parliamentary notables. Trade unions came to develop later, with a great
deal of suspicion of this sort of politics claiming to represent the working
class. Trade unions often came to regard themselves as a kind of industrial,
non-parliamentary working-class party, but also drawing upon revolutionary
history. The latter meant that association as such was not overwhelmingly
important. What also mattered, and to many Latin trade unionists much
more, was the preparation and the preservation of militants and militancy
for the right moment of radical action (for a recent illustration of this, French
steel unions, see Rhodes 1985: 202). Put in other words, what mattered
more were other forms of collective action than associative action, such as
the revolutionary crowd or non-associated movement.
In the Germanic lands a persistent ancient regime type of polity and
rather late but then rapid industrialization and capitalization provided the
basis for a third kind of labour movement. Working-class parties—repres-
enting a new kind of political association, based on firmly structured mass
organization—arose almost simultaneously with trade unions, as part of one
common effort at class organization. The party had the upper hand, but
co-operation was close.
The Germanic labour movement is propitious to participation in ‘cor-
poratism arrangements for two fundamental reasons: (1) more than the
others it tends to create a centralized and uniform trade union organization,
as an expression of an effort at class unification; (2) because of the closeness
of industrial and political association, it is easier for a Germanic labour
movement to make politico-economic deals.
In Germanic countries with a polity significantly different from the
German-Austrian model another ancient social force furthered a politico-
economically unified labour movement with a broad societal vision: the
Catholic Church, and to some extent also the Calvinist one. The very
existence and strength of these forces of socio-political power meant a severe
limitation to a liberal polity of the Anglo-Saxon or French post-Restoration
types, and their vigorous response to beginning autonomous class organiza-
tion by the workers carved up the working class into ideological ‘pillars’ of
organizations, Social Democratic, Catholic, and Calvinist. Such was, to a
large extent, the working-class formation in Holland, (German) Switzer-
land, and, via Flanders and Flemish influence, Belgium (cf. Righart 1986;
Therborn 1989a).
The outcome was labour organizations of basically the same kind as in
Germany, with the, in other contexts than this one certainly non-negligible,
difference that the Social Democratic current had to coexist with a Catholic,
and in Holland and Switzerland also with a Calvinist, one.
30 GÖRAN THERBORN
There is also another reason why ‘corporatism’ is alien, and therefore

usually very controversial and/or fragile, to the Anglo-American world. That
is the legal tradition. The common law tradition of these countries had great
difficulties in dealing with ‘combinations’ in any positive way, geared to the
safeguarding of individual freedoms as it is. The famous British Taff Vale
Settlement of 1906 is an extreme case in point, making the unions exempt
from the, otherwise restrictive, law (cf. Phelps-Brown 1983: ch. 11). The
Continental tradition of public law could easily, and did in fact, positively
provide unions and employers’ organizations with public functions, rights,
and duties. This is clear in particular in the cases of Austria, Belgium,
Germany, The Netherlands, and Switzerland, either in public law institu-
tions of collective bargaining and paritary co-operation (as in Austria,
Germany, and The Netherlands) or in the explicitly formulated and some-
times used possibility of declaring a collective agreement legally binding for
all concerned, organized or not. The latter is stipulated by Belgian (Slomp
and Mierlo 1984: ii. 100) and Swiss (Katzenstein 1984: 145) law, of 1968
and 1947, respectively.
In addition the French legal and administrative traditions provide for an
institutionalized trade union presence, which would hardly be expected from
the limited and unstable membership of French unions and from the hostility
of many French employers to collective bargaining, a resistance second only
to that of American employers. There are regular ‘professional’ elections of
personnel delegates, supervised by the judiciary, and in which the unions run
slates of candidates. An old system of local arbitration leads to another
system of regular industrial elections (élections prud’hommales), in which the
unions operate like some kind of industrial parties (cf. Adam 1983).
Associative action by workers qua workers, i.e. trade unions, is a universal
phenomenon of all democratic countries and of all countries that at least
de facto tolerate autonomous associative action. At least in all democratic
countries there is also a national, inter-trade union centre, and quite a
number of countries have more than one such centre. Associations of
employers as employers are usually derivative from workers’ actions, and are
not universal to the same extent. In the USA there is no central organization
of employers at all, for instance. The reason, of course, is that workers are
more dependent upon collective action as a source of strength than
employers. For the latter, the enterprise is the prime organization (cf. Offe
1985). On the whole, employers’ associations have developed to the extent
that the owners/directors of enterprises have felt unable to deal with workers’
grievances and demands on their own. National inter-industrial associations
of employers have come into being and have been invested with repres-
entative authority only when employers have been confronted with a national
challenge, either directly by working-class industrial organization or more
indirectly via issues of governmental politics. Anglo-Saxon employers have
had neither, nor the Japanese, the West German, nor the Swiss, even though
central organizations of employers do informally co-ordinate employers’
collective bargaining in the latter countries (cf. Sisson 1987; Windmuller and
Gladstone 1984; Katzenstein 1984). Labour government-union accords
seem not to have been very successful in promoting national employers’
unity in Australia, although associative action has increased among them (see
further Archer in this volume).
Centralized employers’ organizations divide into two subgroups, accord-
ing to their historical origin. One subgroup comprises Denmark, Finland,
Norway, and Sweden, the other Austria, Belgium, France (in so far as there
is, intermittently, employer centralization), Italy, and The Netherlands.
State politics has been decisive to both. Without involvement of national
state politics, no employers’ unification. But one route has been political-
cum-industrial association, the other has been an imbrication of industrial
association with structures and concerns of state.
The Danes were the pioneers of national employers’ organizations. In
1898 they set up one that united a large number of small employers, mainly
located in and around the capital (which strongly dominated the non-
agrarian economy) and facing a vigorous labour movement. The initiator
was a constructor-cum-Conservative politician. Upon his instigation the
Norwegian Employers’ Association was set up in 1900. Sweden had larger
enterprises than these countries, and enterprises much more diversified and
dispersed. But the Swedish employers were drawn together by the 1902
strike for universal suffrage. Originally started as an industrial arm of First
Chamber political reaction, the SAF soon became an organization of purely
employers’ interests. Till it started producing its own house-apprenticed top
functionaries, the Swedish Employers’ Association was built and forcefully
led by former civil servants, none of whom had ever had any private
entrepreneurial or managerial experience (see further Therborn 1989b). The
Finnish employers came together in 1907, but Scandinavian patterns of
collective bargaining date from the end of World War II in Finland.
Whereas the Nordic pattern is one of employers associating by transfer-
ring experiences and techniques of mobilization and organization-building
from parliamentary politics and the civil service, the other one is employer
centralization at state instigation. The Austrian Federal Economic Chamber
is governed by public law and is a species apart and between a state body
and a voluntary association. The Joint Commission, where the deals between
capital and labour are struck, is an officious, though legally unregulated,
body (Katzenstein 1984: 59 ff.). Belgian ‘interprofessional’ bargaining was
brought about by the state, under intense working-class pressure for social
change, in 1918 and in 1936, and was recognized by the employers volun-
tarily after World War II, but state direct intervention has remained impor-
tant (Slomp and Mierlo 1984: ii. 208 ff.), and only the experience of the war
32 GÖRAN THERBORN
and the occupation brought the Wallonian and the Flemish employers
together (Slomp and Mierlo 1984: i. 84 ff.). The French employers’ con-
federation was first set up in 1919, at the behest of the Minister of Commerce
(Bunel and Saglio 1984: 236). Neither on the side of business nor on that
of labour has it been possible to establish any unity around broad collective
bargaining, and only in particular socio-political conjunctures has nation-
wide top bargaining taken place, generally with the government involved one
way or the other, in 1936, in 1968 and its aftermath, after Mitterrand’s
victory in 1981. The Italian Confindustria is perhaps the employers’ organ-
ization most similar to the Nordic types, going back to 1910. The centralized
pattern of collective bargaining and the powerful position of Confindustria
derive from the organization’s successful deal with the Fascist regime and
from the legacy of the Fascist form of corporatism (Treu and Martinelli
1984: 265 ff.). The first Dutch employers’ organization arose (in 1899)
with a view to opposing an Industrial Accidents Bill, but no unity ensued.
Major unification and centralization came only after World War II in the
context of a process of politico-economic reconstruction, and centralized
bargaining developed in a framework of legal labour market regulation
(Voorden 1984).
Rather than relations to the world market or the form of trade union
organization, it is relations to politics, political conjunctures, and concerns
of state which have shaped the associative action of employers. So, on both
sides of the class divide, the organizations of capital and labour are decisively
shaped by different historical trajectories and by the relation of both to the
national state. A small set of historically shaped and, by and large, still
reproduced types of labour movements, and the relative power balance
between capital and labour in society, emerge as the two most important
general variables. When both labour and capital are centralized and all-
encompassing or when there is little (successful) associative action and no
central co-ordination, the popular variables of economic corporatists may
yield some confidence-inspiring predictions of likely industrial and eco-
nomic outcomes. But the variety of settings between the poles of Austria and
the United States does not seem graspable with relative size and centraliza-
tion. And, as we shall see below, the latter may even be insufficient in such
cases too.
Historical forms and relations of power, historical or current, only set the
stage. They do not determine action. They make certain courses more or
less likely, difficult, or stable. Herein the two variables interact and should
both be taken into account. The fact that what corporatist theorists refer to
derives from patterns of associative action means that labour and capital
action strategies and forms of bargaining, co-operation, or conflict are not
fixed labels which can be easily stuck to the names of countries and then run
in correlations and regressions. Between the more frozen poles of Austria
and the USA, labour market action varies considerably within the same
country from one year or short period to the next (see further Armingeon
1983; Elvander 19886; and the yearly overview of collective bargaining in
Western Europe, put out by the European Trade Union Institute in
Brussels). The implication of this for empirical investigations into the wage
outcomes of bargaining patterns is that annual (or other short-time periodic)
data of the national bargaining format should be used as the independent
variable. Such a procedure would also have another advantage, of consider-
ably increasing the number of observations.
2.4 PATTERNS OF INDUSTRIAL (DIS)ORDER
In the corporatist literature patterns of associative action have often been

conflated with their hypothesized ordering effects. Phrased in a different way
organizational forms have been conflated with consensus. It is a merit of
Calmfors and Driffill (1988: 24 ff.) that they take issue with that confusion.
On the other hand, neither they nor the corporatist tradition have systematic-
ally addressed the question of different patterns or systems of industrial order
or, for that matter, disorder.
One group of countries has an extensive pattern of procedural rules, also
respected de facto, governing collective bargaining and possible industrial
conflict. They are all countries with what we above called a ‘Germanic’
labour movement. Another pair of states has an elaborate system of public
arbitration.
The world’s first comprehensive national labour market agreement, the
Danish September Agreement of 1899, ended a conflict in the construction
industry that had broadened to an almost general lock-out in the urban
trades. In working out the settlement, which provided the model for later
ones in Norway and Sweden, not only the leaders of the employers’ associa-
tion and the trade union confederation were involved, but also the Conser-
vative Prime Minister, representatives of the banking community, the liberal
bourgeoisie, and the Social Democratic Party leadership (Bertolt et at. n.d.:
254 ff.).
Norway got its Main Agreement in 1935, Sweden in 1938, Switzerland
got a pace-setting agreement in the export industries of watch-making and
machine-making in 1937. In 1958 the same kind of agreement had reached
the construction sector (Parri 1987: 46). During and in the immediate
aftermath of World War II, as expressions of national solidarity, important
capital-labour agreements were made in Finland, The Netherlands, and
Belgium. The German Legien–Stinnes agreement of 1918 had broken down
even before the breakdown of the Weimar Republic, and in the Federal
Republic collective bargaining was institutionalized by law, in 1949. Austria,
34 GÖRAN THERBORN
whose centralized labour and capital organizations and conservative cor-

poratism had gone to civil war in 1934, sealed its post-World War II
understanding with the Raab–Bohm agreement of 1956, which, among other
things, installed the famous Joint Commission.
The French have twice attempted something similar, without success.
First there was the Matignon agreement of 1936, struck in the shadow of the
electoral victory of the Popular Front and the wave of factory occupations in
June 1936, and almost immediately disavowed by a dominant faction of the
employers (Lefranc 1974: 216–17). A similar background, the May events,
led in 1968 to the Grenelle agreement, which, although not repudiated,
never stuck fully (Bunel and Saglio 1984: 241 ff.). In democratic Italy no
industrial relations constitution has ever made it, either to the statute book
or to a comprehensive collective agreement.
In the Anglo-Saxon world, a centralized legal system of arbitration was set
up in New Zealand in 1894 and in Australia in 1904. It gives these countries
a pattern of industrial order which is neither corporatist, in the usual varieties
of that notion, nor decentralized or market-dominated. Its effectiveness in
bringing about its official aim of conciliation has been limited, as appears
from the high strike record, but it certainly orders industrial relations. In
the USA (the Wagner and Taft–Hartley Acts), Canada (with provincial
Wagner-type acts), Britain with the Thatcherite Employment and Trade
Union Acts, and in post-World War II legislation in Eire, and in Japan,
unionization and trade union action are given a framework of minimum
protection, and, in the case of the Taft–Hartley and Thatcher Acts, certain
legislated restraints. In neither of these later cases, however, is an industrial
order of mutual class recognition and orderly conflict provided for (cf.
Armingeon 1988: 168 ff.; Sisson 1987; Windmuller and Gladstone 1984).
Centralized peak-to-peak wage bargaining is not the same as a bargained
order of negotiations. The former occurred nowhere before World War II.
It started in Norway and in a state-guided form in The Netherlands right
after World War II, and began in Sweden in 1952. In the Nordic countries
it became a pattern in the 1950s and 1960s, as well as in Holland and
Austria. Such patterns of centralized negotiations have not been rare in
post-war Belgium, and they have occasionally occurred in Italy and France.
Except for Austria, this pattern has come under increasing strain in the 1970s
and 1980s. Such negotiations have never taken place nor have such agree-
ments been made in Germany, Switzerland, the Anglo-Saxon countries, and
Japan (Armingeon 1983; Elvander 1988b; Sisson 1987; Windmuller and
Gladstone 1984).
The level at which negotiations take place, if at all, seems to depend
mainly, though not exclusively, on employers’ decisions under different
conditions and considerations of power (cf. Sisson 1987). But national
politics and national adaptations to world market competition can also
produce a sense of the nation as one field of negotiation. Post-war central

negotiations in Sweden were pushed through by the employers’ confedera-
tion, but there were demands for centralization also from the (export
industry-dominated) metal workers’ union, which wanted to hold back the
wages of the best-paid domestic industry unions (Rehn 1980). Currently,
after the failure of the big lock-out in 1980, the leading export industry
employers in Sweden are strongly against centralized negotiations, which are
now defended mainly by the unions and by the domestic industry employers.
There is no universal political economy explanation of bargaining levels.
State-guided ‘incomes policies’ have been attempted, with varying suc-
cess, in a number of countries, notably Anglo-Saxon and Germanic
(Elvander 1988b; Flanagan et al. 1983; Archer in this volume). In centralized
Sweden and in decentralized Switzerland, incomes policy is an official taboo,
on the other hand (cf. Elvander 1988a; Katzenstein 1984).
Throughout the post-World War II period, six countries have consistently
had fewer working days lost to industrial dispute than others: Switzerland,
Austria, The Netherlands, Germany, Sweden, and Norway (Korpi 1983:
165; ILO 1986: tables 3 and 28; Shalev 1983). Two of these countries have
competitive unions (The Netherlands and Switzerland), which between
them organize about a third of the workers, one (Austria) has a single
encompassing trade union movement, and the other countries fall in be-
tween. During most of the period Holland had ideologically multiple
employers’ organizations as well. In the mid-1970s–mid-1980s decade,
Japan, which for the whole 1946–76 period had more industrial conflict than
Britain, added itself to the very low conflict league, and so did Denmark,
although at a level above that of Switzerland and Austria. All the time in the
twentieth century Switzerland has been an ultra-peaceful case apart from the
crowd (Korpi 1983: 165) (the defeated general strike of 1918 excepted),
recently accompanied by Austria.
At the other end, as countries of persistently high volumes of industrial
conflict in the twentieth century, are (non-Fascist) Italy, Australia, and
France. Rather large amounts of industrial conflict have also periodically
occurred in Finland, New Zealand, the UK, Canada, Ireland, and Japan.
Occasional big and/or widespread conflicts have taken place in all the
industrialized countries since World War II, except Switzerland, Austria,
Holland, Germany, and Norway. No neat institutional explanation can be
given to this pattern of industrial peace and conflict. The most peaceful
countries all have Germanic-type labour movements, but relative peace
ensued only after pre-World War II trials of strength between capital and
labour. Latin and Anglo-Saxon labour movements and employers are the
most conflict-prone, but the evidence also indicates that such countries with
strong labour movements but no or brittle governmental influence on labour
tend to have most industrial conflict.
36 GÖRAN THERBORN
The relatively peaceful industrial orders are structured in fundamentally

two different ways. One is an institutionalization of partnership and consen-
sus, with formalized bodies of joint decision-making and an extensive legal
back-up. The Netherlands is the prototype of this industrial order, which
had its days of glory from Liberation till the mid-1960s. A ‘guided wage
policy’ governed, and the legal right to industrial conflict was severely
curtailed. There was no associational monopoly, however. Both employers
and employees were divided along legitimate ideological lines into three
confederations each (cf. Windmuller et al. 1983).
Though much less legally formalized, Austria may be located on the Dutch
(now considerably attenuated) side. Switzerland and West Germany also use
the ‘partnership’ vocabulary, but Germany (at least) leans more to the other
end of the peace-conflict spectrum. This second type is organized not in
institutions of consensus and partnership but around an institutionalization,
one might perhaps even say ritualization, of conflict. The Scandinavian
countries, and since World War II Sweden in particular, have developed this
regulation of conflict, which dates back to the Danish September Agreement
of 1899. Here a ‘guided wage policy’ would have a ring similar to ‘guided
democracy’. Capital and labour are not ‘social partners’ but ‘labour market
parties’, whose autonomy and principled conflict rights are held sacrosanct.
Conflict rights in the Swedish public sector are extensive, together with
those in Finland, said to be the most extensive in the world (Elvander 1988a:
318), although the conflict right is bestowed only upon confederations (of
manual workers, white-collar employees, and professionals). Sweden is
probably unique in according the lock-out the status of a constitutional right
(§ 17 of the 1974 Constitution), a weapon occasionally used and always
considered possible and legitimate. The multi-employer lock-out also exists
in Germany, but there it is more contested and legally hedged in. It is not
alien to the industrial relations of the other Nordic countries, but it is rarely
used, whether as a threat or as a practical move, outside Sweden and
Germany (cf. Gladstone 1984:41; Bunn 1984: 194). What has kept Swedish
industrial relations in peace has been a duopolistic balance of power between
two major actors, both taking a rational long-term view, the manual workers’
confederation, LO, and the employers’ association, SAF. Since the mid-
1970s this duopoly of two industrial superpowers has been broken up by the
rise of new actors with an industrial force de frappe, the white-collar con-
federation and, in particular, its resourceful union of private industrial
employees, and the public sector unions, both manual and white collar. This
has led to a number of serious industrial conflicts in the last decade, and
Sweden looks bent upon leaving the industrial peace club. Conflictual
legitimacy and a very strong concern with distributive justice have led to this
erosion of settlement by bargaining only.
By contrast, the Swiss industrial relations system has always harboured a
large number of actors while avoiding conflict. The latter appears to have
less legitimacy in Switzerland, but probably more important are the weak-
ness of the unions and the lesser concern with distributive justice.
There are, then, different kinds of industrial order. Centralized peak-level
bargaining between capital and labour is not the key factor (cf. Pohjola in
this volume). Such negotiations have never occurred in two of the six
outstandingly peaceful industrial orders after World War II, in Switzerland
(cf. Blaas in this volume) and West Germany. They have occasionally
occurred, and led to agreement, however, in the most conflictual of all the
advanced industrial nations, Italy. Nor is consensus very apt as a charac-
teristic of all peaceful industrial relations systems, given the fact of two very
different kinds of industrial cultural systems, one formed around partner-
ship, the other around orderly regulated positive-sum (always potential)
conflict.
What has mattered, it seems, is a socially broad-minded union move-
ment—as a Germanic-type trade union is more or less from its origin—with
a social vision which may be later acquired by political experience, a reaching
of a power balance between labour and capital (which may occur at different
points of relative power, depending upon the point of departure), and an
experience on the parts of labour and capital of belonging to the same
national culture, between the members of which some amount of trust is
possible and between whom some aspects of destiny are common. The
shadows of Fascism and/or foreign threat were decisive or at least significant
in all the most successful and enduring peace settlements. National unity
may be reached under less threatening circumstances, but no simple reshuf-
fling of labour market institutions stands much chance of succeeding in
bringing about new industrial relations. The latter are inextricably em-
bedded in relations of class and nation.
2.5 ASSOCIATIONS AND SOCIAL STEERING: DO POLITICAL

OBJECTIVE AND OTHER INSTITUTIONS MATTER?
Social steering is a concept for relating the political system to broad issues
of social dynamics and refers to attempts by social actors, not only govern-
ments, to affect the dynamics of the development of a society, to steer the
latter into a particular direction and/or at a particular pace. ‘Corporatism’
as concertation (Lehmbruch 1984) or as ‘private interest government’
(Schmitter 1985) implies a mechanism of social steering. Again implications
of and alternatives to ‘corporatism’ have generally been left hanging. But
much can be learnt from a large number of penetrating, often cross-nationally
comparative studies of interrelation between interest organizations and state
authorities on a local or industrial sector level that the corporatist impetus
38 GÖRAN THERBORN
has come to spawn (see e.g. Cawson 1985; Grant 1985). However, at the
same time the general theoretical issues raised or implied need to be pushed
further (cf. Cawson 1986).
The economic literature on corporatism, established as a legitimate dis-
ciplinary genre by Bruno and Sachs (1985) but on track in political science
at least since Schmidt (1982), asserts that corporatism has a major, benefi-
cial, effect on macro-economic steering. It is argued to lead to less unemploy-
ment and to better unemployment-inflation trade-offs, lower ‘misery index’.
Here we have reformulated the conflated corporatist problematic into one
of strategic associative action and its consequences. Clearly, associations
have played, and still play, an important role in twentieth-century social
steering. The most general part is probably the one played in the monitoring
and support of agriculture since the 1930s, where farmers’ associations in all
industrial countries constitute a crucial part of the public steering endeavour
(cf. Peterson 1979).
The contemporary importance of associative action does not limit itself
to involving states and private associations with each other. Associative
action, of coalition-building, interest mediation, and bargaining, is also
increasingly found to characterize the internal operation of public ad-
ministrations. Such was, for instance, the conclusion of the Norwegian
Power Study (NOU 1982: iii. 53) and of Hanf and Scharpf (1978). Even
though there are good reasons not to forget that ‘government is not just
another organization’ (Sharpe 1986: 177, emphasis omitted), to the extent
that the former conclusions are correct, ‘the associative model’ reaches
beyond the state–market–association distinction and points to politico-
economic steering processes of an intricate complexity. Legal traditions
make a difference between states in this respect. For example, the strong
legal Rechtsstaat tradition in Germany leaves public authorities with consid-
erably fewer discretionary powers and less room to manoeuvre in bargaining
than, say, Sweden (labour market authority, Gösta Rehn oral communica-
tion) or the USA (regulation of toxic chemicals, Schneider 1985).
Another direction into which steering issues of state-associations relations
might take us is a perspective on the latter combining both economic and
social policy and structure aspects. ‘Corporatism’ has primarily centred
around macro-economics, labour markets, and industrial policy. On the
other hand, imbrications of states and private associations are important
phenomena and cross-national variables also in social policy and welfare
state organization. In the Latin as well as in the Germanic countries of
Europe, trade unions and employers’ organizations are directly involved in
social policy-making, and in many countries also in the implementation of
public social policy, in the running of public social insurance systems, for
instance. Trade unions run unemployment insurance, with mainly public
finance, in some countries. Denominational organizations play a major part
in the running of schools and hospitals in some countries, in Holland and

Belgium in particular. Help for the elderly and several other social services
are sometimes run by private charity associations of various kinds, but largely
financed by the state, the former constituting para-statal organizations. Such
‘voluntary service’ quasi-public organizations can be found not only in
Europe but also in, for instance, the USA (cf. Franz et al. 1986: 533). A not
insignificant social policy debate today centres around the question whether
a hiving off of social services from the state to private associations should be
promoted. From various ideological positions—religious, marketeering,
Green—the latter looks like an attractive option. However, the implications
of public services by private association are rather obscure and uninvest-
igated. The parallels with ‘corporatism’ or ‘private interest government’ in
economic and industrial areas are obvious, once stated. But so far, they seem
to have been little noticed.
Here, however, the issue is corporatism, or associative action and associa-
tive order, and macro-economic steering. Above we have already questioned
the meaningfulness of the standard ‘corporatist’ variables of centralization,
size, consensus, etc. In other words, the usual explanatory variables in institu-
tional explanations of macro-economic performance have been called into
doubt. But, from the perspective of steering the economy, the main question
is another one. That is, do objectives count?
Corporatist explanations implicitly assume they do not. Only institutional
arrangements, instruments of policy, are taken to count. The implicit assump-
tion seems to be that all governments and major economic actors have the
same socio-economic preferences. This assumption runs counter to capital-
labour conflicts of interest and to the political manifestations of the latter.
Elsewhere I have argued extensively (Therborn 1986) that the institu-
tionalization or not of the objective of full employment was the decisive
explanation for the maintenance of more or less full employment in some
countries during the 1970s–1980s crisis. The goal was the same in the five
persistent full-employment countries, but the means differed considerably,
as did the motivations of the major political actors.
That the same sort of labour market institutions can have vastly different
effects depending upon the objectives to which they are wedded is also
underlined by the different outcomes of centralized wage bargaining in
Sweden and Austria with regard to wage dispersion. In fact, Sweden and
Italy, with very different systems of labour market institutions, are the two
equalizers of wage differentials, whereas Austria rather increases differentials
(Hedstrom and Swedberg 1985; Rowthorn, and Guger, in this volume). In
stressing this difference between corporatist Austria and corporatist Sweden,
Rowthorn departs from the usual framework of corporatist analysis.
The corporatist idea of macro-economic steering has an unwarrantedly
narrow focus also in another respect. Corporatist economists assume that
40 GÖRAN THERBORN
only wages and union behaviour matter in setting levels of (un)employment.

Other institutions are, tacitly, held unimportant. The empirical evidence
against that assumption, and in favour of the importance of central banks,
public employment leeway, and gender relations affecting women’s be-
haviour with regard to the labour market, is overwhelming (cf. Scharpf 1987;
Therborn 1988; and Glyn in this volume).
2.6 LESSONS AND FUTURES
In one sense, the lessons of corporatist theorizing are strong and straight-
forward. Most of it has run into a dead end of confusion and neglect of
alternative perspectives. An nth rerun of corporatism and whatever variable
between the OECD countries is hardly likely to add very much to our
understanding of anything. Corporatism, in any of its usual variants of
definition, is not an adequate variable capable of accounting for industrial
order or macro-economic performance. On the other hand, corporatist
theoretical endeavours do contribute to a social scientific understanding of
reformulated issues of associative action, industrial order, and socio-
economic steering. However, ‘corporatism’ does not deserve to survive as an
explanatory variable. The major lessons of corporatist theorizations are their
side effects.
The types of industrial associative action, which so far have characterized
industrial relations of the industrial countries, have come under pressure
from different trajectories of post-industrial development. The two polar
types of industrial peace, the Dutch and the Swedish, have both already run
into trouble. In Sweden, hitherto alone among the classical countries of
industrial peace, a significant amount of industrial conflict is becoming
endemic.
On the other hand, the legacy of industrial society is most likely to keep
an enduring impact. In the foreseeable future, national post-industrial soci-
eties are likely to remain distinguishable by the form and order of the
associative action of the given national industrial society. Relative de-
industrialization has not (yet) meant any disaggregation of the industrial
class organizations, in particular not in the countries which since World War
II have had a successfully institutionalized industrial order, i.e. the countries
most often called corporatist. But in the longer run, new, post-industrial
settlements have to follow. Public service employment and public sector
unions have already begun to rearrange the industrial order, not the least in
Sweden. Though their impact is still limited in most countries, high-
technology industries and private services pose a serious challenge to the
unions everywhere. Industrial orders have so far been national (or sub-
national). If and when major corporations become fully de-nationalized—with
trans-nationally composed and mobile headquarters—a new era of industrial

relations has opened. However, we are not there yet, and it would be unwise
to jump into far-flung speculation. The patterns we have discussed here are
likely to remain for the rest of this millennium, and it is as strains and
challenges to them that new tendencies had better be analysed and dealt with
by politico-economic actors. And, like the present and the past, the future
will be plural.
REFERENCES
ADAM, G. (1983), Le Pouvoir syndical, Paris: Dunod.

ALEXANDER, J., etal. (1987) (eds.), The Micro-Macro Link, Berkeley, Calif.: Univer-
sity of California Press.
ARMINGEON, K. (1983), Neo-korporatistische Einkommenspolitik, Frankfurt: Haag und
Hurchen.
_____ (1988), ‘Politische Regulierung industrieller Beziehunge: von Kaiserreich zur
Bundesrepublik’, in Schmidt (1988).
BERGER, S. (1981) (ed.), Organizing Interests in Western Europe, Cambridge:
Cambridge University Press.
BERTOLT, O., et al. (n.d.), En bygning vi rejser, i, Copenhagen: Fremad.
BEYME, K. VON (1982), Parteien in zvestlichen Demokratien, Munich: Piper.
Basil Blackwell.
BUNEL, J., and SAGLIO, J. (1984), ‘Employers associations in France’, in Windmuller
and Gladstone (1984).
BUNN, R. (1984), ‘Employers associations in the Federal Republic of Germany’, in
Windmuller and Gladstone (1984).
CALMFORS, L., and DRIFFILL, J. (1988), ‘Bargaining structure, corporatism and
macroeconomic performance’, Economic Policy, 6: 14–61.
CAWSON, A. (1985) (ed.), Organized Interests and the State, London: Sage.
____ (1986), Corporatism and Political Theory, Oxford: Basil Blackwell.
ELVANDER, N. (1988a), Den svenska modellen, Stockholm: Allmänna förlaget.
____ (1988b) (ed.), Förhandlingssystem, inkomstpolitik och arbetskonflikter i Norden,
Stockholm: Norstedts.
FLANAGAN, R. J., SOSKICE, D. W., and ULMAN, L. (1983), Unionism, Economic
Stabilization, and Incomes Policies, Washington, DC: Brookings Institution.
FRANZ, H.-J., et al. (1986), ‘Associations and co-ordination’, in Kaufmann et al.
(1986).
GIDDENS, A. (1984), The Constitution of Society, Cambridge: Polity Press.
____ and TURNER, J. (1987) (eds.), Social Theory, Cambridge: Polity Press.
GLADSTONE, A. (1984), ‘Employers associations in comparative perspective: func-
tions and activities’, in Windmuller and Gladstone (1984).
GOLDTHORPE, J. (1984) (ed.), Order and Conflict in Contemporary Capitalism, Oxford:
Clarendon Press.
42 GÖRAN THERBORN
GRANT, W. (1985) (ed.), The Political Economy of Corporatism, Basingstoke: Mac-

millan.
HANF, K., and SCHARPF, F. (1978) (eds.), Inter-organizational Policy-Making: Limits
of Co-ordination and Central Control, London: Sage.
HARDIN, R. (1982), Collective Action, Baltimore: Johns Hopkins University Press.
HECHTER, M. (1987), Principles of Group Solidarity, Berkeley, Calif.: University of
California Press.
HEDSTRÖM, P., and SWEDBERG, R. (1985), ‘The power of working class organiza-
tions and the inter-industrial wage structure’, Journal of Comparative Sociology, 26:
90–9.
ILO (1986), Yearbook of Labour Statistics, Geneva.
KATZENSTEIN, P. (1984), Corporatism and Change: Austria, Switzerland and the Politics
of Industry, Ithaca, NY: Cornell University Press.
____ (1985), Small States in World Markets: Industrial Policy in Europe, Ithaca, NY:
Cornell University Press.
KAUFMANN, F. X., et al. (1986) (eds.), Guidance, Control, and Evaluation in the Public
Sector, Berlin: de Gruyter.
LEFRANC, G. (1974), Histoire du Front Populaire, Paris: Payot.
LEHMBRUCH, G. (1984), ‘Concertation and the structure of corporatist networks’,
inGoldthorpe(1984).
____ (1985), Neo-Corporatism in Western Europe: A Reassessment of the Concept in
Cross-National Perspective, Paris: IPSA XIII Congress.
____ and SCHMITTER, P. (1982) (eds.), Patterns of Corporatist Policy-Making, Lon-
don: Sage.
MCCALLUM, J. (1986), ‘Unemployment in the OECD countries in the 1980s’,
conomic Journal, 96: 942–60.
MAIER, C. (1984), ‘Preconditions for corporatism’, in Goldthorpe (1984).
NOU (1982), Maktutredningen, iii, Oslo: Universitets förlaget.
OECD (1987), Historical Statistics 1960–1985, Paris.
OFFE, C. (1985), ‘Two logics of collective action’, in C. Offe, Disorganized Capitalism,
Cambridge: Polity Press.
OLSON, M. (1965), The Logic of Collective Action, Cambridge, Mass.: Harvard
University Press.
PARRI, L. (1987), ‘Staat und Gewerkschaften in der Schweiz 1873–1981’, Politische
Vierteljahresschrift, 28.
PESTOFF, V. (1977), ‘Voluntary associations and Nordic party systems’, Diss. Stock-
holm University: Dept. of Political Science.
PETERSON, M. (1979), International Interest Organizations and the Transmutations of
Postwar Society, Stockholm: Alqvist & Wicksell.
PHELPS-BROWN, H. (1983), The Origins of Trade Union Power, Oxford: Clarendon
Press.
REHN, G. (1980), ‘Idéutvecklingen’, Lönepolitik och solidaritet, Stockholm: LO.
RHODES, M. (1985), ‘Organized interests and international crisis management:
restructuring the steel industry in West Germany, Italy and France’, in Cawson
(1985).
RIGHART, H. (1986), De Katholieke zuil in Europa, Amsterdam: Boom Meppel.

SCHARPF, F. (1987), Sozialdemokratische Krisenpolitik in Europa, Frankfurt: Campus
Verlag.
SCHMIDT, M. (1982), ‘Does corporatism matter? Economic crisis, politics and the
rise of unemployment in capitalist democracies since the 1970s’, in Lehmbruch
and Schmitter (1982).
____ (1986), ‘Politische Bedingungen erfolgreicher Wirtschaftspolitik’, Journal fur
Sozialforschung, 26: 251–73.
____ (1988), Staatstätigkeit, Opladen: Westdeutscher Verlag.
SCHMITTER, PH. (1981), ‘Interest intermediation and regime governability in con-
temporary Western Europe and North America’, in Berger (1981).
____ (1985), ‘Neo-corporatism and the state’, in Grant (1985).
____ and LEHMBRUCH, G. (1979) (eds.), Trends toward Corporatist Interest Intermedia-
tion, London: Sage.
SCHNEIDER, V. (1985), ‘Corporatist and pluralist patterns of policymaking for
chemicals control: a comparison between West Germany and the USA’, in Cawson
(1985).
SHALEV, M. (1983), ‘Strikes and the crisis: industrial conflict and unemployment in
the Western nations’, Economic and Industrial Democracy, 4: 417–60.
____ (1988), Class Conflict, Corporatism and Comparison: A Japanese Enigma,
Jerusalem: Dept. of Sociology, Hebrew University.
SHARPE, L. (1986), ‘Inter-governmental policy-making: the limits of subnational
autonomy’, in Kaufmann et al. (1986).
SISSON, K. (1987), The Management of Collective Bargaining, Oxford: Basil Blackwell.
SLOMP, H., and MIERLO, T. VAN (1984), Arbeidsverhoudingen in Belgie, 2 vols.,
Utrecht: Het Spectrum.
THERBORN, G. (1985), ‘The coming of Swedish Social Democracy’, in Annali della
Fondazione Giangiacomo Feltrinelli 1983/84, Milan: Feltrinelli.
____ (1986), Why Some People Are More Unemployed Than Others, London: Verso.
—— (1988), ‘Does corporatism really matter? The economic crisis and issues of
political theory’, Journal of Public Policy, 7: 259–84.
____ (1989a), ‘ “Pillarization” and “popular movements”, two variants of welfare
capitalism: The Netherlands and Sweden’, in F. Castles (ed.), The Comparative
History of Public Policy, Cambridge: Polity Press.
____ (1989b), Borgarklass och byråkrati i Sverige, Lund: Arkiv.
TOCQUEVILLE, A. DE (1986), De la democratic enAmerique, 2 vols., 12th edn., Paris:
Folio, Gallimard.
TREU, T., and MARTINELLI, A. (1984), ‘Employers associations in Italy’, in
VOORDEN, W. VAN (1984), ‘Employers associations in the Netherlands’, in
WINDMULLER, J., and GLADSTONE, A. (1984) (eds.), Employers Associations and
Industrial Relations, Oxford: Clarendon Press.
____ et al. (1983), Arbeidsverhoudingen in Nederland, Utrecht: Spectrum.
3
Corporatism and Wage Bargaining
Matti Pohjola
3.1 TO CENTRALIZE OR DECENTRALIZE? ALTERNATIVE WAYS OF

ORGANIZING THE BARGAINING STRUCTURE
Efficient wage-bargaining institutions operate so as to channel the self-

interested behaviour of individual employer and employee organizations away
from purely redistributive activities into well co-ordinated and socially pro-
ductive ones. The success of an economy’s bargaining structure in achieving
this objective can be regarded as a major determinant of its employment, in-
come, and inflation performance and, therefore, of social welfare. Its role in
this task can be considered as equal in importance to well-designed eco-
nomic policies on the part of the government. This chapter looks for such
an efficient bargaining structure in the theoretical and empirical literature
on trade union behaviour.
The traditional view of economists is that decentralized wage fixing and
price setting promote competition and, hence, efficiency. Redistributive
issues do not arise in a competitive economy because all agents are unable
to affect and are not interested in affecting—directly or indirectly through
their own actions—the decisions of others. Workers, for example, do not care
about the decisions made by their employers because at prevailing market-
determined wages each one of them is indifferent between his current job
and the next-best job alternative. Given well-known qualifications regarding
externalities and scale economies, private non-co-operative decision-making
results in an efficient and effectively co-operative solution for wage and price
decisions on the scale of the economy as a whole. Market wages and prices
convey all relevant information when no agent can exert market power and
everyone must react only to the aggregate decisions of others.
According to this view, which we may here call market liberalism, the lack
of competition in the product and labour markets is the reason for poor eco-
nomic performance. The enhancement of competition is seen as the desir-
able solution to unemployment, inflation, growth, and productivity problems.
As regards the labour market and the view that unemployment results from
excessively high wages, this policy prescription means that trade union power
Helpful comments from other participants in the project are gratefully acknowledged, as is
financial support from the Jenny and Antti Wihuri Foundation. The usual disclaimer applies.
CORPORATISM AND WAGE BARGAINING 45
should be reduced in one way or another. By such power is generally meant

unions’ ability to push wages above market-clearing levels.
From this perspective the future of the union movement does not look at
all promising if, as many argue, competition between firms in product mar-
kets is becoming more and more intense because of growing international
trade and increasing economic integration.1 The fact that product markets
are monopolistic and labour markets have monopsonistic elements has tradi-
tionally been regarded as the basic economic justification for trade union
organization. Under such circumstances employers always have incentives
to restrict output and employment in order to reduce wage rates. Monopo-
listic power on the part of employees can restore their bargaining position.
But market liberalism immediately begs the question why it is not possible
to bargain over wages and possibly prices centrally by mimicking decentral-
ized markets whenever it is desirable and by even doing better in the case of
market failures. Centralized wage bargaining has recently been praised for
the economic success of so-called corporatist economies in a number of
empirical studies (see Calmfors and Driffill 1988 for an excellent survey).
The starting-point of these enquiries is the view that competition in the
markets of real economies is not perfect. Rather, markets are dominated by
the behaviour of rent-seeking organizations which waste resources in influ-
encing other agents’ decisions and try to distort allocations so as to generate
outcomes with distributive effects favourable to themselves. As, for example,
Olson (1982) claims, it may be more profitable for an interest group to fight
for redistribution than to fight for the growth of output. Small interest groups
with narrow interests are also more easily organized than large groups which
pay attention to broader interests. Consequently, real economies do not
consist of atomistic agents but of a large number of small and influential in-
terest groups whose uncoordinated non-co-operative behaviour can be re-
sponsible for poor economic performance. If this is the case, then there are
two obvious policy recommendations concerning the efficient design of
bargaining institutions: either to centralize to such an extent that broader
macro-economic interests come to be represented in bargaining or to de-
centralize even more to the extent that interest groups become uninfluential.
We shall here call an economy corporatist if its wage-bargaining structure
is centralized, that is, if workers and employers are organized into a few
powerful national federations negotiating wages and other terms of employ-
ment on behalf of their members, and if, in addition, the government is
prepared to share some political space with these organizations. This can
happen, for example, through participation in wage bargaining. Austria and
1
Economic interdependence has certainly increased in recent years. The roots lie in
improvements in international transportation and communication. Differences in comparative
costs have most likely been reduced by the diffusion of production and management techniques.
In addition, both firms and consumers are likely to have better knowledge about the oppor-
tunities of producing and buying abroad.
46 MATTI POHJOLA
the Nordic countries (Denmark, Finland, Norway, and Sweden) are often
regarded as examples of such societies (see e.g. Calmfors and Driffill 1988).
It is common in economic policy debates to regard centralized wage
bargaining and competitive wage fixing as the only alternative and mutually
exclusive ways of organizing labour market institutions. In surveying the
recent economic performance of selected OECD countries and reviewing
empirical and theoretical works on bargaining, we shall in fact see in sections
3.2 and 3.3 of this chapter that under certain circumstances both centralized
and extremely decentralized (i.e. enterprise-level) wage bargaining generate
equally good economic performance whereas intermediate degrees of cent-
ralization result in inferior outcomes. Consequently, an obvious policy
recommendation is that countries with either centralized or decentralized
bargaining structures should stick to their prevailing institutions while inter-
mediate countries should either centralize or decentralize. There exist, how-
ever, realistic circumstances under which the centralized structure is even
preferable to the decentralized one, in which case the policy adviser should
recommend increasing centralization. This is because all-encompassing
unions are likely to recognize the fact that in modern societies their members
must support the unemployed, not directly but via income tax. This under-
standing serves to moderate excessive wage demands.
But what then explains the observed tendency towards greater decen-
tralization even in the corporatist countries? Corporatism seems to be in
retreat throughout the OECD, with the exception of Australia (see Archer
in this volume). One possible explanation, offered in the concluding section
of this chapter, is the ongoing restructuring of industrial production. Accord-
ing to Piore and Sabel (1984), Gustavsen (1986), Piore (1986), and many
others, an increasing number of firms are abandoning Fordist mass produc-
tion modes and Taylorist hierarchical work organizations in favour of flex-
ible specialization, organizations, and job demarcation. Behind such a change
of business strategy is the view that responding quickly to shifting market
conditions and product demand has become more important in rapidly
changing and uncertain market and technological environments. Flexible
organization structures are regarded as being superior to the traditional highly
integrated, hierarchical ones and also as requiring a new kind of labour
market flexibility, i.e. the freedom to deploy labour within the enterprise.
In a series of papers Aoki (1986, 1988, 1989) has argued that the new
mode of production has to rely more than the traditional one on workers’
participatory information processing and communication capabilities. Such
integrative skills generate informational rents and make workers constituent
members of the firm entitled to sharing the rents through participatory local
bargaining. This can serve as the new economic justification for trade union
organization in the future if the old one is losing its relevance. But it may be
that corporatist labour market institutions have to be adapted accordingly.
It is the aim of this paper to investigate which one—if any—of the proposed
solutions to labour market institutions is desirable under either all or some
specific circumstances. To keep the exposition within reasonable limits, we
have to concentrate on the issue of centralized versus decentralized bargain-
ing in the traditional production mode. It would also be premature to
attempt to provide a comprehensive survey of the new participatory theory
of the firm and its implications for the corporatism debate.
3.2 BARGAINING STRUCTURE AND MACRO-ECONOMIC

PERFORMANCE: EMPIRICAL EVIDENCE
There exist a number of studies on the macro-economic performance of

corporatist economies. Three of the most recent ones are Calmfors and
Driffill (1988), Freeman (1988), and Jackman (1988). Corporatism is here
narrowly defined as indicating a high degree of centralization of wage
bargaining between workers and employers. This is rather restrictive because
no attention is paid to the degree of co-operation or ‘consensus’ between the
bargaining parties.2 These issues have been dealt with empirically, for
example, in Tarantelli (1986) whose work will be reviewed in the next
section. However, bargaining structures do differ between countries in a
distinct way, and these differences are easier to measure than those in the
degree of co-operation. Consequently, we shall take a brief look here at the
existing evidence on the macro-economic performance of selected econom-
ies with different wage-bargaining structures.
Figs. 3.1 and 3.2 illustrate differences in employment performance be-
tween seventeen OECD countries. The first figure displays the average
unemployment rate and employment as a percentage of working-age popula-
tion in the period 1980–5.
Both of these measures are needed in characterizing the unemployment
experience because unemployment can be reduced by means which dis-
courage participation in the labour force but which are then partly revealed
as low employment. The south-east corner of this figure sets out the ‘star’
performers: Japan, Norway, Sweden, and Switzerland. Austria, Finland, and
New Zealand have also performed well whereas Belgium, Italy, and The
Netherlands have done rather badly. Fig. 3.2 presents the changes in these
measures from their average values in the period 1968–73.
By considering changes we can eliminate some structural and cultural
effects, like the size of the labour market and differences in employees’ search
behaviour. The effect of the bargaining structure remains if it displays itself
as a smooth adjustment to the oil and other crises of the 1970s. The star
performers are again found in the south-east corner: Norway and Sweden.
2
See Therborn in this volume for various approaches to the definition of corporatism.
48 MATTI POHJOLA
FIG. 3.1 Average employment and unemployment in 17 OECD countries in

1980–1985
Source: OECD (1987b)
FIG. 3.2 Changes in average employment and unemployment rates between the
periods 1968–1973 and 1980–1985
Source: OECD (1987b)
Austria, Canada, Finland, Japan, and the United States have also done rather
well.
Our attention is drawn to the impact of the bargaining structure by the
fact that Norway, Sweden, and—on looser criteria—Austria, Denmark, and
Finland belong to the group of star performers. The relationship between
centralized bargaining structure and good employment performance be-
comes even more distinct if the non-European countries are excluded from
the comparison.
We can try to confirm the hypothesis that good employment performance
correlates with the degree of centralization of the bargaining structure by
examining the relationship between employment measures and various
indices of centralization. Calmfors and Driffill (1988) have constructed one
rather mechanically as the sum of two indices. The first indicates the levels
of co-ordination within national union confederations and within national
employer organizations (3 = national level, 2 = industry level, 1 = enterprise
level, 0 = occupational level). The second reflects the number of existing
central union organizations and the extent of their co-operation as well as
the number of existing employer organizations and their co-operation (3 =
one dominating union and one dominating private sector employer organiza-
tion, 2 = 2–5 union and/or employer organizations, 1 = no central organiza-
tion on one or both sides of the market). Table 3.1 shows the combined index
on the basis of which the countries under consideration have been classified
into three groups with centralized, intermediate, and decentralized bargain-
ing structures.
The first category consists of countries with centralized wage bargaining
at the national level. In each of these, both employees and employers have
organized themselves into nation-wide unions, the rates of unionization are
high among both white-collar and blue-collar workers as well as employers,
wage negotiations take place centrally at the national level, and the outcomes
of these negotiations also determine the wages of those workers who are not
directly represented. As will be argued later in detail, this kind of bargaining
structure can be conducive to the greater harmonization of private and social
interests in wage fixing, and thus it can contribute to more successful
economic performance. The second group of countries consists of those with
bargaining mainly at the industry level. The last category is formed by
countries with decentralized bargaining at the enterprise level. There is, of
course, a certain element of arbitrariness in such a classification, but there is
no need here to go into a detailed institutional description of the countries.
The idea is simply to demonstrate that there may be a positive correlation,
or even a causal relationship, between the degree of centralization of wage
bargaining, as measured by the extent of both inter-union and inter-
employer co-operation, and good macro-economic performance.
Figs. 3.3 and 3.4 bring out an interesting feature which has received
50 MATTI POHJOLA
TABLE 3.1 Classification of countries

according to their degree of centralization
Bargaining structure
Centralized
Austria 6
Norway 5
Sweden 5
Denmark 5–
Finland 5–
Intermediate
Germany 5––
The Netherlands 4+
Australia 4
Belgium 4
New Zealand 4
Decentralized
France 3+
Italy 3+
UK 3+
Japan 3
Switzerland 3
Canada 2
Note: The measure of the degree of centralization

has been constructed as the sum of two indices.
The first indicates the level of co-ordination of
decision-making within the bargaining structure
(3 = nation, 2 = industry, 1 = enterprise, and 0 =
occupational level). The second measures the
number of existing central union and employer
organizations (3 = one organization on each side,
2 = 2–5 organizations on each side, 1 = no central
organization).
Source: Calmfors and Driffill (1988).
considerable attention in the literature: the relationship between employ-

ment and the degree of centralization of bargaining may have a U-shaped
form.3 Figs. 3.5 and 3.6 indicate that the corresponding inverted U-shape
between unemployment and the degree of centralization appears only when
changes in the unemployment rates are considered (Fig. 3.6).
Both employment and unemployment performances are impressive in the
countries with centralized wage bargaining. Unemployment was low in the
early 1970s and still so in the early 1980s, with the exception of Denmark.
3
The fitted curves are quadratic polynomials and should be interpreted as illustrations of
the arguments raised, not as serious estimations. See Rowthorn (in this volume and 1990) for
such estimations.
FIG. 3.3 Centralization of the bargaining structure and average employment in

1980–1985
Sources: Calmfors and Driffill (1988); OECD (1987b).
FIG. 3.4 Centralization of the bargaining structure and changes in employment

between the periods 1968–1973 and 1980–1985
52 MATTI POHJOLA
FIG. 3.5 Centralization of the bargaining structure and average unemployment

in 1980–1985
FIG. 3.6 Centralization of the bargaining structure and changes in

unemployment between the periods 1968–1973 and 1980–1985
The employment rates, as measured here by the ratio of employment to

working-age population, are high and have been either steady or rising in the
period under consideration, Austria being the exception. This performance
can only be matched by a few countries in the other groups. Switzerland has
done particularly well in terms of unemployment but not so well if employ-
ment is used as the criterion. Japan’s record is outstanding and New
Zealand’s good in terms of both indices. Canada and the United States have
been able to increase their employment rates over this period, but the levels
of both visible and hidden unemployment are still relatively high in both
countries.
Now, if non-European countries are excluded from the analysis, as for
example in Jackman (1988), on the basis that European countries are more
similar to each other, the conclusion is reached that good employment
performance and the centralization of wage bargaining go together—the
celebrated U-shaped relation disappears. According to Figs. 3.3–3.6, there
is no significant difference in the records of the intermediate and decentral-
ized countries. However, if the non-European countries are included, as in
Calmfors and Driffill (1988), then good performance can also be interpreted
to be consistent with the decentralization of wage bargaining. Freeman
(1988) reaches a very similar conclusion from a cross-country regression
analysis by using wage dispersion as the indicator of the labour market
structure: employment performance is good in the countries with either low
or high dispersion but poor in those where dispersion is moderate. Con-
sequently, a more detailed examination is needed to unravel the relationship
between bargaining structure and macro-economic performance although if
low wage dispersion is a desirable policy objective in itself then the central-
ized countries clearly outperform the rest (see Rowthorn (in this volume and
1990) for more detailed arguments).
The next possible step might be an analysis of the inflation record of the
countries under investigation. This is justified on the grounds that un-
employment over a finite period can be seen as an investment in reducing
the inflation rate. Such an examination is not, however, attempted here
because previous analyses have reached the conclusion that there is no clear
difference between the groups in terms of either inflation or the misery index
(see Calmfors and Driffill 1988 and Jackman 1988). The differences in
inflation records can be better explained by the nature of exchange rate
policies than by bargaining structure (Jackman 1988). Instead, we shall look
more closely at the growth performance of these countries to find out
whether a good employment record is obtained in the centralized group at
the expense of investment for the future. Some Scandinavian economists
(Calmfors 1982 and Lindbeck 1979, for example) have been critical of the
centralization of bargaining because they see it as eroding economic efficiency
and competitiveness and resulting in an excessively large public sector.
54 MATTI POHJOLA
These outcomes follow if unions push up real wages and if the government
acts as the employer of last resort.
Fig. 3.7 shows the scatter diagram representing real GDP growth in the
period 1973–84 and the centralization index. Although there are cross-
country differences, the economies with centralized wage bargaining have
not done any worse than the rest—indeed, their output performance is better
than that of the other European countries. There again, the non-European
countries have kept up with the centralized ones although there seems to be
no hump-shaped relationship between output performance and the degree
of centralization of bargaining. The records of the Nordic countries and
Austria look even better if the comparisons are made in terms of GDP per
capita growth rates because of their relatively low population growth rates.
Jackman (1988) has also compared the major European countries in terms
of productivity growth, and capital formation and exports as proportions of
GDP. His conclusion is that countries with centralized wage bargaining do
not differ in any significant way from others. Consequently, the Nordic
economies and Austria have been able to achieve and sustain better employ-
ment records than the rest of Europe without having to sacrifice other aspects
of economic performance.
The kind of casual empiricism of Figs. 3.3–3.6 does not, of course, prove
that centralized or decentralized wage bargaining is the cause of lower
unemployment in any of the countries. Testable hypotheses and more refined
FIG. 3.7 Centralization of the bargaining structure and the growth of real GDP
in 1973–1984
Sources: Calmfors and Driffill (1988); Freeman (1988).
techniques of analysis are needed. There already exist a number of eco-

nometric studies testing the specific hypothesis that corporatism is con-
ducive to real wage restraint and, consequently, to good employment
performance. The basic explanation for the unemployment problems ex-
perienced in the OECD countries is the large and unexpected increases in
the 1970s in the relative prices of raw materials produced outside the region,
requiring reductions in standards of living in the economies importing these
commodities. But real wages, as measured in terms of consumption prices,
have been inflexible downward, thus lowering the demand for labour and
increasing unemployment. International differences in real wage flexibility
can therefore be considered to be a major reason for international differences
in employment performance.
Fig. 3.8 shows the differences in a measure for real wage rigidity which
has been constructed by the OECD for the year 1986 using econometric
wage equations for individual countries.4 The measure is simply formed by
dividing the short-term (single year) price elasticity of the nominal wage rate
by the sum of its short-term elasticities with respect to productivity and
unemployment. These three factors are thus used as indicators of the ability
of the labour market to react to disturbances and emerging disequilibria.
Real wages are rigid if nominal wages react quickly to changes in the price
level. This adjustment is complete in the long run but there are considerable
FIG. 3.8 Centralization of the bargaining structure and real wage rigidity
Sources: Calmfors and Driffill (1988); OECD (1987a).
4
I am grateful to Seppo Leppanen for providing me with this unpublished material. See also
OECD (1987a) for similar calculations.
56 MATTI POHJOLA
differences in the speed of adjustment. According to OECD (1987a), the

proportion of total adjustment in the first year is 67 per cent in Finland, 28
per cent in the USA, and 100 per cent in Germany. A slow reaction of wages
to prices may contribute to labour market stability, for example, in the case
of rising prices as a result of raw material price shocks. Real wages are also
rigid if nominal wages react sluggishly to changes in labour productivity and
unemployment. Rapid reactions would enhance labour market stability. The
proportion of total adjustment in the first year to changes in productivity is
61 per cent in Finland, 27 per cent in the USA, and 65 per cent in Germany.
The corresponding responses to unemployment are 49, 60, and 6 per cent,
respectively. The value of the real wage rigidity is consequently 0.67/(0.61
+ 0.49) = 0.61 for Finland, 0.28/(0.27 + 0.60) = 0.33 for the USA, and
1.0/(0.65 + 0.06) = 1.4 for the Federal Republic of Germany.
The hump-shaped relation appears again in Fig. 3.8, implying that both
decentralized and centralized bargaining are conducive to real wage flex-
ibility and, therefore, possibly to labour market stability. Given this correla-
tion and those of the previous figures, it is rather surprising that no clear
evidence has been found for the significance of the bargaining structure in
econometric studies proper.
Bruno and Sachs (1985) and McCallum (1983, 1986) have estimated
cross-country Phillips-type price equations using a measure of corporatism
as one of the explanatory variables. Bean, Layard, and Nickell (1986) and
Newell and Symons (1987) have estimated wage equations on time-series
data for individual countries. In general, corporatism turns out to be strongly
significant. In these studies, however, corporatism is defined in a more
general way than merely as a measure of the degree of centralization of wage
bargaining. In particular, a measure of consensus between labour and
employers is used in one way or another. But as Calmfors and Driffill (1988)
rightly point out, there are a number of problems associated with such vague
measures of corporatism. In their re-estimation of these models, the results
turned out to be very sensitive to the corporatism index and no clear evidence
was found for either a monotonic or hump-shaped relationship between
good employment performance and the degree of centralization of wage
bargaining.
Freeman’s (1988) study can be used as an example of the problems
involved. Using the cross-country data, he regressed the change in employ-
ment on changes in real GDP and on changes in real wages. The conclusion
was reached that, conditional on the growth of output, there is a clear
negative correlation between employment changes and real wage changes
across the countries. A 1 per cent increase in the real wage rate induces an
approximate 0.5 per cent decline in employment. He further concluded that,
because the cross-country differences in GDP growth are smaller than those
in real wage growth, the latter can be taken to explain the diversity of
employment performance. The problem here is the fact that both employ-
ment and real wages are endogenous, and in no sense can the latter be viewed
as the cause of the former. As Nickell (1988) points out, the key correlation
is the strong positive one between real wage changes and productivity
changes. Thus, for given output, employment and real wages must be
negatively correlated. As for the analysis of the bargaining structure, the
problem now is that this negative correlation can arise in almost any kind of
economy: it appears in a perfectly competitive environment as well as in the
case where wages are fixed and prices marked up on average costs. It can be
argued that unions have a more significant effect on the persistence than on
the level of unemployment. This is likely to be the case if unions restrict job
turnover by making the job-separation rate inefficiently low. If this also
affects the job-finding probabilities through more selective employment
policies on the part of firms, then the gross labour turnover rate goes down
making unemployment more persistent but having an indeterminate effect
on its average rate. But it can also be argued that unionization does not result
in labour turnover restrictions, if bargaining takes place centrally so that the
resulting externalities can be internalized. In a recent paper, Barro (1988)
has tested these hypotheses using aggregate unemployment data from
nineteen countries. His analysis supports the view that unionization has a
positive effect on the persistence of unemployment among economies that
lack a centralized wage-bargaining structure.
There also exist a number of studies in political science testing Olson’s
(1982) theory about the rise and decline of nations (see, for example, Wibe
1987 for a survey). As was mentioned in section 3.1 above, Olson argues
that there is a tendency for the number of small and narrow interest groups
to increase in stable, democratic societies. Such distributional coalitions
distort the allocation of resources and, consequently, reduce economic
growth by pushing for special interests, thus bringing about national decline.
The conclusion is thus reached that the mere existence of trade unions has
a negative impact on economic growth and, thus, on employment. The larger
the number of unions, the greater the adverse effect. Olson, however,
qualifies the conclusion by adding the reverse hypothesis that encompassing
unions promote economic growth. Accordingly, there should exist a U-
shaped relationship between economic growth and the degree of unioniza-
tion or centralization of wage fixing.
Olson’s theory has been tested in a number of ways. The most relevant
one for the topic of this paper is the study by Lane and Ersson (1986). They
tried to explain the differences in GDP growth rates among the OECD
countries (like those shown in Fig. 3.7) using a number of both economic
and political variables. Of these variables, an index of institutional sclerosis
turned out to have the largest explanatory power. As for unionization,
the results indicate that, when the initial level of development and the
58 MATTI POHJOLA
institutional sclerosis are controlled, union membership has a negative

impact on growth, whereas centralization contributes to growth. The effects
are, however, statistically insignificant.
3.3 CONSENSUS AND ECONOMIC PERFORMANCE
The analysis of the preceding section can be criticized on the grounds that
in concentrating on the role of centralization it excludes all other relevant
features of the industrial relations system. Corporatism is a broader concept
than a centralized wage-bargaining structure. As defined in the introductory
section, a corporatist system is one in which the interest organizations either
share a vision of economic policy similar to that of the government or,
through interest intermediation, reach outcomes which are essentially
similar to the ones obtainable under the sharing of objectives. Centralized
bargaining structure may be necessary to enforce such outcomes, but
centralization without consensus is not likely to work. For example, the
results of Newell and Symons (1987), referred to above, indicate that the
economic performance of the same bargaining structure may be quite
different when consensus prevails from the way it is under conflict.
As will be argued in greater detail in the next section, the basic theoretical
argument in favour of the centralized bargaining structure includes elements
of both the prisoner’s dilemma and the public goods problem. A kind of
prisoner’s dilemma arises in wage setting if individual unions bargain over
wages with employers independently from other unions. Although all
employees would benefit from wage restraint in some circumstances, any
individual group might be significantly worse off if it were to accept wage
moderation while other groups were able to obtain large pay increases. The
public goods problem arises because the benefits of nominal wage restraint
do not flow only to the employees bearing its costs but also to the whole
society in the form of a reduction in inflation and an increase in output and
employment. Both problems can be successfully solved by co-operation which
can be enforced via a centralized bargaining structure. But, as Tarantelli
(1986) argues, wage stability can be considered to be a public good only in
so far as the relevant interest groups broadly agree on the distribution of
income. If not, then attempts by any union to change, say, the wage structure
trigger wage increases aimed at re-establishing the relative differentials. In
this case wage stability is no longer a public good in itself because instability
is considered a way of changing income distribution.
Tarantelli (1986) has attempted to capture the effects of these other
aspects of the industrial relations system by constructing an index measuring
centralization, consensus, and arbitration rules. The scale used ranges from
1 to 5 for each of the three dimensions. The centralization measure differs
from Calmfors and DriffilPs (1988) in that Tarantelli also accounts for
factors like the existence of government institutions designed to assist
collective bargaining. His consensus index is a rather subjective one designed
to measure the degree to which there is not only a high degree of ideological
and political consensus (e.g. in the USA) but also a high integration of
interest groups with the economic and political machinery of the government
(e.g. in Austria, Japan, and Germany). The arbitration index accounts for
the actual process of labour dispute settlement in the case both of conflicts
of interest and of conflicts of right. The aggregate measure is reported in
Table 3.2, which also classifies the countries into high, intermediate, and
low corporatism groups.
The most striking difference compared with the classification used by
Calmfors and Driffill (Table 3.1) is that Tarantelli regards Japan as the third
most corporatist country. Its ranking is high along all three dimensions. The
TABLE 3.2 Classification of countries

according to their degree of corporatism
Degree of corporatism
High
Austria 15
Germany 14
Japan 13
Denmark 12
Norway 12
Sweden 12
Intermediate
Australia 10
Finland 10
The Netherlands 10
Belgium 9
Canada 9
USA 9
Low
New Zealand 8
France 7
UK 5
Italy 4
Note: The index has been constructed by
ranking the countries according to three criteria:
centralization of wage bargaining, degree of
ideological and political consensus, and
existence of arbitration rules. The scale ranges
from 1 to 5 for each dimension of corporatism.
Source: Tarantelli (1986).
60 MATTI POHJOLA
relative positions of Canada and the USA are also better here than in Table
3.1 because of the high degrees of ideological and political consensus in these
countries. It comes as no surprise, therefore, that the U-shaped relationship
between employment and corporatism or the hump shape between real wage
rigidity and corporatism can no longer be detected in Figs. 3.9 and 3.10.
Employment performance correlates positively with the degree of
FIG. 3.9 Corporatism and average employment in 1980–1985

Sources: Taranteli (1986); OECD (1987b).
FIG. 3.10 Corporatism and real wage rigidity

Sources: Leppänen (1988); Tarantelli (1986); OECD (1987a).
corporatism. Tarantelli (1986) finds equally clear relationships between his

index and the misery index (i.e. the sum of unemployment and inflation
rates) in the periods 1968–73, 1974–9, and 1980–3. What is interesting here
is that in the post-first and second oil crisis periods the relationships between
these two indices shifted in a way implying a higher misery index for any
given degree of corporatism. But the slope changed so that the comparative
advantage of the most corporatist countries became increasingly greater.
Corporatism seems to matter most when things get worse.
By way of a conclusion, we can say that the empirical evidence presented
in this and the previous section in support of centralized bargaining or, more
generally, of a high degree of corporatism is suggestive but so far too weak
to be taken as conclusive. This is hardly surprising since there is no reason
to expect that just a few variables would explain the differences in perform-
ance of such a number of countries. What is needed, therefore, is theoretical,
preferably empirically testable frameworks to identify the channels through
which the degree of centralization of wage bargaining and, possibly, the
degree of consensus between the bargaining parties affect employment.
3.4 THEORETICAL INSIGHTS
3.4.1 Short-run analyses

There exist a variety of partial or general equilibrium arguments claiming
that centralized wage fixing is associated with better or worse employment
performance than decentralized wage bargaining. It is the purpose of this
section to take stock of them. The analysis is first confined to the case of a
fixed capital stock.
We borrow a simple model from game (or oligopoly) theory to illustrate
the arguments. Suppose that there is an economy consisting of two industries
having identical production structures. Each one is assumed to be dominated
by a monopoly union, i.e. by a union able to impose the nominal wage rate on
the firms. The latter choose employment levels so as to maximize their profits.
The unions have identical preferences over real wages and employment:
Ui(w1, w2) = U(wi/p, Li) (3.1)
where wi denotes the wage rate of the members of union i, i = 1, 2, Li, the
demand for labour in industry i and where p is a consumer price index. Now
it is evident that the choices of the two unions become interdependent
through p and Li—the consumer price index and the labour demands are
generally functions of both w1 and w2. The union utility function U( · ) is
assumed to be increasing and concave in its arguments.
Viewed in isolation, each union i—being in a strong enough bargaining
62 MATTI POHJOLA
position—would set its nominal wage rate wi so as to maximize (3.1) for any
given choice of wj subject to the labour demand equation:
Li = (wi/pi), (3.2)
where pi is the price of the product of industry i. It is here assumed that there
are no intermediate products. The demand side could be modelled to
determine how prices depend on wages, but there is no need here to go into
the details. This relationship can be kept in an implicit formp, pi = pi(w1, w2).
It is well known from the monopoly union theory of the trade union that for
realistic preference characterizations unemployment results in industry i if the
marginal productivity of labour is diminishing, in which case ∂Li/∂wi < 0.5
Unions ‘cause’ unemployment by creating real wage pressures, i.e. by
pushing wages above the market-clearing levels.
This prediction survives a number of extensions concerning assumptions
about wage setting and conditions of production. First, unions are seldom
in a position to set nominal wages. These are normally bargained over with
employers. However, unemployment follows even under bargaining if the
firms can unilaterally set the levels of employment (Nickell and Andrews
1983). Second, a qualitatively similar conclusion is obtained when unions
can bargain over employment as well as wages, as in the efficient bargain
model of McDonald and Solow (1981), if the labour supply constraint is
taken into account in a proper way (Layard and Nickell 1987).6 Third,
competitive labour markets can be introduced by assuming that industry i
is unionized whereas j is not. Then a wage differential in favour of the
unionized workers is likely to arise. Unemployment is created if the other
sector is not able to take up the slack because of, say, unemployment
insurance preventing wages in the non-unionized sector from declining
sufficiently. Finally, the assumption concerning the diminishing marginal
productivity of labour can be replaced by the view that firms’ marginal
revenues are diminishing when firms operate in imperfectly competitive
markets without any need to revise the neo-classical conclusion that unit
labour costs have to decline for employment to increase. It is not, however,
immediately clear what happens to real wages in such a ‘New’ Keynesian
world as that of Meade (1982), Layard and Nickell (1985), Solow (1986),
and others because price formation is endogenous.
Armed with the conclusion that wage fixing plays an important role, at
least in the short run, in the explanation of unemployment, we can turn to
consider the effects of the bargaining structure. Although the significant cor-
relation between high real wages and large unemployment seems undeniable,
5
See e.g. Oswald (1985) and Farber (1986) for more extensive surveys.
6
The over-employment result of McDonald and Solow (1981) is obtained in a partial
equilibrium framework and cannot be realized at the macro-economic level simply because
employment cannot exceed labour supply.
it is incorrect to accept the naive view that real wages are exogenous and
determine unemployment because both are outcomes. Even if prices are
exogenously determined in the world markets, money wage rates are influ-
enced by labour market conditions either through competition or through
union behaviour. Consequently, the bargaining structure affects wage
determination and thus employment. It—rather than real wage rates—can
be regarded as the exogenous factor.
Decentralized bargaining is considered first. This case can be analysed by
the Cournot–Nash equilibrium of the non-co-operative game between the
two unions, which is obtained by assuming that each union maximizes the
preference function (equation 3.1) with respect to its own wage rate, taking
the other union’s wage as fixed. As illustrated by point N in Fig. 3.11, the
resulting equilibrium pair of wages, ( 1, 2), is defined by the intersection
of the reaction functions:
Ri (Wj) = [ wi \ ∂Ui(w1, w2)/∂wi = 0 ] (3.3)
for i, j = 1,2. Here, 1 = 2 because of the assumed symmetry of the model.

It is clear that this kind of non-co-operative behaviour is not optimal for
the unions from the viewpoint of their common interest. This is because
there is an externality involved here. When each union sets its wage rate
independently of the other, it ignores the effect of its choice on the welfare
of the opponent. Such externality can be internalized only through
FIG. 3.11 Co-operative (C) and non-co-operative (N) solutions to union rivalry
64 MATTI POHJOLA
co-operative behaviour—by choosing w1 and w2 jointly to maximize a weighted

sum of the union welfare functions. As the problem is assumed to be sym-
metrical, the symmetrical Pareto solution (w1*, w2*) for which w1* = w2*
can be chosen as the relevant co-operative solution. This outcome can be
enforced if the two unions amalgamate into one centralized organization.
It is obvious that the internalization of the described externality through
the centralization of wage fixing can improve the welfare of both unions. But
does centralization result in lower or higher wages and, thus, in higher or lower
employment? That is, does point C corresponding to (w1*, w2*) in Fig. 3.11
lie to the south-west or north-east of the equilibrium point N? The answer
depends on many unspecified characteristics of the model. Union preferen-
ces, production, labour demand, and competition in the product markets
should be specified in greater detail. It is not possible to do all this here, and
we have to be content with a brief survey of more extensive studies.
First, Calmfors and Driffill (1988) have argued that both a completely
decentralized (i.e. firm-level unionization) and a totally centralized (one
national union) system are likely to be conducive to wage restraint and,
consequently, will result in equally good employment performance. They
use the utilitarian preference function (see, for example, Oswald 1985):
U(wi/p)Li = u(wi/p)Li + v(b)(Ni – Li) (3.4)
Here u( · ) is the utility function of an individual union member who is
employed, whereas v(b) is the utility function of a member who is unem-
ployed and receives an alternative wage (e.g. unemployment benefit) b. Ni is
the total number of members of union i and Li, the number of them who are
employed. All members are assumed to be alike and to supply a fixed amount
of labour when employed. Calmfors and Driffill further assume perfect
competition to prevail in all industries and characterize the output demand
function on the basis of consumer preferences. Employment is determined
by equation (3.2) and is randomly allocated among union members. Output
demand and supply then determine prices as functions of wage rates: pi =
pi(wl, w2).
The outcomes of centralized and decentralized bargaining are identical if
unemployment benefits are financed out of a profits tax. The explanation is
that in both cases the unions effectively choose consumption real wages wi/p
under the constraint that the relative prices pi/pj are constant. In completely
decentralized bargaining this result follows because the output price is given
for an individual firm, whereas in complete centralization all wage and,
consequently, price increases are equal in all sectors of the economy because
of the symmetry of the model.
It is more difficult to obtain any definite conclusions about the effects of
changes in the degree of centralization. Calmfors and Driffill demonstrate
that if dUj/dwi = dUi/dwj < 0 at the non-co-operative equilibrium point
( 1, 2) of the model, then co-operation between the two unions (i.e.

centralization of bargaining) results in lower wages and higher employment
than in the non-co-operative (i.e. decentralized) system. The sign of dUj/dwi
is thus critical. There are two forces here that work in opposite directions,
as can be seen from equation (3.4). First, an increase in wi raises pj by an
amount which depends on how close substitutes the goods i and j are in
consumption. This price effect increases production in sector j and, there-
fore, employment Lj. Second, an increase in wi results in a higher value of
the consumer price index p, and thus it has a negative effect on the utility of
union j. The net effect may go in either direction. The overall outcome is
very much an empirical question. Calmfors and Driffill, therefore, resort to
numerical simulations in their analysis and demonstrate for detailed struc-
tures of consumption, given total nominal demand, and for plausible
production function parameters that both extensive centralization and
decentralization can be conducive to wage restraint and, therefore, to high
employment. If nothing else, their results must be taken as evidence for the
view that the relationship between employment and the degree of centraliza-
tion of bargaining is not necessarily monotonic, but may be hump-shaped.
The second theoretical conclusion concerning the effect of the bargaining
structure is that the symmetry between centralized and decentralized bar-
gaining breaks down when fiscal externalities are considered. These arise if
unemployment benefits are financed by an income tax which is also paid by
the employed workers. Unions now recognize that, even if their members
are not called directly to support the unemployed, they must do so indirectly
via the income tax. This understanding serves to moderate the unions’ wage
demands. It increases employment and the tax base, reduces expenditure on
unemployment benefits, and may even increase the net consumption real
wage in the case where the consumption real wage is falling. Such fiscal
externalities can be internalized through higher degrees of centralization in
wage fixing, meaning that employment is likely to be higher in the centralized
than in the decentralized system. This outcome is confirmed by the simula-
tions of Calmfors and Driffill (1988) as well as by the theoretical analyses of
Kemp, Leonard, and Long (1987) and Mulder (1988).
The view that income taxes are taken as endogenous by centralized and
encompassing unions is the most convincing argument found so far in our
survey for the claimed success of corporatist economies. Besides the de-
scribed theoretical support for it, there exists some empirical evidence in
Bean, Layard, and Nickell (1986). Rowthorn and Glyn’s (1990) praise for
the success of the Swedish economy can also be interpreted by applying this
argument.
The third theoretical result is another plausible and theoretically sound
argument in favour of centralized bargaining. It is obtained if the model is
extended to cover imperfect competition between firms. Perfect competition
66 MATTI POHJOLA
was assumed to prevail above, implying that firms cannot individually

change their output prices when wages change. In the case where firms have
such market power, industry-wide or enterprise-level unions can take
advantage of this fact and set wages at higher levels than are optimal for a
completely centralized union. This kind of an economy has been studied by
Strand (1987) by applying the monopoly union theory. Assuming produc-
tion functions to be linear and firms to be able to set output quantities, he
demonstrated that centralized, encompassing unions pursue the least aggress-
ive wage policies and industry-wide unions the most aggressive policies, while
enterprise-level, local unions were found to use intermediate strategies.
Thus, again both centralized and decentralized bargaining structures do well
in terms of employment performance, the former being likely to outperform
the latter.
Jackman (1988) has also examined the role of imperfect competition in a
framework similar to Calmfors and DriffiU’s. He modelled an open economy
with only one domestic producer in each sector facing monopolistic com-
petition only from abroad and setting prices as mark-ups over wages. In
contrast to Calmfors and DriffiU’s results, Jackman obtained a monotonic
relationship between employment and the degree of centralization of wage
fixing, centralized bargaining being a force for wage moderation. This
monotonicity, however, breaks down once more domestic firms (or plants
of the same firm) are introduced in each sector (Calmfors and Driffill 1988).
The conclusions obtained here can be criticized on the grounds that the
bargaining power of trade unions has been treated as constant and inde-
pendent of the level of bargaining. In applying the monopoly union model,
for example, it was assumed that unions can impose wages on firms at any
given level of aggregation. It may be that in the real world of imperfect
competition decentralization of wage fixing reduces workers’ relative bar-
gaining power whereas centralization increases it. The effects of variable
bargaining power, however, escape the analysis at the current stage of
modelling because there is no adequate theory for it. It is also possible that
union preferences are not independent of the level of aggregation. Central-
ized unions may pay more attention to unemployment problems than local
unions do in which case unemployment is less persistent in centralized
economies. Finally, the role of asymmetric information has not been studied.
It is well known in bargaining theory that, together with individual rationality
constraints, this kind of information may hinder efficiency. Contracts would
not be signed even if they turned out to be efficient expost, i.e. after all private
information has been revealed. Individual rationality means that no one can
be forced into signing an agreement if it is not in his or her best interests to
do so. Under such circumstances centralization of decision-making can
7
See also Davidson (1988) for similar results in a bargaining framework.
improve efficiency even if the higher-level decision makers do not have as

good information as the individual bargaining units (Farrell 1987).
It is tempting on the basis of the analysis of this section to draw the
following policy conclusion. The countries with a centralized bargaining
structure should stick to it, the countries with decentralized bargaining
should preserve their systems, especially if product markets are competitive,
while the intermediate countries should abandon theirs in favour of central-
ized bargaining. Such policy prescriptions are, however, premature because
they are based on short-run considerations obtained by assuming the capital
stock to be fixed. Unions may have little impact if capital is mobile. In fact,
as Jackman (1988) emphasizes, there is a unique equilibrium real wage for
a small open economy where output prices are given by world markets, where
capital is supplied elastically at a constant world rate of interest, and where
production takes place under constant returns to scale. Consequently, there
is nothing unions can do in the long run if they are not able to restrict capital
mobility. In any case, the choice between bargaining structures should not
be based solely on short-run considerations, but long-run factors have been
much neglected in the literature on this topic.
There is, however, an additional issue on which a short-run analysis of the
bargaining structure can shed some light—that is, the relationship between
economic policy and wage setting. This has received considerable attention
from some Scandinavian economists (for example, Hansen 1958; Johansen
1977; Soderstrom and Viotti 1979; Calmfors 1982; Calmfors and Horn
1985, 1986; Hoel 1987). The argument is simple. Suppose that government
policy reacts to aggregate wages and/or employment. In a decentralized
bargaining system, each union will regard the impact on aggregate variables
of its wage claims as negligible. Consequently, unions view macro-economic
policies as independent of their wage claims. In a centralized bargaining
structure, however, the union movement is assumed to perceive this feed-
back and to act accordingly. The lesson learnt is that the effects of macro-
economic policy vary with respect to the degree of centralization of
bargaining. Consequently, the differences in employment behaviour shown
in the figures of sections 3.2 and 3.3 may be better explained in terms of
economic policies. It can be recalled here that most European economies
pursued high budget deficit policies in the late 1970s, but reduced deficits
in the early 1980s. If such policies were conditioned on bargaining outcomes
in the way that deficits are reduced in the case of ‘excessive’ wage increases,
8
It is interesting to observe that, for example, in Finland, which is one of the most
strike-prone European countries, most strikes are associated with local (i.e. enterprise- or
plant-level) bargaining. The asymmetry of information is a possible explanation for this. Its role
is important at the local level and strikes may be necessary instruments for information
revelation. At the national or industry level, the union negotiators have to use public information
which needs no revelation mechanisms.
68 MATTI POHJOLA
then the relative success of the Nordic and Austrian economies can be
attributed to their centralized bargaining structures.
There is also another reason for the need to study the relationship between
wage setting and economic policy. As has been demonstrated in the foregoing,
wage stability is a kind of public good: the benefits in the form of higher
employment and lower inflation do not accrue exclusively to those who
provide it. But it is also a very special kind of public good in the sense that,
while everyone may benefit from it, its provision depends on the behaviour
of a particular group of society: the employees and their organizations. The
workers, however, are not only interested in wages and employment but also
in ‘macrogoods’ like unemployment benefits, pension funds, retraining
schemes, health insurance, labour legislation, taxes, etc. The supply of such
goods is under the control of the government. This fact serves as the basis
of government involvement in wage fixing as a bargaining party. There does
not exist any theoretical work on this topic, but common sense suggests that
bargaining between the government and the labour movement is much easier
if the bargaining structure is centralized than if it is decentralized. In addition,
agreeable outcomes are easier to achieve when the parties share common
objectives than in the case where they represent opposing political views.
3.4.2 Long-run considerations: instability of consensus

As was demonstrated in the preceding section, co-operation between trade
unions results in greater welfare for their members than is obtainable under
rivalry. Such co-operation, however, requires commitments, i.e. binding
agreements, preventing the parties from cheating each other. Each union has
an incentive to raise its wage if the opponent is known to adhere to the
co-operative solution. But such considerations then lead the players to
choose the Cournot-Nash strategies instead of the co-operative ones. The
commitments required can be obtained by centralizing the bargaining struc-
ture, i.e. by joining the two unions into one encompassing organization. This
is the method used by Calmfors and Driffill (1988), Jackman (1988), and
Strand (1987) who, consequently, do not discuss the stability issue.
It is, however, well known in oligopoly theory that binding contracts are
not necessary to enforce co-operative solutions if the game relationship is a
repeated one. The unions do, in fact, confront each other as well as the
employers from year to year. Even if there exist a number of independent
unions, they may then achieve tacit or implicit agreement on co-operation
without binding, formal arrangements. All that is needed is a mechanism,
included in the efficient equilibrium strategies, which deters an individual
union from cheating. Such a mechanism can be constructed by adopting
punishment strategies designed to be used against the defecting union in
periods following cheating. If punishments inflict large enough welfare losses
on the defecting players, they prevent cheating and, thus, make co-operation
sustainable.
By applying this idea to the union rivalry model, it is possible to show that
the co-operative solution (w1*, w2*) (point C in Fig. 3.11) can be sustained
as an equilibrium. The following trigger strategy pair (Friedman 1971), for
example, yields this result if the discount rates of the unions are not too high:
wi(t + 1) = wi* if wi(t) = wi* for i = 1, 2,
= i otherwise for a fixed punishment interval. (3.5)
Here t denotes the time period. The discount rates play a role because the
benefits of cheating are obtained in the current period, whereas the welfare
costs of punishments fall on the players in the future. The unions are here
assumed to be maximizing the discounted sum of the single-period utilities:
δ being the discount rate.

An interesting conclusion can be drawn from this extension of the basic
one-shot game: bargaining structure may not matter at all, but the ‘mode of
play’ is important. If the unions possess complete and perfect information
and if they are not too myopic, then they should be able to reach consensus
on wage policies and obtain efficient outcomes without any binding con-
tracts. Consequently, if co-operation between unions is conducive to wage
restraint and good employment performance, then even the countries with
a moderate number of industry-wide unions—the intermediate group of
Table 3.1—should be able to display good macro-economic performance at
least in stable environments such as those which prevailed in the growing
world economy of the 1960s.
It is, however, often claimed (see, for example, Newell and Symons 1987)
that there exist variations in the observed modes of play between the unions
and, possibly, other organizations. Countries may experience alternating
periods of consensus and conflict. Endogenous changes in the degree of
co-operation between the unions can be explained by the aid of the model
described above if informational imperfections are included in it. Suppose
that union i cannot directly observe the wage rate set by union j, but has to
rely on indirect observation obtained through the price level p. This may be
the case if the contracts between employers and employees are so complic-
ated, as they are in real life, that it is very hard for an outsider to measure
their impact on labour costs. Suppose further that the price level is also
influenced by some other factors which can be regarded as random from the
70 MATTI POHJOLA
unions’ viewpoint. Thus, when a union observes an increase in the price level
or inflation rate, it cannot be certain whether this rise is caused by random
factors or by an increase in the wage set by the opponent. This uncertainty
may cause the union to switch from an initial co-operative mode of play to
the non-co-operative one for a fixed period during which the punishments
of equation (3.5) are executed. As demonstrated by Green and Porter (1984)
in an oligopoly model, this happens if the observed price or inflation shock
is large enough. Given that positive inflation shocks are more likely during
economic booms than recessions, we obtain the empirically testable con-
clusion that corporatism, i.e. a co-operative wage restraint, is much harder
to achieve and sustain when economic activity is high than when it is low.
And, inversely, corporatism is likely to emerge when activity is slack.
This conclusion is very similar to Tarantelli’s (1986) view that under
stable and low inflation a large wage request made by an individual union is
easily noticed by others and, therefore, deterred by a mechanism like the one
presented in equation (3.5). But if the initial inflation rate is high and
considered to be uncertain in the future, then any high wage demand by a
particular group of employees is less likely to be observed. There is, thus, a
greater incentive for free riding because the expected costs of cheating are
considered to be lower. Consequently, price shocks, like those created by the
oil crises of the 1970s, increase the likelihood of non-co-operation between
the interest organizations.
The conclusion of this section is that bargaining structure may matter only
in turbulent environments. Under ‘normal’ circumstances of stable eco-
nomic growth and inflation, countries with a number of strong unions should
also be able to achieve good macro-economic performance. The oil crises
may then explain the divergent performance in the 1970s and 1980s of
countries with centralized and intermediate bargaining structures. The
economic history of the intermediate countries should also exhibit endogen-
ously alternating corporatist and non-corporatist periods during these two
decades. Hence, we are likely to learn more about corporatism by investigat-
ing the experiences of individual countries than by making direct cross-
country comparisons.
3.4.3 Economic growth and the bargaining structure

Let us now turn our attention to long-run analysis. It can be argued that the
choice of the bargaining structure should be based on considerations which
are of longer-term nature than those examined in the preceding sections.
These issues have been much neglected in the literature on the economic
consequences of trade unions. Most studies have been carried out in a
9
See Rowthorn (1990) for further developments of this argument.
CORPORATION AND WAGE BARGAINING 71
framework that is very different from the union-firm set-up studied earlier—
that is, in the context of two-class growth theory. There do exist a few
attempts to generalize the monopoly union or the union-firm bargaining
model to cover investments and the demand for labour over time (see, for
example, Ploeg 1987), but these partial equilibrium studies do not much
improve upon our understanding obtained from static analyses of the
relationship between bargaining structure and economic performance.
Consequently, only studies in the growth theory tradition will be sur-
veyed here.
A striking new conclusion catches the eye immediately: it may be the case
that unions do not have any power in the long run—the best they can do is
to set the real wage rate at the competitive level equalizing the demand for
and the supply of labour. This result is obtained by Kemp and Long (1987)
for a two-class, constant returns-to-scale neo-classical growth model in
which the workers (i.e. the all-encompassing union) set the minimum rate
of real wages so as to maximize a discounted sum of the wage bill over time,
firms set employment levels so as to maximize profits, and only the capitalists
save and automatically invest part of their income (i.e. profits). The con-
clusion is rather intuitive if the elasticity of substitution between labour and
capital is equal to or greater than one. In this case, a rise in the real wage rate
above the free market level must both reduce output (by creating unemploy-
ment) and decrease (or leave unchanged) labour’s share of it. By forcing
down the return on capital and, hence, saving, such a real wage pressure also
reduces wage shares in the future. Therefore, a rational union will not set
the real wage rate above the market-clearing level.
A more elaborate model is needed to prove the result for the case in which
the elasticity of factor substitution is less than one. But we can understand
the Kemp-Long result without going into the details if it is recalled from
growth theory that the capital-labour ratio k maximizing worker consump-
tion in balanced, full-employment growth is obtained as the solution to f’ (k)
= n/s, where /( · ) denotes the production function in its intensive form, n is
the natural growth rate, and 5 the capitalists’ savings rate. It is this steady
state level of k, if any, which is optimal for the union to aim at in the long run.
A very similar result was obtained by Hamada (1967) for a model in which
the labour market is assumed to be competitive, but where the workers can
set lump-sum transfer payments from capitalists to workers so as to max-
imize the sum of discounted worker consumption over an infinite horizon.
He showed that the workers cannot be made better off by a positive or a
negative income transfer in the long run if their discount rate is zero. For
positive transfers to be optimal, it is essential that they indicate some time
preference or impatience.
According to these studies, it is optimal for the union movement not to
exercise its market power in the long run. Similar conclusions were already
72 MATTI POHJOLA
hinted at in the preceding sections when allocational distortions were

discussed. If capital is mobile, unions may not be able to extract any rents.
This outcome of the dynamic models corresponds to the U-shape hypothesis
of the static analysis: economies with strong union movements are likely to
perform as well as those with competitive labour markets.
A number of qualifications are, however, needed before unions are dis-
missed in favour of free competition. First, it was assumed that the union
movement can set the real wage rate. This is questionable even in the case
where there is a strong and unified union movement, and the outcome may
turn out to be worse in the case of no unions if the product market is not
perfectly competitive. Trade unionism may be necessary for the equality of
the wage rate with the marginal productivity of labour to be achieved in the
long run. Second, the conclusions refer only to the steady state. Unions have
power in the short run. In the Kemp and Long model, the union sets the
minimum real wage at a level exceeding the competitive real wage, thus
causing unemployment. In the Hamada model, it is optimal for the workers
to suppress their consumption at an initial state of development, by trans-
ferring wages to capitalists, in order to attain the desired capital-labour ratio
in the most rapid way. Thus, the achievement of steady state outcomes is
likely to depend, among other things, on the number of unions and the
degree of co-operation between them. In this sense the bargaining structure
or corporatism matters in the long run as well. The final qualification
concerns production technology. Unions may have power even in the very
long run if returns to scale in production are variable, rather than constant.
Palokangas (1989) has made some initial experiments with this modification
to the Kemp and Long model. Production and organizational technologies
are also likely to be a function of the bargaining structure and not exogenous
as was assumed in the studies mentioned. Thus bargaining and participatory
systems have an impact on technical progress.
Let us illustrate the importance for economic growth of the bargaining
structure or co-operation between unions by considering the transition to
steady state growth. Let us simplify the Kemp and Long model by assuming
that only one production technique exists. Thus, the elasticity of factor
substitution is zero, and the firms can be left out of the model—the demand
for labour is proportional to the capital stock. Suppose that initially labour
supply is not binding, i.e. the economy is of the labour-surplus type. Then,
if the working class can set real wages and if the capitalists’ savings rate is
fixed, the workers can control accumulation and employment. They do this
so as to maximize a discounted sum of worker consumption over time. Now
it is possible to show by applying dynamic game theory that growth is sub-
optimal from the workers’ viewpoint if they do not act as a single decision-
making unit but rather organize themselves into a number of non-co-operating
trade unions (Pohjola 1984). Slow growth, large unemployment, and high
real wages are the consequences of such decentralization of bargaining. It is

interesting to note that the correlation between labour costs (i.e. high wages)
and slow growth (and large unemployment) can be explained in terms of the
institutional structure or the mode of play between the unions. High real
wages do not here ‘cause’ unemployment.
This outcome can be explained in terms of a dynamic externality. One
union’s decision to save for the future by adopting moderate wage increases
is affected by the fact that the accumulated amount may be expropriated by
other unions by their increasing the wages of their members. This analysis
can be interpreted as a dynamic extension of Sen’s (1961, 1967) isolation
paradox. He has demonstrated how private saving versus consumption
decisions are fundamentally different from collective ones and how in-
dividual decisions may result in an insufficient amount of investments for
the future. Assuming that, given the decisions of others, each individual
would be better off by not saving an additional unit whereas all would be
better off by saving collectively, it follows that no one will save although
everyone would have preferred one more unit of saving by each individual.
Because of the external effects between individuals’ decisions and the lack
of influence on others’ actions, nobody wants to invest himself, although
each individual would like to see others invest, i.e. all try to shift the burden
of saving on to others.
Pohjola (1984) has further demonstrated that the inefficiency described
above becomes more severe as the number of independent unions grows.
There is no U-shaped relationship here between the degree of centralization
and economic performance. It is possible that decentralization or the lack of
co-operation between unions can prevent the economy from ever reaching
the full-employment steady state.
The dynamic externality described above can be internalized by centraliz-
ing the bargaining structure. But can the same result be achieved if the unions
retain their independence but co-operate voluntarily? It was suggested in the
preceding section that, in the absence of informational imperfections, co-
operative outcomes can be supported as equilibria in repeated games by the
aid of trigger strategies that threaten to revert to non-co-operative play when
one of the players deviates from co-operative behaviour. Things are, how-
ever, different if the game relationship is structurally dynamic. A repeated
game consists of the repeated play of a particular one-shot, i.e. static, game.
The repetitions are structurally independent, meaning that in each period
players’ pay-offs depend only on their actions in the current period. Struc-
turally dynamic games are different in the sense that these pay-offs also
depend on the whole history of the game. The basic difference lies in the
fact that the sustainability of a co-operative policy at any given time now
depends on the values of state variables summarizing the history. Since the
states evolve over time as functions of the players’ controls and the states
74 MATTI POHJOLA
themselves, it may happen that co-operation between players breaks down

endogenously in the course of the game when a previously sustainable
policy becomes unsustainable or that co-operation emerges from a non-
co-operative mode of play when state variables achieve some critical levels.
Consequently, it is possible to observe endogenous variations in the degree
of co-operation between players even in the absence of informational imper-
fections.
These kinds of phenomena are not yet well understood even in game-
theoretic literature, but there already exist two preliminary applications to
the modelling of growth and distribution (Haurie and Pohjola 1987, Kaitala
and Pohjola 1990). As these ideas have some relevance to corporatism, we
briefly summarize the Kaitala-Pohjola paper. It is based on the Hamada
(1967) model described earlier. The working class is one of the players. It is
assumed to be able to set income transfers from capitalists so as to maximize
a discounted sum of workers’ consumption over time. Only the capitalists
save, and the original model is extended by assuming that capitalists also
optimize dynamically. They choose their saving strategy so as to maximize a
discounted sum of capitalist consumption over time. Non-co-operation
between these players has the most drastic consequences: the economy does
not grow. Instead, the capital stock decays at a constant rate. A similar
outcome can be obtained for the corresponding game formulation of the
Kemp and Long (1987) model (see Kemp and Shimomura 1987). Con-
sequently, both social classes have strong incentives to devise mechanisms
which would sustain some form of co-operation.
Kaitala and Pohjola (1990) assume that workers and capitalists are in-
duced to use trigger strategies which specify that co-operative policies are
followed as long as both sides comply but that the non-co-operative equi-
librium strategies will be used for the rest of the game once cheating is
detected. They demonstrate that efficient equilibria exist in such a game if
the natural growth rate is positive. Thus, growth is essential for redistribu-
tion. They further show that an expanding economy is likely to be less
conflict-ridden than a stagnating one, and that sustainable co-operation is
likely to imply income transfers from capitalists (i.e. the rich) to workers (the
poor) in the long run.
These predictions are not very strong, but they may shed some light on
the nature of economic policies in countries, like the Nordic ones, where
redistribution through taxation and collective bargaining has been at the
centre of policy-making. In particular, the connection between economic
development and the nature of sustainable policies is an interesting one for
the following reason. As discussed in the preceding sections of this chapter,
the received view in the literature on social corporatism is that centralized
bargaining in a co-operative (consensus) framework is the best institutional
arrangement in terms of economic performance. Causation here runs from
co-operation and centralization to performance. The conditions under

which such an institutional arrangement would emerge have not been
explored. Kaitala and Pohjola show that for co-operation to emerge and to
be sustainable it has to imply income redistribution in favour of the working
class in the long run, although both workers and capitalists benefit as
compared to the non-co-operative mode of play. Admittedly, however, the
capitalists’ relatively weak position is explained by the specific feature of their
model that investment is irreversible.
Schott (1984) has also discussed the corporatism issue in the Lancaster
(1973) model in which workers control die wage share and capitalists the
share of investments in output. She defined working-class power by means
of a model parameter (the upper limit on the wage share) and showed that
countries may differ from each other in terms of economic performance
because of differences in labour power. The Kaitala and Pohjola analysis is
different in the sense that they have not regarded class power as an exogenous
parameter but have instead examined how it arises endogenously as a
function of economic development.
The possibility of class conflict over income distribution does not neces-
sarily imply that capitalist societies are dynamically inefficient in the sense
of Lancaster (1973). The fact that either of the social classes can, at any time,
impose the conflict outcome on the economy only affects the set of sus-
tainable efficient equilibria by making it smaller. All co-operative policies are
not agreeable because they are not sustainable as equilibria of the game. A
social contract, precommitting both sides to a particular solution, is not,
therefore, a necessary institution for obtaining efficiency. Here the con-
clusions differ from, for example, Mehrling’s (1986) views on the role of
government in enforcing a social contract. No enforcement mechanism is
required if the contract belongs to the set of efficient equilibria.
Although this analysis was performed in the context of a worker-capitalist
game, similar conclusions should be obtainable for the case of union rivalry.
The implication for empirical studies is again that we are likely to learn more
about corporatism by investigating the experiences of individual countries
rather dian by making direct cross-country comparisons.
3.5 CONCLUSIONS
In this chapter we have surveyed empirical and theoretical literature on

corporatism viewed either as a centralized wage-bargaining structure or as
an industrial relations system in which consensus prevails. The study leads
to the following conclusions. First, empirical evidence for the traditional
view that centralized wage fixing contributes to more successful economic
performance is suggestive but so far too weak to be taken seriously—the
76 MATTI POHJOLA
econometric results obtained do not survive standard stability tests. This fact
may, however, reflect more the weakness of the empirical models than the
non-existence of the hypothetical relationship. Most studies have, for
example, concentrated on the relationship between the level of unemploy-
ment and a measure of centralization, whereas it can be convincingly argued
that the persistence of unemployment could be a more relevant measure of
unemployment performance.
Second, even theoretical studies do not support the view that there exists
a monotonic relationship between the degree of centralization of wage
bargaining and some measure of economic performance—extensive decen-
tralization can also be a route to full employment. However, most authors
seem to agree that intermediate degrees of centralization are harmful. But
this view was challenged in section 3.4 above where it was argued that it is
not die bargaining structure as such that matters but the degree of co-
operation or ‘consensus’ between the parties involved. In normal circum-
stances, they should be able to reach an implicit agreement on the nature of
wage-fixing strategies without any binding agreement or centralization of the
bargaining structure. It seems clear that the degree of centralization is at best
a poor predictor of economic performance. For this reason it was suggested
that more could be learnt about corporatism by examining the experiences
of individual countries—alternating corporatist and noncorporatist episodes—
instead of making direct cross-country comparisons.
Third, given the reservations made, if the policy makers of an intermediate
country were to make a decision about changing the degree of centralization,
the analysis of sections 3.2–3.4 would advise them to go for greater
centralization. There are two empirically relevant factors which make cent-
ralization preferable to decentralization. The first one is the existence of fiscal
externalities and the second the role of imperfect competition in output
markets. Fiscal externalities arise if unemployment benefits are financed by
an income tax which is also paid for by the employed workers. Taxes are thus
endogenous, and their level depends on the policies pursued by the unions.
Such externalities can be internalized through higher degrees of centraliza-
tion of bargaining, meaning that wage fixing is likely to be more conducive
to high employment in a centralized than in a decentralized system. Imper-
fect competition on the other hand provides employers with incentives to
hold wages back by restricting employment. This fact can be regarded as the
economic justification for the existence of unions. This policy conclusion
must, however, be qualified in the sense that it is based on the assumption
of a fixed capital stock. In addition, the analysis of section 3.4 says that the
beneficial effects of greater centralization should be obtainable without any
change in the bargaining structure if the mode of play between the unions
can be made more co-operative.
Fourth, it was suggested that government policy effects are dependent on
the bargaining structure or the degree of co-operation between unions. More

specifically, it was argued that the effects of contractive policies are not so
harmful to employment in the centralized system as they are in the decentral-
ized system. These policy connections have not been explored much in the
literature, and they would merit more attention.
The final conclusion concerns the long run. It was argued that the bargain-
ing structure may not matter in the very long run—the best the workers can
do is to set the real wage rate at the competitive, full-employment level.
However, the very long run is far away, and the transition to a particular steady
state growth path is certainly influenced by the structure of bargaining. It was
pointed out that it may be impossible to achieve full employment growth
without a union movement. In the long run, corporatism is also endogenous,
but the circumstances under which it is likely to arise or decay have not been
much explored.
The analysis of this chapter was based on a set of rather restrictive
assumptions and needs to be qualified in a number of respects. First, it was
assumed that the relative bargaining power of the interested parties is inde-
pendent of the degree of centralization. It maybe the case that workers are in
a weaker position in a decentralized system than in a centralized one. Second,
information has been assumed to be public knowledge. It is well known in
bargaining theory that the existence of private information and the ‘freedom
to choose’ (i.e. to walk out of negotiations) can hinder efficiency. Under such
circumstances centralized decision-making may result in socially desirable
outcomes even if bureaucrats cannot induce the bargaining parties to reveal
their information but have to rely on imperfect public knowledge (Farrell
1987). Third, co-ordination also pays in the case where information is
imperfect in the sense that some relevant facts are not yet known when
decisions have to be taken. Cremer (1988) has demonstrated how the whole
organization composed of independent decision makers benefits from com-
mon information (i.e. measurements) on such uncertain events if they care
sufficiently about the variability of their decisions (or rather outcomes). Thus,
even a decentralized union movement would benefit from a centralized
co-ordination of information if labour market solidarity is an objective shared
by the independent unions. In typical circumstances, individual unions know
less than would be optimal were they perfecdy rational, and union organiza-
tion finds it beneficial to co-ordinate the limited knowledge of its members.
By ‘corporatist culture’ we mean the generation of such common knowledge
and the sharing of solidaristic objectives. In such culture it is not necessary
for an individual union to know other unions’ decisions in order to be able to
predict the outcomes because it knows the way decisions are taken and
because information is common.
All these qualifications are arguments in favour of some form of cent-
ralization. But is centralization compatible with the ongoing industrial
78 MATTIPOHJOLA
restructuring towards flexible specialization and work organizations?10 What

has been said in this chapter about bargaining structure and macro-
economic performance was based on analysing their relationship via the
average or aggregate real wage. Nothing has been said about the wage
structure, i.e. about relative wages, or about other conditions of work.11
Casual empiricism suggests that there exist pressures towards greater inter-
and intra-industry wage dispersions as well as greater flexibility in the
deployment of labour which are associated with decentralized wage setting.
Bargaining at the level of the firm does indeed allow greater scope for firm-
or industry-specific factors than can be taken into account in a more
centralized structure. These pressures are likely to grow in the future if
competition between firms is becoming more intense, as some have argued.
But does the possible erosion of monopolistic competition mean that
unions lose the economic justification for their existence? In a series of
interesting studies, Aoki (1986, 1988, 1989) has studied the foundations of
the bargaining theory of the firm. He claims that more intense competition
between firms leads them to adopt new forms of organizations which rely
more on workers’ participatory information-processing and communication
capabilities than the traditional hierarchical organization structure. Emerg-
ing, new participatory modes of production generate information rents
which can be bargained over by the parties involved—the owners who supply
capital, the management who provide the organization framework, and the
employees who offer their information-processing capacities. Although Aoki
does not regard the generation of these rents as a justification for union
representation, it makes the workers entitled to their share of such rents. We
may, therefore, give a negative answer to the above question, and conclude
that there is a role for the union movement in the participatory mode of
production as well. Information rents are firm-specific but collective and
cannot, therefore, be written into individual wage contracts. But wage
bargaining has to take place at the level of the firm for conditions for an
efficient generation of such rents to arise. If Aoki’s view is correct, it is the
most relevant argument in favour of decentralized bargaining. The question
then arises which form of corporatism, if any, can survive in this new
environment. Is the corporatist system flexible enough to be able to face
these new challenges or will it be buried along with the old production
methods and forms of organization?
10
See e.g. Piore and Sabel (1984) and Piore (1986) for descriptions of the changing models
of production.
11
Rowthorn (in this volume and 1990) has taken up the wage dispersion issue.
REFERENCES
AOKI, M. (1986), ‘Horizontal vs. vertical information structure of the firm’, American
Economic Review, 76: 971–83.
____ (1988), Information, Incentives and Bargaining in the Japanese Economy, London:
____ (1989), ‘The participatory generation of information rents and the theory of the
firm’, in M. Aoki, B. Gustafsson, and O. E. Williamson (eds.), The Firm as a Nexus
of Treaties, London: Sage.
BARRO, R. J. (1988), ‘The persistence of unemployment’, American Economic Review
(Papers and Proceeding), 78: 32–7.
BEAN, C. R., LAYARD, P. R. G., and NICKELL, S. J. (1986), ‘The rise in unemploy-
ment: a multi-country study’, Economica, 53: S1–S22.
Basil Blackwell.
CALMFORS, L. (1982), ‘Employment policies, wage formation and trade union
behaviour in a small open economy’, Scandinavian Journal of Economics, 84:
345–73.
____ and DRIFFILL, J. (1988), ‘Bargaining structure, corporatism and macro-
economic performance’, Economic Policy, 6: 14–61.
____ and HORN, H. (1985), ‘Classical unemployment, accommodation policies and
the adjustment of real wages’, Scandinavian Journal of Economics, 87: 234–61.
____ (1986), ‘Employment policies and centralized wage-setting’, Economica,
53:281–302.
CRÉMER, J. (1988), ‘Common knowledge and the coordination of economic
activities’ (mimeo).
DAVIDSON, C. (1988), ‘Multiunit bargaining in oligopolistic industries’, Journal of
Labour Economics, 6: 397–422.
FARBER, H. S. (1986), ‘The analysis of union behavior’, in O. Ashenfelter and R.
Layard (eds.), Handbook of Labour Economics, Amsterdam: North-Holland.
FARRELL, J. (1987), ‘Information and the Coase theorem’, Journal of Economic
Perspectives, 1: 113–29.
FREEMAN, R. B. (1988), ‘Labour market institutions and economic performance’,
Economic Policy, 6: 64–80.
FRIEDMAN, J. W. (1971), ‘A non-cooperative equilibrium for super-games’, Review
of Economic Studies, 38: 1–12.
GREEN, E. J., and PORTER, R. (1984), ‘Non-cooperative collusion under imperfect
price information’, Econometrica, 52: 87–100.
GUSTAVSEN, B. (1986), ‘Evolving patterns of enterprise organisation: the move
towards greater flexibility’, International Labour Review, 125: 367–82.
HAMADA, K. (1967), ‘On the optimal transfer and income distribution in a growing
economy’, Review of Economic Studies, 34: 295–9.
HANSEN, B. (1958), The Economic Theory of Fiscal Policy, London: Allen & Unwin.
HAURIE, A., and POHJOLA, M. (1987), ‘Efficient equilibria in a differential game of
capitalism’, Journal of Economic Dynamics and Control, 11: 65–78.
80 MATTI POHJOLA
HOEL, M. (1987), ‘Endogen makrookonomisk politikk med fagforeninger’,

Sosialekonomen, 7: 26–33.
JACKMAN, R. (1988), ‘Wage formation in the Nordic countries viewed from an
international perspective’ (mimeo).
JOHANSEN, L. (1977), Lectures on Macroeconomic Planning. Part I: General Aspects,
Amsterdam: North-Holland.
KAITALA, V., and POHJOLA, M. (1990), ‘Economic development and agreeable
redistribution in capitalism: efficient game equilibria in a two-class neoclassical
growth model’, International Economic Review, 31: 421–38.
KEMP, M. C, LEONARD, D., and LONG, N. V. (1987), ‘Trade unions, seniority and
unemployment’, European Economic Review, 31:1093–112.
____ and LONG, N. V. (1987), ‘Union power in the long run’, Scandinavian Journal
of Economics, 89: 103–13.
____ and SHIMOMURA, K. (1987), ‘A differential game between organized labour and
organized capital’ (mimeo).
LANCASTER, K. (1973), ‘The dynamic inefficiency of capitalism’, Journal of Political
Economy, 81:1092–109.
LANE, J.-E., and ERSSON, S. (1986), ‘Political institutions, public policy and
economic growth’, Scandinavian Political Studies, 9: 19–34.
LAYARD, R., and NICKELL, S. (1985), ‘The causes of British unemployment’,
National Institute Economic Review, 111: 62–85.
____ (1987), ‘Is unemployment lower if unions bargain over employment?’
(mimeo).
LEPPANEN, S. (1988), ‘Wage bargaining and economic performance’ (in Finnish),
Finnish Economic Journal, 84: 271–8.
LINDBECK, A. (1979), ‘Imported and structural inflation and aggregate demand: the
Scandinavian model reconstructed’, in A. Lindbeck (ed.), Inflation and Employment
in Open Economies, Amsterdam: North-Holland.
LOVEMAN, G. W., and TILLY, C. (1988), ‘Good jobs or bad jobs? Evaluating the
American job creation experience’, International Labour Review, 127: 593–611.
MCCALLUM, J. (1983), ‘Inflation and social consensus in the seventies’, Economic
Journal, 93: 784–805.
____ (1986), ‘Unemployment in the OECD countries in the 1980s’, Economic
Journal, 96: 942–60.
MCDONALD, I. M., and SOLOW, R. E. (1981), ‘Wage bargaining and employment’,
American Economic Review, 71: 896–908.
MEADE, J. E. (1982), Wage-Fixing, London: Allen & Unwin.
MEHRLING, P. G. (1986), ‘A classical model of the class struggle: a game-theoretic
approach’, Journal of Political Economy, 94: 1280–303.
MULDER, C. B. (1988), ‘Wage moderating effects of corporatism: decentralized
versus centralized wage setting in a union, firm, government context’, Tilburg
University, Department of Economics, Research Memorandum 316.
NEWELL, A., and SYMONS, J. S. V. (1987), ‘Corporatism, laissez-faire and the rise in
unemployment’, European Economic Review, 31: 567–614.
NICKELL, S. (1988), ‘Discussion’, Economic Policy, 6: 78–9.
____ and ANDREWS, M. (1983), ‘Trade unions, real wages and employment in
Britain 1959–79’, Oxford Economic Papers (supplement), 35: 183–206.
OECD (1987a), Structural Adjustment and Economic Performance, Paris.
____ (1987b), Historical Statistics 1960–1985, Paris.
OLSON, M. (1982), The Rise and Decline of Nations, New Haven, Conn.: Yale
University Press.
OSWALD, A. J. (1985), ‘The economic theory of trade unions: an introductory
survey’’, Scandinavian Journal of Economics, 87: 160–93.
PALOKANGAS, T. (1989), ‘Union power in the long run reconsidered’, Scandinavian
Journal of Economics, 91: 621–35.
PIORE, M. J. (1986), ‘Perspectives on labour market flexibility’, Industrial Relations,
25: 146–66.
____ and SABEL, C. F. (1984), The Second Industrial Divide: Possibilities for Prosperity,
New York: Basic Books.
PLOEG, F. VAN DER (1987), ‘Trade unions, investment, and employment: a non-
cooperative approach’, European Economic Review, 31: 1465–92.
POHJOLA, M. (1984), ‘Union rivalry and economic growth: a differential game
approach’, Scandinavian Journal of Economics, 86: 365–70.
ROWTHORN, B. (1990), ‘Wage dispersion and employment: theories and evidence’,
Department of Applied Economics Working Paper, University of Cambridge.
____ and GLYN, A. (1990), ‘The diversity of unemployment experience since 1973’,
in S. A. Marglin and J. B. Schor (eds.), The Golden Age of Capitalism, Oxford:
Oxford University Press.
CHOTT, K. (1984), ‘Investment, order, and conflict in a simple dynamic model of
capitalism’, in J. H. Goldthorpe (ed.), Order and Conflict in Contemporary
Capitalism, Oxford: Clarendon Press.
SEN, A. K. (1961), ‘On optimising the rate of saving’, Economic Journal, 71: 479–96.
____ (1967), ‘Isolation, assurance and the social rate of discount’, Quarterly Journal
of Economics, 81:112–24.
SÖDERSTRÖM, H., and VIOTTI, S. (1979), ‘Money wage disturbances and the
endogeneity of the public sector in an open economy’, in A. Lindbeck (ed.),
Inflation and Employment in Open Economies, Amsterdam: North-Holland.
SOLOW, R. E. (1986), ‘Unemployment: getting the questions right’, Economica, 53:
S23–S34.
STRAND, J. (1987), ‘Oligopoly with monopoly unions’, FIEF Working Paper No. 33.
TARANTELLI, E. (1986), ‘The regulation of inflation and unemployment’, Industrial
Relations, 25: 1–15.
WIBE, S. (1987), ‘Empirical evidence of Mancur Olson’s theory of the rise and decline
of nations’ (mimeo).
4
Corporatism and Labour Market Performance
Bob Rowthorn
4.1 INTRODUCTION
The present chapter is concerned with labour market performance in cor-

poratist economies. The term ‘corporatist’ has many different meanings, but
in this chapter it will be used to denote capitalist economies in which there
are strong, and relatively centralized, employers’ and workers’ organizations.
The economies which readily fit this description, and to which the chapter
is primarily devoted, are Austria and the Nordic countries—Denmark, Fin-
land, Norway, and Sweden.1 In recent years, the corporatist economies have
attracted increasing attention from social scientists, especially Sweden, about
which there is now a voluminous literature. Amongst the reasons for such
interest are the success exhibited by some of these economies in responding
to external shocks during the turbulent years since 1973, and the apparently
good labour market performance they have displayed over this period.
Labour market performance can be measured in a variety of ways. At one
time, international comparisons focused mainly on unemployment rates as
the simplest and most readily available index of performance. However, it is
now accepted that this measure has major drawbacks which seriously under-
mine its usefulness in international comparisons. For example, Switzerland
experienced a huge loss of jobs in the crisis of the mid-1970s, but because
of the way Swiss statistics are compiled, there was virtually no rise at all in
the official unemployment rate, and the country was widely, but mistakenly,
praised for its handling of the crisis.2 Denmark, on the other hand, is often
classified as a failure because of its high unemployment rate (8 per cent in
1987). Yet in terms of jobs, the country’s record has been quite good by inter-
national standards, and the high rate of unemployment in Denmark is partly
due to a high rate of labour force participation. Because of these limitations,
I should like to thank the participants of this project for their helpful comments, especially
Andrew Glyn, Jukka Pekkarinen, and Matti Pohjola.
1
In addition to the countries just listed Australia has also recently developed a centralized
system of labour market institutions, whose features are described in a companion essay by
Robin Archer elsewhere in this volume. However, the Australian system is still in its infancy and
has not been in existence long enough to produce a noticeable impact in the time period covered
by this chapter, which ends in 1985. Special attention will not therefore be devoted to Australia.
2
For a discussion of the Swiss case see Schmidt (1987), Therborn (1986), Rowthorn and
Glyn (1990), and Wolfgang Blaas in the present volume.
LABOUR MARKET PERFORMANCE 83
the use of unemployment statistics for international comparisons has been

increasingly questioned, and attention is shifting towards alternative indices of
labour market performance, such as the level or growth rate of employment.3
However, even these newer indices of performance are seriously flawed,
since they are primarily concerned with the total number of jobs in existence,
and take little or no account of what kind of jobs are involved. In particular,
they ignore variations in pay between one kind of job and another. Some
countries have achieved high levels of employment in recent years, but many
of the jobs in question are very badly paid, with workers receiving only a
fraction of the national average. The result is now a large underclass of
‘working poor’ in these countries. In certain other countries, employment is
just as high—or even higher—but wage rates are comparatively uniform and
even low-paid workers earn a wage fairly close to the average. Such inter-
national differences in wage dispersion are extremely important in human
terms, and also to some extent in efficiency terms. They should clearly be
taken into account when comparing labour market performance.
In the present article evidence on both employment and wage dispersion
is examined, and the performance of the corporatist economies is evaluated
taking both of these variables into account. The picture revealed by our
analysis is in some ways rather surprising, both confirming and refuting
certain common beliefs. It also reveals a clear division within the group of
social corporatist economies. The main findings can be summarized as
follows. Austria, which has been frequently admired for its labour market
performance since 1973, does badly and emerges with a rating which is at
best mediocre by international standards. This is explained primarily by the
fact that employment opportunities for women are severely limited in Austria
and their pay is on average well below that of men. In addition, wage
dispersion amongst men is considerable and there is a substantial body of
low-paid male workers. Conversely, all of the Nordic countries, without
exception, do well and enjoy a clear lead over the rest of the OECD during
the period since 1973. All of them have managed to achieve, or maintain,
comparatively high levels of employment for both men and women, and as
a rule wage dispersion is relatively low; in particular, the earnings gap
between men and women is smaller than in most other countries. Thus, in
terms of employment and wages, the Nordic countries are strikingly more
egalitarian than Austria. This points to a fundamental difference between
social corporatism of the Nordic variety and that which obtains in Austria.
Later we shall speculate briefly on what is responsible for this difference.
Before considering at length the evidence on wage dispersion and employ-
ment, some preliminary remarks on the subject of centralization will be
useful. These will clear the ground for the later discussion.
3
See Bruno and Sachs (1985), Rowthorn and Glyn (1990), and Therborn (1986).
84 BOB ROWTHORN
4.2 CENTRALIZATION
The corporatist economies are distinguished above all by the nature of their
labour market institutions. In these economies, workers and employers are
organized into a small number of centralized federations, which are able to
exert a significant degree of control over the behaviour of their members.
Sometimes these federations negotiate directly with other bodies on behalf
of their members, and sometimes they merely provide a framework and
guide-lines for more decentralized bargaining. In each case, the central
federations have various means at their disposal to influence the course of
events, and they can normally ensure that the final outcome is broadly in
accordance with their overall objectives. Their powers are not absolute and
on occasion these federations can virtually lose control over events. But for
the most part their influence is considerable.
One important feature of the corporatist economies is the role of the state
in wage determination. This varies considerably through time and space but
is frequently of great significance. Sometimes the state participates directly
in negotiations, offering various inducements, like tax cuts or increased
welfare expenditure, in return for wage restraint. Sometimes it imposes a
compulsory incomes policy on the parties concerned—though such action
is rarely taken without some kind of formal or informal negotiations. Policies
of this type have been imposed on a number of occasions in both Denmark
and Norway. Finally, even where it does not overtly intervene in wage
bargaining, the state may still play a major role simply by threatening to
intervene if the various parties cannot reach an acceptable agreement.
Threats of this type have helped to shape the evolution of wage bargaining
in Sweden.4
To summarize, one could say that the political economy of the corporatist
economies is characterized by bargaining between centralized Labour and
centralized Capital, with a significant, though variable, role for the state. The
content of bargaining may vary through time and space, but always includes
wages and conditions of work. It may also include various aspects of
government policy.
A crucial aspect of the corporatist economies is that wage bargaining is
dominated by a small number of powerful agents. Because of their individual
power and small number, such agents have an incentive to behave ‘stra-
tegically’, taking explicit account of how their actions will affect others—and
what the resulting feedback to themselves will be. Moreover, the interactions
between these agents are normally of an ongoing variety and continue over
indefinitely long periods of time. As the theory of repeated games shows,
such conditions are conducive to compromise and to forms of’co-operative’
4
See Braun (1986) for information on the role of the state in wage determination in the
corporatist economies.
behaviour which are to the mutual benefit of all concerned. The fact that
bargaining is dominated by a handful of powerful agents does not in itself
automatically ensure that there will be compromise and co-operation, but it
does mean that such behaviour is more likely than under a more atomistic
regime. It is for this reason that many economists and political scientists
believe that centralized wage bargaining is a major factor behind the fre-
quently good labour market performance of corporatist economies.5 How-
ever, although highly plausible, this is only a partial explanation which leaves
many important questions unanswered. Why, for example, has less priority
been attached to full employment in Denmark and Finland than in Norway
and Sweden? Why has unemployment risen significantly in Austria and,
more recently, Norway? Why does Austria exhibit such a persistently high
wage inequality and such a low level of employment for women? Such
questions cannot be answered simply by reference to some general theory of
centralized bargaining, but require an examination of specific features of the
countries concerned, including such items as the balance of power within
the centralized bodies and their resulting priorities.
As Göran Therborn points out elsewhere in the present volume, the term
‘corporatism’ is used in a multitude of different ways by political scientists.6
However, amongst economists the term is generally given a rather specific
meaning, being used merely as a shorthand for centralized wage bargaining.
My usage in the present chapter is consistent with the latter approach,
although I would take a somewhat wider view and argue that bargaining may
cover more items than simply wages, and frequently includes government
policy. Indeed, I would emphasize the often crucial role of the state, which
is stressed by political scientists, but usually ignored or downplayed by
economists.
In emphasizing centralization, I am excluding decentralized systems of
consensus formation of the type which, according to Katzenstein (1984),
apparently obtains in Switzerland.7 Under certain conditions, such a system
can produce outcomes similar to those which theory predicts are likely in
more centralized economies. For this reason, Switzerland will on occasion
be singled out for special attention in the following discussion, along with
Austria and the Nordic countries which are our main concern.
In this chapter it will be helpful to have a numerical index of centralization
in wage bargaining. Of the various measures available, the most satisfactory
is the index constructed by Calmfors and Driffill (1988). Their index has
several advantages over its potential rivals, of which the most important is
5
For a critical review of these ideas see the chapter by Matti Pohjola in the present volume.
6
For a heroic, but ultimately unconvincing, attempt to distil some common element from the
numerous definitions of corporatism put forward by political scientists see Williamson (1989).
7
Katzenstein uses the term ‘liberal’ corporatism to denote such decentralized, consensus-
generating systems.
86 BOB ROWTHORN
the importance they attach to employers’ organizations, which are weighted

equally in this index with workers’ organizations. Most other measures of
centralization, by contrast, take little or no account of the degree of organiza-
tion amongst employers.8 The Calmfors–Driffill Index is not perfect and like
any other measure it has drawbacks, but it is the most suitable available, and
wherever a numerical index of centralization is required, it is the one I shall
use.9 Details of the Calmfors–Driffill Index and its construction are given in
Appendix Table 4.A1. It is interesting to note that the five corporatist coun-
tries all rank at the top of the scale according to this index, whilst Switzerland
is close to the bottom. Throughout the present article, where comparative
international statistics are tabulated, the order of countries will normally be
in accordance with the Calmfors–Driffill ranking, with the most centralized
(Austria, Norway, etc.) at the top and the least centralized (the USA and
Canada) at the bottom.
4.3 WAGE DISPERSION: CONCEPTS AND EVIDENCE
The subject of wage dispersion raises a number of conceptual and measure-

ment problems. First, there is the question: what do we mean by wages?
Which items should be included in income and which excluded? Should we
include the social costs paid by employers on behalf of workers? Should we
deduct taxes paid by workers or the costs they have incurred in their training?
Should we be concerned with earnings at a given moment of time or with
earnings over the whole of a person’s working life? These are not trivial
questions because the magnitudes involved are often very large.
Some idea of their scale can be gauged from the following example drawn
from a study commissioned by the Swedish Employers’ Federation (SAF).10
According to this study, the total cost to a Swedish firm in 1982 of employing
a graduate economist, including all payroll and other social charges, was on
average 2.25 times as much as the cost of employing an industrial worker.
When social charges are deducted, this differential falls to 2.15, and when
8
For a description of the various indices proposed by political scientists see Pryor (1988). It
should be stressed that these indices were never intended simply to measure centralization in
wage bargaining, but were designed for the far more ambitious, and perhaps futile, task of
ranking entire national systems of political economy. Even so, a casual treatment of employers’
organizations is a weakness apparently shared by all of these indices.
9
One possible criticism of the Calmfors–Driffill Index concerns Italy, whose wage-bargaining
system is ranked as one of the most decentralized in Europe according to this index. Such a
ranking ignores the role of political parties, especially the Communist Party, which exert a
significant centralizing influence over wage bargaining in Italy. However, to avoid confusion,
no attempt is made to remedy this defect, and the ranking used here is based on the original
Calmfors–Driffill Index. Fortunately, none of our main statistical conclusions would be affected
by modifying the Calmfors–Driffill value for Italy to reflect more accurately the situation in that
country.
10
Jonsson and Siven (1986).
income taxes are also deducted it falls still further to 1.56. Finally, when
account is taken of the fact that graduate economists spend longer in
education and have a shorter working life than industrial workers, the
differential falls to 1.32 for post-tax, lifetime earnings. These numbers
should be treated with caution as they have been calculated by authors who
are highly partisan in their advocacy of greater differentials, but even so they
do indicate the orders of magnitude potentially involved. Moreover, this is
not the end of the problem. One could also take account of the so-called
‘social wage’, which is provided by the state in the form of free or highly
subsidized social services for workers and their families. In some countries,
especially in Western Europe, this social wage is often very large and is
available either to all workers irrespective of their financial situation or else
preferentially to poorer workers. As a result, the true standard of living of
workers in these countries is often more uniform than data on cash earnings
might suggest.11
Despite these qualifications, our discussion of wage dispersion will be
based entirely on pre-tax earnings, excluding employer payroll charges and
the like. This is not an ideal measure of income, but it is the only one for
which anything like comprehensive international statistics are available.
Moreover, although not ideal, this measure is adequate for our purposes and
provides a reasonable indication of what has been happening internationally
to the kinds of wage differentials which interest us. In particular, it allows us
to identify important differences between countries with diverse bargaining
structures.
4.3.1 Occupational differentials

Readily available international data on wage dispersion are of two kinds:
occupational and industrial. The most useful source of information on
occupational differentials in earnings is an article specifically devoted to this
topic in the OECD Employment Outlook (1987). Unfortunately, this infor-
mation has certain limitations. In particular, international differences in
methods of occupational classification are considerable, so that cross-
country comparisons of occupational differentials can only be impression-
istic. On the other hand, this information does give a reasonable indication
of what has happened to such differentials through the course of time within
particular countries. Even here there are difficulties, but the general picture
is fairly clear.
In the Nordic countries there has been a marked reduction in the differ-
ential between manual and non-manual workers over the past two decades,
larger in fact than in most other countries. In the case of Sweden this
11
This point is discussed in the closing section of the present chapter.
88 BOB ROWTHORN
development is a continuation of a trend dating back at least to the 1940s.

In the other Nordic countries its origin is somewhat later. Evidence pre-
sented in the OECD study suggests that differentials have also narrowed
within the two broad occupational categories. In Sweden, for example, the
gap between skilled and unskilled manual workers has diminished, whilst in
both Norway and Sweden earnings differentials between various grades of
salaried employee have narrowed over the past twenty years. Although the
OECD provides no information on the other Nordic countries, it seems
likely that similar developments have occurred there as well.
These developments are not unique to the Nordic countries and several
other countries have also experienced a decline in occupational differentials.
Even so, the decline in the Nordic countries is striking in scale and has oc-
curred over a period when occupation differentials have remained constant
or even widened in countries like Canada, the USA, Germany, and the UK.
Another striking feature of this decline has been its ending and even partial
reversal in the 1980s. For the time being, at least, the process of equalization
in the Nordic countries has apparently come to a halt.
The experience of our other corporatist country, Austria, over the past
twenty years has been very different from that of the Nordic countries. There
has been no move at all towards equalization, differentials have been main-
tained, and the structure of occupational earnings has remained largely un-
changed since the late 1960s. For example, the ratio of non-manual to
manual earnings in Austria has shown only minor fluctuations during this
period with no sign of any real trend. This finding is consistent with the
evidence on inter-industry wage dispersion which, as we shall see, reveals no
major changes in the overall structure of earnings in Austria.
4.3.2 Inter-industry wage dispersion

The most comprehensive and useful information for international com-
parisons of wage dispersion refers to industrial, as opposed to occupational,
earnings. This information is collated by international agencies such as the
UN, Eurostat, and the ILO or by certain governmental or private agencies,
such as the American Bureau of Labor Statistics or the Swedish Employers’
Federation. Most of these agencies publish series of average hourly earnings
for a range of countries and for a range of manufacturing industries, together
with mining, construction, and sometimes transport. The data I shall use are
drawn mainly from the ILO Yearbook of Labour Statistics. There are a few
gaps in the ILO data, some of which have been filled by interpolation using
information drawn from the UN Yearbook of Industrial Statistics or from
material provided by national governments and private organizations. The
result is a complete and, hopefully, internationally comparable data set on
average hourly earnings (and employment) for the years 1973 and 1985. It
covers manufacturing, mining, and construction in seventeen of the most

advanced OECD countries. The data are differentiated by gender and
broken down into approximately fifteen to twenty-five industries, depending
on the country concerned.
A common method for dealing with data of this kind is to compute the
coefficient of variation of earnings across industries. This coefficient, which
is equal to the standard deviation divided by the mean, provides a simple
and illuminating measure of wage dispersion which can be readily used for
international comparisons. There are various ways of computing this coeffi-
cient. The conventional approach is to use unweighted, gender-blind in-
dustrial averages. Thus, all industries are given equal weight, no matter how
large or small they may be, and no distinction is made between male and
female earnings, which are both subsumed under the same industry-wide
average.
The drawback of the unweighted method is obvious. Because it assigns
equal weight to all industries irrespective of size, the resulting measure of
dispersion is highly sensitive to the presence of minor outliers, where
employment is extremely small and earnings are way out of line with those
elsewhere in the economy. To deal with this problem we shall weight all
observations by the number of people employed in the category concerned.
This greatly reduces the influence of minor outliers and gives a more accurate
picture of the true distribution of earnings.12
Our second departure from the conventional approach is to distinguish
explicitly between men and women. Not only do we compute coefficients of
variation for the two groups individually, we also compute an overall coeffi-
cient of variation, which is derived by regarding earnings figures for men and
women in each industry as quite separate observations. The resulting
measure of dispersion takes account of many important male-female wage
differentials which are ignored by the conventional approach.
Our final departure concerns coverage. Most analyses of inter-industry
wage dispersion consider only the manufacturing sector. Throughout the
present paper, we shall consider a wider group of industries covering
construction and mining (including petroleum extraction) in addition to
12
A striking example of the distortions arising from the use of unweighted observations is
provided by New Zealand, which is often classified as a country with a very high and increasing
level of industrial wage dispersion (see Freeman 1988). In fact, this is an illusion caused by the
extremely high earnings of a small number of workers in petroleum refining. If petroleum is
excluded, wage dispersion in New Zealand is about average by international standards. Instead
of excluding such outliers, an alternative and more satisfactory approach is to weight observa-
tions by the number of people employed in each industry. This greatly reduces the influence of
small outliers and is the approach we shall normally use in the present chapter. If observations
are weighted by employment, the coefficient of variation is reduced by a half, and New Zealand
is then reclassified as a country with an about average level of wage dispersion. For evidence on
this point see Rowthorn (1990).
90 BOB ROWTHORN
manufacturing. This wider grouping, which we shall refer to as INDUSTRY,

is more representative than manufacturing alone.
Thus, the measure of wage dispersion we shall use is based on separate
observations for men and women in approximately fifteen to twenty-five dif-
ferent industries in manufacturing, mining, and construction; all observa-
tions are weighted by employment. One defect of this measure is the exclusion
of services. This is not by choice, but is due to the paucity of suitable data on
the service sector. Despite a considerable effort, using a variety of published
and unpublished material from national sources, adequate information on
earnings in the service sector could be obtained for only twelve of the coun-
tries in our sample. Suitable data do not seem to be available for the remain-
ing five countries. It is for this reason that we restrict our attention primarily
to the narrower range of activities covered by the heading INDUSTRY,
where data are available for all the countries in our sample.
A useful check on the likely errors arising from this restriction is to see how
measured wage dispersion is affected by the inclusion of services in the twelve
countries where a comparison is feasible. Table 4.1 compares our measure
of wage dispersion, which is based on INDUSTRY, with the corresponding
measure obtained when services are also included. The main effect of in-
cluding services is to raise the absolute level of wage dispersion somewhat.
However, the ranking of countries remains much the same and the cross-
country correlation between the two measures is extremely high (R-squared
= 0.91). This suggests that our measure, despite its limitations, is a good
proxy for the more comprehensive measure obtained when the service sector
is also included.13
4.3.3 Statistical evidence

After these methodological points let us now examine the statistical informa-
tion on wage dispersion which is provided in Tables 4.1 and 4.2. Further
information is given in Appendix 4. A. It would be tedious to go through these
tables in detail, so I shall merely summarize the main points here.
Countries are ranked in the tables according to their degree of centraliza-
tion in wage bargaining, as indicated by the Calmfors–Driffill Index, with
the most centralized at the top. Table 4.1 shows the overall position in 1985.
It gives figures for wage dispersion, which are derived mainly from the ILO
Yearbook, together with the corresponding information on male–female wage
differentials. To complement these figures, the table also provides some
information on earnings in the service sector, where wage dispersion is often
higher than in industry and a large number of low-paid women are frequently
employed.
13
For a more comprehensive comparison of various indices of wage dispersion see Rowthorn
(1990).
TABLE 4.1 Dispersion of hourly earnings in the late 1980s

(mostly manual workers)
Coefficient of F/M ratio
variation (%)a (%)
ILO NAT. NAT. ILO NAT
IND. IND. PS and IND. PS and
1985 I. 1985 I.
(1) (2) (3) (4) (5) National data
Austria 21 18 27 70 64 Arbeiter 1986
Norway 13 12 15 82 87 Non-supervisors 1985

Sweden 8 9 10 87 87 Wage earners 1987
Denmark 9 13 14 85 85 Wage earners 1987
Finland 16 76 (skilled separately)
Germany 12 73
The Netherlands 11 11 17 77 72 Workers 1985
Belgium 15 77
New Zealand 19 72
Australia 18 17 18 74 80 Non-managers 1984
France 12 19 18 82 82 Ouvriers 1982
UK 17 18 21 73 64 Manual 1988
Italy 8 86 (part-time separately)
Japan 29 34 35 49 53 Regular workers 1985
Switzerland 16 17 18 70 71 Ouvriers 1985
USA 22 20 30 68 68 Hourly paid 1987
Canada 22 21 28 64 67 Hourly paid 1986
a
Weighted by employment, male and female earnings counted as distinct
observations.
Notes: IND. = manufacturing, mining, and construction; PS and I. = private services
(excluding health and education) plus IND.; NAT. = national data sources, data for late
1980s; F/M ratio = ratio of average female to average male hourly earnings. Countries in this
and all similar tables are ranked according to the Calmfors and Driffill index of centralization.
Source: ILO and national sources.
Comparing the various columns in this table, we find a broad measure of

agreement. Most European countries have rather similar levels of wage
dispersion and are in the middle range by international standards. In
Denmark and Sweden, wage dispersion is well below the European average.
The same is apparently true of Italy, although because of deficiencies in the
Italian statistics, the figure shown may not be very accurate. Outside of
Europe, in North America and Japan especially, dispersion is typically
higher. Indeed, if the service sector is included, the level of wage dispersion
in North America and Japan is three times the level observed in Sweden and
approximately twice the European average. One factor behind this difference
92 BOB ROWTHORN
TABLE 4.2 Decomposing changes in wage dispersion, 1973–1985

Coefficient of % Breakdown of % Other
variationa change change
empl f/m
1973 1985 eff Eff
Austria 22.0 21.3 –3.2 –0.4 –6.2 3.5
Norway 14.3 13.4 –6.6 –0.9 –13.8 8.1
Sweden 11.8 8.5 –28.4 –3.9 –15.4 –9.2
Denmark 11.1 8.6 –22.8 –3.7 –11.5 –7.2
Finland 20.2 15.9 –21.6 –2.9 –17.7 –1.0
Germany 13.5 12.4 –7.9 –5.5 –4.4 2.0
The Netherlands 8.8 10.5 19.2 6.1 2.2 11.0
Belgium 20.6 14.8 –27.9 –5.9 –17.1 –5.0
New Zealand 17.8 18.8 5.7 0.6 –8.9 14.0
Australia 15.3 17.6 15.2 8.8 –11.5 17.9
France 12.8 12.1 –5.4 3.8 –8.6 –0.6
UK 19.6 17.0 –13.5 –3.6 –23.2 13.3
Italy 15.3 7.7 –49.8 –5.8 –17.8 –13.3
Japan 27.1 29.3 8.1 –0.9 7.7 1.3
Switzerland 15.7 16.3 3.5 –1.0 –8.6 13.1
USA 26.7 22.2 –16.8 0.4 –9.0 –8.2
Canada 26.9 22.5 –16.5 –1.9 –9.6 –5.0
Geometric average 16.8 14.8 –12.0 –1.0 –10.0 1.0
a
INDUSTRY (manufacturing, mining, and construction), employment
weighted, male and female earnings regarded as separate observations.
Source: Mainly ILO, supplemented by data from UN, OECD, and national sources.
is the relatively high level of female wages in many European countries.

Women’s hourly pay in these countries is often approaching 80 per cent or
more of the male average, as compared to less than 70 per cent in North
America and 50 per cent in Japan. Indeed, such cross-country variations in
the male-female differential explain a considerable part of all the interna-
tional variations in wage dispersion. Countries with a very low ratio of female
to male wages normally have a very high level of inter-industry wage
dispersion, and vice versa. This is only to be expected with our measure of
dispersion, which is specifically designed to highlight differentials between
men and women. However, it also applies when more conventional measures
are used. No matter what measure is used, there is a clear inverse correlation
between the relative pay of women and the level of inter-industry wage
dispersion.14
14
See Rowthorn (1990).
Perhaps the most surprising feature of Table 4.1 is the extremely high level
of wage dispersion in Austria. No matter which indicator we look at and no
matter which data source we use, Austria has a wage dispersion similar to that
observed in Canada or the United States. It is exceeded significantly only by
Japan. The example of Austria illustrates that centralized wage bargaining is
not synonymous with egalitarianism. In the Nordic countries, centralized
wage bargaining has been accompanied by a policy of deliberately squeezing
differentials. This has not been the case in Austria, where there has been no
systematic attempt to reduce differentials which remain as high as they were
in the 1960s, and sometimes even higher.
4.3.4 Changes through time

Table 4.2 provides information on how wage dispersion has changed since
1973 and on some of the factors responsible for these changes. In most
countries wage dispersion has fallen, sometimes dramatically. However, the
pattern of reductions is interesting. All of the Nordic countries, with the
exception of Norway, have experienced a large fall in dispersion. So too have
a number of other countries, of which Italy is the most striking case.
According to our measure, the USA and Canada have also experienced
noticeable reductions, although as already mentioned, the level of dispersion
in these countries is still very high. Moreover, in the case of the United States,
other evidence contradicts our findings and suggests that wage dispersion
has in fact increased somewhat over the past fifteen years.15 Two countries
also worth mentioning are Austria and Japan, both of which have extremely
high levels of wage dispersion and neither of which has experienced any
significant reduction since 1973. The picture is similar if we look at other
indicators of dispersion. Most countries have experienced a considerable fall
in the male-female differential since 1973. In Austria, however, there has
15
Some authors, such as OECD (1985), Bell and Freeman (1985), and Freeman (1988),
base their claim for increasing wage dispersion in the USA on the behaviour of the unweighted
inter-industry coefficient of variation of hourly earnings in the manufacturing sector, making
no distinction between male and female earnings. This measure indicates a rise in dispersion
over the period 1973–85. However, when observations are weighted by employment and
differentiated by sex, the inter-industry coefficient of variation in US manufacturing remains
more or less constant over the period 1973–85. If coverage is extended to include mining and
construction—as in the present chapter—this coefficient falls over the period (see Table 4.2).
More substantial evidence for increasing dispersion is provided by Harrison and Bluestone
(1988 and 1989), who argue that there has been a polarization of earnings amongst workers of
the same sex employed in the same industry, with an increase in the proportion of high- and
low-paid workers and a decline of people in the middle. Their evidence on this point is.
convincing and points to weakness in the measure used in the present chapter. However, despite
the different picture they give concerning changes over time, the various indicators available for
measuring wage dispersion all agree on one thing. The USA has consistently exhibited a very-
high level of dispersion by international standards over the period covered by the present article.
This is, perhaps, the fundamental point.
94 BOB ROWTHORN
been almost no reduction in this differential and the average female wage is
still only 64 per cent of its male counterpart. Even more striking is Japan,
which has the dubious distinction of being the only country where the
male-female wage differential has actually widened since 1973.
To round off the analysis, let us consider the results of a simple ‘shift-
share’ analysis which are reported in Table 4.2. This table decomposes
changes in the coefficient of variation since 1973 into three components. The
first, labelled ‘emp eff, indicates to what extent changes in the coefficient of
variation are due to shifts in the pattern of employment and hence in the
weights used to compute this coefficient. The second, labelled ‘f/m eff,
measures the influence of changes in the male-female wage gap. Finally, the
third component is a residual reflecting changes in male and female earnings
differentials; it also includes interaction effects, which are typically very
small. It is clear from the table that the effect of changes in employment
structure on wage dispersion has been small. Far more important in most
countries has been the narrowing of the male-female wage gap, which is the
main factor responsible for the widespread reduction in wage dispersion over
the period concerned. Other factors have also influenced the evolution of wage
dispersion to a significant degree, but no clear systematic pattern is observable.
Most of the preceding conclusions, it must be stressed, are extremely
robust. No matter what indicators of dispersion we consider, including those
not discussed here, the general picture is always the same. Austria, North
America, and Japan always have very high wage dispersion. Conversely,
Sweden and Denmark have very low wage dispersion, and most other
European countries are somewhere in the middle. Moreover, changes in the
male-female wage gap are always the main systematic factor responsible for
international differences in wage dispersion and changes in dispersion since
1973.16
4.3.5 Centralization and wage dispersion

Some of these findings are summarised in Figs. 4.1a and 4.16, which show
the situation with regard to wage dispersion in 1973 and 1989. They also
illustrate the relationship between wage dispersion and centralization in
bargaining. In both years there is an inverse correlation between these two
16
In the conventional measure of dispersion, earnings are not distinguished by gender but
simply averaged across all workers in the industry concerned. Even in this case, however,
declining male-female wage differentials are the main factor statistically responsible for the
observed fall in wage dispersion in OECD countries. The point is that an increase in the relative
pay of women will affect various industries differently. In industries employing a large number
of women the average rate of pay will rise relative to the national average, whilst the opposite
will occur in industries employing a low proportion of women. As a result, inter-industry wage
dispersion will fall, even though the measure concerned takes no explicit account of gender-
based differentials. This point is made in Bell and Freeman (1985).
variables. In particular, the relatively centralized economies of Norway,

Denmark, and Sweden have comparatively low wage dispersion, whilst the
highly decentralized economies of North America and Japan have extremely
high dispersion. However, there are some major outliers. For example, there
is Italy, whose bargaining system is normally classified as decentralized, but
which in 1985 had a level of wage dispersion as low as Sweden. At the other
extreme, highly centralized Austria has consistently had a very high level of
dispersion. However, despite there exceptions, there does appear to be
something of an inverse relationship between centralization and wage dis-
persion.
The second point concerns changes through time. As can be seen from
Table 4.2, the picture here is very confused and no clear pattern can be
observed. At all levels of centralization, there are countries which have
experienced a notable reduction in wage dispersion since 1973. In most
Nordic countries, solidaristic bargaining has been deliberately used to
squeeze wage differentials, but a significant reduction, albeit from a much
higher level, has occurred spontaneously in some more decentralized eco-
nomies. Likewise, at all levels of centralization, there are economies where
wage dispersion has remained roughly constant or even increased. Such
diversity invalidates the notion of a universal link between centralization and
wage equality. It also points to an important diversity within the social
corporatist group—between the Nordic countries and Austria. Centralized
bargaining has certainly been used to reduce wage differentials in the former
countries but not in Austria where they remain very high. This suggests the
existence of two different varieties of corporatism, which are both compar-
atively centralized, but which differ in their social objectives and, presum-
ably, in the nature of their institutions and in the social compromises which
underlie them. We shall return to this point below.
4.3.6 Centralization and employment

In recent years there has been talk of a U-shaped relationship between
employment and centralization of wage bargaining.17 Both highly centralized
and highly decentralized economies, it is said, have above average levels of
employment, whilst employment is normally much lower in the intermediate
countries. The amount of empirical support for this proposition depends on
the time period concerned. If we go back to 1973, there is only weak evidence
of any U-shaped relationship between the two variables. This is clear from
Fig. 4.1 and is confirmed in Table 4.3, which reports the results of a
regression analysis in which the employment rate was regressed against the
17
For a critical survey of this literature see the chapter by Matti Pohjola in the present
volume.
96 BOB ROWTHORN
FIG. 4.1 Bargaining structure and wage dispersion
first and second powers of a variable DECEN, representing decentralization

in wage bargaining.18 In no case are any of the relevant coefficients for 1973
statistically significant—though in the case of women they are at least of the
18
The variable DECEN is merely the rank of a country according to the Calmfors–Driffill
Index of centralization. The more decentralized (less centralized) a country is, the greater is its
numerical rank and hence the greater is the value of DECEN. Thus, for the least decentralized
(most centralized) country, Austria, the value of DECEN is 1; for the most decentralized (least
centralized) country, Canada, die value of DECEN is 17.
TABLE 4.3 Regression analysis: employment rate and centralization

(a) No Austrian dummy
Independent variables
DECEN DECEN SQ. FEM R2
1973
Males 0.01(0.01) 0.00(0.07) 0.01
Females –2.92(1.42) 0.13(1.18) 0.17
Pooled –1.45(1.22) 0.07(1.05) –36.4**(13.23) 0.85
1985
Males –2.54*(1.90) 0.14*(1.95) 0.21
Females –4.83*(2.13) 0.23*(1.87) 0.27
Pooled –3.68**(–2.82) 0.18**(2.62) –23.2**(7.71) 0.69
Change
1973–85
Male –2.55**(3.03) 0.14**(2.98) 0.40
Females –1.90(1.55) 0.10(1.50) 0.15
Pooled –2.23**(3.09) 0.11**(3.01) 13.2**(7.93) 0.71
(b) Including dummy for Austria
Independent variables
DECEN DECEN SQ. AUS FEM R2
1973
Males 0.26(0.17) –0.01(0.08) 2.1(0.28) 0.02
Females –4.31(1.71) 0.19(1.50) –12.3(0.96) 0.22
Pooled –2.03(1.38) 0.09(1.23) –5.1(0.68) 36.4**(13.11) 0.86
1985
Males –3.43*(2.10) 0.18*(2.16) –7.9*(0.95) 0.26
Females –8.59**(3.86) 0.40**(3.51) –33.5** (2.94) 0.56
Pooled –6.01** (4.18) 0.29**(3.94) 20.1**(2.82) –23.2 (8.55) 0.76
Change
1973–85
Males –3.68**(4.01) 0.19**(3.97) –10.1**(2.15) 0.55
Females –4.28**(4.07) 0.21**(3.85) –21.1**(3.93) 0.61
Pooled –3.98**(5.73) 0.20**(5.53) –15.6**(4.39) 13.2**(10.10) 0.82
Notes: The dependent variable in these equations is the employment rate (employment as a
percentage of the relevant population aged 15–64 years). For clarity the constant term is not
shown in these equations. Number of countries =17. Pooled regressions combine male and
female observations in one sample, giving a total of 34 observations. DECEN is a measure of
decentralization in wage bargaining (= Calmfors–Driffill rank as given in Table 4.2) and
DECEN SQ. is its squared value; AUS is a dummy variable for Austria, FEM is a dummy
variable for female observations. Absolute t-values are shown in brackets; * (**) denotes
significance at the 1% (5%) level on a one-tailed t-test.
98 BOB ROWTHORN
right sign. The situation, however, is quite different for 1985, where there is
a well-defined U-shaped relationship between the centralization and the
employment rate (Fig. 4.26). This is true for both men and women taken
separately, and when male and female observations are pooled in one
sample. It is also true when we look at changes in the employment rate since
1973. As a general rule, both highly centralized and highly decentralized
FIG. 4.2 Bargaining structure and employment

economies have experienced larger than average increases in the employ-

ment rate (or less than average reductions in the case of men), whilst
intermediate countries have for the most part performed much worse. This
is evident from Table 4.3, where the relevant coefficients are invariably of
the correct sign and for the most part statistically significant.19
Thus, the U-shaped relationship does seem to be well established for the
period since 1973. However, there are several major exceptions. Of particular
interest is Austria, whose employment performance since 1973 has been
poor despite its centralized bargaining structure. The employment rate for
men in Austria has fallen roughly in line with the international average, whilst
it is one of the few countries to have experienced an actual reduction in the
female employment rate. This reduction is striking in view of the already low
employment rate for women in Austria in 1973. Thus, in terms of both
employment and wage dispersion, Austria is a clear exception amongst the
corporatist economies.
Since it is such an exceptional country, Table 4.3 also reports the results
of regression analysis when a dummy variable for Austria is included. For
changes in the employment rate over the period 1973–85, the dummy
coefficients for Austria are uniformly large, negative, and statistically sig-
nificant. The same is true for absolute levels in 1985. In the case of women,
the employment rate in 1985 is a full 33 percentage points below the value
expected given the country’s level of centralization, whilst for men the
shortfall is 8 points. When the Austrian dummy is included, our previous
conclusions concerning the U-shaped relationship are unaffected. For 1973
this relationship remains weakly defined. And in the case of the other
regressions it becomes even more pronounced than before. Indeed, for
changes over the period 1973–85, coefficients of DECEN and DECEN SQ.
are now significant at the 0.1 per cent level. These results provide additional
evidence for the existence in recent years of a U-shaped relationship between
employment and centralization.
4.4 MEASURING LABOUR MARKET PERFORMANCE
Most conventional measures of employment look simply at the number of

people in work, making no allowance at all for the quality of the jobs in
question. This is the procedure we followed in the preceding section.
However, it is open to objection on several counts. First, it ignores the
meaning of employment from the worker’s point of view. Some jobs are
clearly more desirable than others because of their intrinsic qualities or the
19
Similar results are obtained if employment is measured in terms of full-time equivalents,
although the regression results in this case are not reported here.
100 BOB ROWTHORN
income they yield. If a country creates employment by forcing people into

low-paid, marginal jobs, its performance from a human point of view is
inferior to that of a country which provides an equivalent amount of
well-paid, secure employment. Moreover, there is also an important effici-
ency dimension to consider. Low-paid jobs are often low-productivity jobs,
and economies with an extremely high dispersion of wages will almost
certainly contain a great deal of allocative inefficiency. There will be many
people working in low-paid, low-productivity activities who could be more
efficiently employed elsewhere in the economy. There will also be many
inefficient firms kept in existence or subsidized by their ability to pay wages
far below the average.
In some cases, inefficiency arises from the misallocation of the existing
work-force with its existing skills. Some low-productivity workers could
almost certainly be redeployed elsewhere in the economy with little retrain-
ing or habituation, provided the obstacles to their redeployment were
removed. In other cases, workers may lack the skill or motivation required
for more productive employment elsewhere. Such deficiencies may be the
result of some innate incapacity in the workers concerned, in which case their
low productivity is not a symptom of allocative inefficiency. More frequently,
however, these deficiencies are the result of social and economic mechanisms
which inhibit the acquisition of skills and motivation. Failure to acquire skills
and lack of motivation are really a form of allocative inefficiency, since they
imply a waste of productive potential and a misdirection of human abilities.
This is increasingly recognized in the case of women, whose past and present
exclusion from skilled employment in many countries represents a prodigi-
ous misallocation of resources, but it also applies to many other disad-
vantaged groups and individuals.
4.4.1 The INDEED Index

The above discussion suggests that a very high wage dispersion is prima-facie
evidence that existing employment is highly unsatisfactory from a human
point of view, and of a widespread misallocation of labour and waste of
human potential. Such failings should clearly be taken into account when
assessing the employment performance of a country. How can this be done?
In the present chapter, we shall use a composite index of employment
performance which is similar in form to Okun’s ‘misery index’, though based
on different variables. Our measure will be called the ‘Index of Employment
and Earnings Dispersion’, or INDEED for short. It is defined as follows:
INDEED = ER – DISP
where ER is the employment rate (employment as a percentage of the
population aged 15–64 years), and DISP is a measure of wage dispersion
(inter-industry coefficient of variation in hourly earnings, also expressed as

a percentage). In evaluating this index, an adjustment will be made for
part-time working by measuring employment in full-time equivalents.20
The index just defined has the following properties. It is equal to 100 when
all people aged 15–64 are in full-time employment, no one else is employed,
and everyone in employment receives the same hourly pay. The two variables
which define this index influence it in opposite directions. Other things being
equal, an increase in employment or a reduction in wage dispersion will
increase the value of the index, whilst a reduction in employment or an
increase in dispersion will have the opposite effect.
As just defined, the INDEED Index applies to the entire population aged
15–64 in a country. However, it can also be applied, with suitable modifica-
tions, to subgroups of the population.
For example, for males the following formula provides an index summariz-
ing their employment situation:
INDEEDm = ERm – DISPm
where the subscript ‘m’ indicates that the variable concerned is evaluated for
the males only.
A similar formula could also be used for women, but it is more illuminating
to use a slightly modified version. The labour market situation of women
certainly depends on their employment rate and on the extent of wage
dispersion within the female work-force. However, it also depends on the
relative wage of women as compared to men. To take account of this
differential, we shall use the following modified version of the basic formula:
INDEEDf = (ERf – DISPf) × F?M
where F/M is the average female wage as a percentage of the average male
wage, and the subscript ‘f’ indicates that the variable concerned refers to
females only. This modified index is equal to 100 when all women aged
15–64 are in full-time employment; no other women are employed; and all
women receive the same hourly pay, which is also equal to the average hourly
wage of men.
4.4.2 The situation in 1985

Information about the INDEED Index, and its construction, are given in
Table 4.4 and displayed visually in Figs. 4.3a and 4.3b. As might be
expected, the Nordic countries do best according to this index. In 1985,
Sweden has the highest value, followed by Denmark and then Finland and
Norway. Switzerland also does quite well. Apart from these countries, the
20
Full-time equivalents are calculated by counting each part-time worker as half an employ-
ment unit.
102 BOB ROWTHORN
TABLE 4.4 Aspects of labour market performance, 1985
Abbr. CD Emp. Emp. rate Wage INDEED

rank rate (FT equ.) disp. Index
Austria Aus. 1 63 60 21 39
Norway Norw. 2 75 64 13 51
Sweden Sw. 3 80 70 8 61
Denmark Den. 4 74 65 9 57
Finland Fin. 5 73 70 16 54
Germany FRG 6 59 55 12 42
The Netherlands Neth. 7 51 46 11 35
Belgium Belg. 8 54 52 15 37
New Zealand NZ 9 62 58 19 39
Australia Austral. 10 64 58 18 41
France Fr. 11 58 55 12 43
UK UK 12 65 59 17 42
Italy It. 13 52 51 8 43
Japan Jap. 14 71 67 29 38
Switzerland Switz. 15 71 65 16 49
USA US 16 68 63 22 40
Canada Can. 17 66 60 22 38
Notes: CD rank = rank according to the Calmfors–Driffill Index of centralization in wage

bargaining; Emp. rate = total employment as a % of population aged 15–64 years; Emp. rate
(FT equ.) = Emp. rate but measured in terms of full-time equivalents (part-time
employment counts as half a unit); Wage disp. = inter-industry coefficient of variation of
hourly earnings; observations weighted by employment; male and female earnings recorded
separately; covers INDUSTRY (= manufacturing, mining, and construction); measured in
%; INDEED Index = Emp. rate (FT equ.) - Wage disp.
performance of the rest is remarkably uniform. If we look behind the aggre-
gate figures, the striking feature about the Nordic countries is their ability to
combine (1) relatively high employment for both men and women, (2) low
to average wage dispersion within the two groups, and (3) a low male-female
wage differential. It is this combination which explains why the INDEED
Index for these countries is so high. These various features are most consis-
tently displayed in Sweden and rather less so in the other Nordic countries.
It is interesting to note that Denmark scores highly on our index despite
having an unemployment rate of around 8 per cent. This is explained by the
fact that our index is based on employment rather than unemployment as
conventionally defined. Denmark is an unusual country because of its very
high participation rates for both men and women. These ensure that meas-
ured unemployment is considerable even though the country has one of the
highest employment rates in the OECD. If we had based our index on
unemployment instead of an employment, Denmark’s performance would
have appeared less impressive, though still good by international standards.
FIG. 4.3 Bargaining structure and labour market performance
Behind the apparent uniformity of the remaining countries there is, of

course, diversity. North America and Japan, for example, combine high
levels of employment with high wage dispersion. This gives them about the
same overall score as countries in the European core (and Australasia), where
employment and wage dispersion are both much lower. These differences
indicate the very different patterns of inequality in the two groups. In the
104 BOB ROWTHORN
European core, the greatest inequality in the potential work-force is between

those with jobs and those without. Those fortunate enough to have jobs are,
for the most part, moderately well off. This is not the case, however, in the
United States, where there is a large underclass of what might be called the
‘working poor’. The same is also true, though to a lesser extent, in Canada
and Japan. The achievement of the Nordic countries has been their ability
to minimize both kinds of inequality by providing a large number of com-
paratively well-paid jobs. If it is to mean anything at all, the term ‘full
employment’ must surely signify not just any old job at any old wage, but a
reasonable job for all at a wage reasonably close to the average. According
to this interpretation of the term, no country has really achieved full employ-
ment, although Sweden is nearest to this goal and the other Nordic countries
all do relatively well by international standards.
As always, Austria is the exception amongst the corporatist countries.
With its low measured unemployment rate, the country was for long
regarded as a model of what ‘social partnership’ could achieve in the realm
of employment. In recent years this reputation has become somewhat
tarnished as redundancies in the smokestack industries have highlighted the
structural problems of the Austrian economy. However, our analysis raises
the more fundamental question as to whether this reputation was ever
justified in the first place. Employment rates in Austria have never been
unduly high by international standards, and the country’s low measured
unemployment during the 1970s was due primarily to low participation
rates, especially amongst women. Moreover, many of the men and women
actually in work were employed in very low-paid jobs in the service sector
and elsewhere, whilst the average pay of women was (and still is) well below
that of men. To call such a situation ‘full employment’ is something of a
misnomer.
Although its institutional structure is very different, the Austrian labour
market is rather similar to the US labour market in terms of outcomes. This
can be seen clearly from Table 4.5, comparing wage dispersion in the two
economies. As a contrast it also includes Sweden. According to every single
indicator shown in this table, the level of wage dispersion is dramatically
higher in both Austria and the USA than in Sweden, often by a factor of three
or four. Comparing Austria and the USA, the only noticeable difference is
in male wages, which are somewhat more evenly distributed in Austria,
where the coefficient of variation for industry and private services combined
is 18 per cent, as compared to 24 per cent in the USA. In Sweden the
corresponding figure is 9 per cent. In the case of females, wage dispersion in
industry and private services is the same in Austria and the USA at around
24 per cent. When both sexes are pooled together, wage dispersion in the
two countries is 27 per cent and 30 per cent respectively, as compared to 10 per
cent in Sweden. Thus, despite its impressive array of centralized bargaining
TABLE 4.5 Dispersion of hourly earnings in three countries in the late 1980s
Coefficient of variation (%)

Males Females Pooled F/M ratio (%)
Manufacturing
Austria 13 18 20 71
USA 13 14 21 67
Sweden 7 6 8 90
Industry
Austria 11 19 18 70
USA 13 16 20 71
Sweden 7 6 9 88
Industry and private services
Austria 18 24 27 64
USA 24 23 30 68
Sweden 9 6 10 87
Notes: Industry: mining, manufacturing, and construction; Private services: excludes private
health and education; coefficient of variation is calculated with observations weighted by
employment; in the column headed ‘pooled’, male and female earnings are regarded as
separate observations. The final column shows average female wages as a percentage of the
male average.
Sources: Austria: estimated from data in Statistical Handbook of Austria (Arbeiter 1986;
earnings for Vienna, hours and employment for Austria); USA: unpublished; supplied by US
Bureau of Labor Statistics (hourly paid workers 1987); Sweden: Central Statistical Bureau
(wage earners 1987).
institutions, the Austrian labour market has produced wage outcomes very
similar to its decentralized American counterpart. Moreover, in the case of
employment, the Austrian labour market has performed noticeably worse
than that of the USA, primarily because of its failure to generate female jobs.
4.4.3 Changes through time

So far we have been talking mainly about absolute levels in the mid-1980s.
It is also interesting to examine what has happened to labour market
performance during the preceding decade, for it is changes in performance
rather than absolute levels which indicate the capacity of an economy to
withstand shocks of the kind experienced since 1973. The most striking
feature is how badly most European countries have done. Most have ex-
perienced a huge decline in the INDEED Index for men, with this index
typically recording a fall of between 10 and 19 percentage points over the
period 1973–85 (see Table 4.6). The only exceptions are the Nordic
countries and Italy. Even Finland does quite badly over this period, with the
male INDEED falling by 8 points, mainly due to a sharp fall in the male
106 BOB ROWTHORN
TABLE 4.6 INDEED indices, 1973–1985
1985 Changel973–85
Male Female All Male Female All
Austria 65 20 39 –10 –2 –4
Norway 68 30 51 –2 9 6
Sweden 74 45 61 –2 11 6
Denmark 71 39 57 –3 6 0
Finland 63 42 54 –8 7 3
Germany 65 22 42 –15 –4 –8
The Netherlands 56 17 35 –19 –1 –10
Belgium 54 22 37 –12 4 –2
New Zealand 64 21 39 –11 1 –4
Australia 61 24 41 –18 1 –7
France 57 29 43 –14 –2 –7
UK 62 22 42 –17 1 –4
Italy 65 22 43 –1 11 6
Japan 71 17 38 –3 –1 –3
Switzerland 76 24 49 –15 –9 –11
USA 58 23 40 1 6 8
Canada 60 17 38 –3 5 6
Note: For a definition of the INDEED Index see Table 4.4; the indices given here are
based on employment measured in full-time equivalents.
employment rate in the mid-1970s, since when labour market performance

has been extremely good. Amongst females, there has been an almost
universal improvement in Europe, but the ranking of countries is similar,
with Italy and the Nordic countries recording the largest increases in the
INDEED Index, and other European countries some way behind.
It is interesting to note how well Canada and the USA have done since
1973 and how badly Austria has performed. According to the INDEED
Index, labour market performance is still only about average in the North
America countries, but their improvement over the period 1973–85 has been
massive. In Austria on the other hand, the employment rate has fallen, whilst
wage dispersion has remained unchanged. This is reflected in a reduction in
the INDEED Index for Austria since 1973.
The overall picture of labour market performance in the OECD may be
summarized as follows. The Nordic countries started with a high value of
the INDEED Index and this value has increased considerably since 1973 so
that they now enjoy a commanding lead over almost all other OECD
countries. Austria has experienced a fall in this index and now has a roughly
average value. Italy, Canada, and the USA have experienced an increase in
the INDEED Index similar to that of the Nordic countries, but because of
their very low starting-point are still well behind these countries. Most of the
remaining countries have experienced some reduction in the INDEED
Index since 1973, mainly due to a very large fall in male employment, which
has not been fully offset by improvements in other aspects of performance
such as more jobs for women or greater wage equality. A major feature of
developments in the INDEED Index since 1973 has been a general conver-
gence of different countries towards a common value for this index.
Countries well below the average in 1973 have experienced very large
increases in the value of this index, whilst most other countries have
experienced either no change or some reduction in its value. As a result, most
of the major differences visible in 1973 have now largely disappeared. The
only real exceptions to this convergence process are the Nordic countries
which have increased their lead over the rest even further.
4.5 EMPLOYMENT AND WAGE DISPERSION
We have seen how, since 1973, a U-shaped relationship has apparently

emerged between the degree of decentralization in bargaining and the level
of employment. The question we shall now consider is: what role, if any,
does wage dispersion play in explaining this U-shaped relationship?
Let us first examine briefly how different mechanisms of wage bargaining
impinge on total employment. These mechanisms are of two kinds: macro
and micro. The former operate at an economy-wide level and involve the
deliberate manipulation of average wages to help achieve some macro-
employment objective, whilst the latter operate in local markets for particular
kinds of labour. By their nature, macro-mechanisms presuppose a fairly high
degree of centralization in wage bargaining, since it is only in a centralized
system that the average level of wages can be a target variable. In a highly
decentralized system, the individual micro-bargaining units will usually have
preferences concerning developments at a macro-level, but they will be
individually powerless to implement these preferences. Their bargaining
behaviour will perforce be guided primarily by their own local situation.
Thus, in such a decentralized system, the average wage level is not a policy
variable of individual bargaining units, but is the purely arithmetic con-
sequence of a whole series of uncoordinated local bargains.
What are the implications of this for wage dispersion and inequality? The
answer is not clear cut. In a centralized bargaining structure, where the
centre has the formal ability to control relative wages, the level of wage
dispersion will depend on the policy objectives of this centre. If a major
objective is to achieve an egalitarian wage structure, as it has been in the
Nordic countries, especially Denmark and Sweden, then dispersion will be
low. On the other hand, if the objective is to achieve some given, but
108 BOB ROWTHORN
inegalitarian, structure, then wage dispersion will be high. Finally, the centre
may have no strong preferences with regard to wage dispersion, and its policy
may be simply to manipulate the average wage level, whilst modifying relative
wages in response to local conditions. In the latter case, the centre retains
formal control over relative wages, but their ultimate pattern is determined
by local bargaining and market forces. According to some accounts this is
the position in Austria.21
Where relative wages are determined primarily by local bargaining and
market forces, the level of wage dispersion may be high or low depending on
the extent of labour market segmentation, and also to a lesser extent on the
level of aggregate demand. If the labour market is highly segmented, there
will almost certainly be large inequalities in the work-force, with some
workers enjoying privileged access to well-paid jobs and others crowded into
low-paid activities. This will be especially true in periods of recession when
aggregate demand is depressed, but wage dispersion will also be high even
in times of boom. The United States provides a clear example of such an
economy, where a combination of segmentation (including unequal access
to skills) and decentralized bargaining results in a high level of wage disper-
sion at all stages of the business cycle.
Extreme inequality is not an inevitable result of highly decentralized
bargaining. Even in decentralized economies, a concerted drive to reduce
segmentation may have a significant equalizing effect on wages. If dis-
crimination and other major obstacles to labour mobility are tackled, and
good facilities for retraining, the acquisition of skills, child-care, and the like
are provided, many structural inequalities in the labour market will be
eliminated. Micro-flexibility in response to local markets may then result in
a reasonably egalitarian wage distribution. Thus, despite the absence of any
formal wages policy, decentralized economies can make considerable
progress towards wage equality through active measures to reduce labour
market segmentation and tackle the economic and social conditions which
put women, ethnic minorities, and certain other groups at a disadvantage in
the labour market. Moreover, even in more centralized economies, solidaris-
tic wage bargaining of the Nordic type is only fully effective where it is
accompanied by such measures. In the absence of measures to eliminate
segmentation, a solidaristic wages policy may simply reduce employment
opportunities for the disadvantaged by pricing them out of the market. The
most effective way to equalize wages is through a combination of centralized,
solidaristic bargaining with active labour market policies to reconcile supply
and demand in particular markets. Active labour market policies on their
own, even without solidaristic bargaining, can certainly reduce wage disper-
sion, but they are unlikely to achieve the combination of wage equality and
high employment observed, above all, in Sweden. Even if active labour
21
See Pollan (1980) and also OECD (1985).
market policies were retained in Sweden, it seems likely that the abandon-
ment of solidaristic bargaining would lead to some increase in wage disper-
sion—though not to the levels observed in North America and Japan.
Thus, high wage dispersion normally comes about through a combination
of labour market segmentation (including unequal access to skills) and
micro-flexibility of wages in local markets. In theory, if segmentation could
be eliminated, then market forces would spontaneously generate a fairly
uniform distribution of wages. There would still remain differentials for skills
and hard work, but for the most part these would be comparatively small.
The labour market would, in fact, resemble the Walrasian ideal of perfect
competition. This, at least, is the theoretical possibility. The reality, how-
ever, is rather different. As a general rule, in economies where local market
forces play a major role in wage determination, there are also major obstacles
to labour mobility and the acquisition of skills. Such a combination of
labour market segmentation and relative wage flexibility leads inevitably to
high wage dispersion. Fortunate workers earn high wages and enjoy secure
employment, whilst many others are condemned to a life of low pay and
often insecurity. Sometimes these inequalities are the result of unequally
distributed bargaining power. Sometimes they are the consequence of dis-
crimination, as in the case of women in Japan or blacks in the United States.
And sometimes they are the outcome of a culture of deprivation which
prevents the disadvantaged from acquiring the skills and habits required to
land and hold a good job. Thus, whatever the theoretical possibilities, those
countries where relative wages are determined primarily by market forces
are, in practice, also countries with considerable labour market segmentation
(broadly defined to include lack of skills), and it is this combination which
produces such a high level of wage dispersion.
The above discussion suggests that highly decentralized economies are
likely to exhibit high wage dispersion. Moreover, centralized economies will
also exhibit high wage dispersion if, as seems to be the case in Austria, they
combine segmented labour markets with micro-flexibility of wages in res-
ponse to local conditions. What are the implications of flexibility in relative
wages for total employment? Most neo4iberals take it for granted that
flexibility in this sense leads to more employment. Their basic argument is
that ‘deregulating’ labour markets, and allowing wages to vary in response
to market forces, will lead to the absorption of excess labour into low-paid
jobs, primarily in the service sector. Although at first sight convincing, this
argument contains a potentially serious flaw. It ignores the fact that flexibility
is a two-way process. Regulatory mechanisms, such as solidaristic bargaining,
often operate on both ends of the wage spectrum at once. Whilst keeping
wages up at the bottom end, they may also keep wages down at the top end.
These effects have opposing consequences for employment. Higher wages
at the bottom end tend to reduce employment in otherwise low-paid jobs,
110 BOB ROWTHORN
whilst wage restraint at the top encourages expansion in another direction.

The overall impact on employment depends on the relative strength of these
two effects. If wage restraint at the top end is the dominant feature, an egalit-
arian form of regulation will mean not only lower wage dispersion, but also
more employment. Under these conditions, deregulation and greater flex-
ibility in relative pay will lead to increased wage dispersion, higher average
wages, and less employment. On the other hand, suppose that regulation
functions primarily to pull up wages at the bottom end without exerting
significant restraint at the top. Then deregulation will certainly lead to more
wage dispersion, but it will also increase total employment, since wages will
fall at the bottom end and new, low-paid jobs will be created without a
significant loss elsewhere in the economy.
4.5.1 Empirical evidence

There have been several studies seeking to assess the impact of relative wage
flexibility on employment. In their examination of OECD experience, Klau
and Mittelstadt (1986) found some evidence of a positive relationship be-
tween these two variables, suggesting that the high wage dispersion observed
in countries like Japan and the USA is partially responsible for their good
employment performance. However, as the authors themselves point out,
the evidence they present is weak and their conclusions are tentative. In a
more recent article, Freeman (1988) argues that there is a U-shaped relation-
ship between employment and wage dispersion, which he uses as an in-
dicator of relative wage flexibility. This finding can be criticized on several
grounds. His measure of wage dispersion is based on unweighted observa-
tions and is therefore unreliable. Moreover, in his statistical tests, he does
not make adequate allowance for the effects of bargaining structure on
employment and wages. As we have seen earlier, and Freeman himself points
out, wage dispersion is inversely correlated with the degree of centralization
in bargaining. The U-shaped relationship between employment and wage
dispersion, which he identifies, may therefore be spurious. It may, in fact, be
merely a reflection of another U-shaped relationship—between employment
and centralization—which has already been discussed above. Indeed, this
point was implicitly accepted by Freeman in the discussion following a
conference presentation of his arguments.22
In their paper on the US economy in the 1970s, ‘Does a flexible industry
wage structure increase employment?’, Bell and Freeman (1985) reached
just the opposite conclusion. They found that the upward flexibility of wages
in technically dynamic industries predominated over downward flexibility in
stagnant industries, and the employment-reducing effects of the former
See the discussion following Freeman’s paper (1988).

outweighed the employment-enhancing effects of the latter. Similar results
were obtained in a study of seven EC countries by Dewatripont (1988).
These studies are open to criticism, especially with regard to their theory of
causation. They assume that productivity is the autonomous variable, and
that differential wage growth is a response to differential growth in produc-
tivity. Thus, causality runs from productivity to wages. However, causality
could just as well run in the opposite direction, with differential growth in
productivity being partially a response by employers to autonomous, but
non-uniform, growth in wages. Even so, their conclusions are extremely
interesting. If correct, they imply that the high level of wage dispersion
observed in countries like the USA is a consequence of their particular labour
market institutions, but does not help to explain their high level of employ-
ment. Such countries certainly have an unduly large number of people
employed in low-paid, and often low-productivity, jobs. By the same token,
they have considerably less employment in high-wage, high-productivity jobs
than would have been the case had they developed with a more egalitarian
wage structure. This, at least, is the implication of the two studies just
described.
4.5.2 Regression analysis

To explore this question further, let us consider the regression results
presented in Table 4.7. The underlying equations are the same used pre-
viously for Table 4.3, with the main difference that an additional explanatory
variable, DISP, has been added. This variable is merely our normal measure
of wage dispersion.
Evidence presented earlier indicates that wage dispersion is influenced by
the structure of bargaining. To allow for this, there is an additional equation
for DISP as a function of the variable DECEN Finally, all equations have a
dummy variable for Austria, which is included because the country is clearly
such a special case.
As far as the effects of centralization on employment are concerned, these
regressions confirm our earlier results. The coefficients of DECEN and
DECEN SQ. are much the same as before, with similar levels of significance.
For women, there is some evidence of a U-shaped curve in 1973, but not for
men. In 1985 the curve is well defined for both men and women. The
regressions also indicate the existence of a strong and remarkably stable inverse
relationship between centralization and wage dispersion. The coefficients of
DECEN in the equations for DISP are highly significant and virtually ident-
ical in the two years. Moreover, these equations indicate that Austria had a
level of wage dispersion some 11–12 percentage points greater than expected
given the degree of centralization in its bargaining structure. The final point
to note concerns the coefficients of the variable DISP in the employment
112
TABLE 4.7 Regression analysis: employment rate, centralization, and wage dispersion
Dependent variables Explanatory variables
DECEN DECENSQ DISP AUS FEM R2
1973
ER males 0.11(0.07) 0.01(0.12) –0.22(0.58) 3.8(0.45) 0.04
ER females –3.95(1.54) 0.15(1.10) 0.56(0.90) –16.5(1.20) 0.27
ER pooled –1.92(1.27) 0.08(0.99) 0.17(0.46) –6.3(0.78) 36.4**(12.93) 0.86
DISP 0.79**(3.22) 11.3*(2.22) 0.45
1985
BOB ROWTHORN
ERmales –3.46*(2.19) 0.16*(1.99) 0.48(1.39) –12.9(1.46) 0.37
ER females –8.63**(4.03) 0.38**(3.38) 0.67(1.43) –40.4**(3.37) 0.63
ER pooled –6.04**(4.39) 0.27**(3.77) 0.58*(1.91) –26.7*(3.45) 23.2**(8.93) 0.79
DISP 0.75**(2.89) 12.2*(2.26) 0.41
Notes: For clarity the constant term is not shown in these equations. DISP = percentage coefficient of inter-industry variation in hourly earnings in
INDUSTRY (mining, manufacturing, and construction); observations are weighted by employment; male and female earnings are regarded as
distinct observations. Other variables are as in Table 4.3. For each year, errors in the equations for dispersion are assumed to be independent of the
errors in other equations; this assumption permits estimation by OLS. Absolute f-values are shown in parentheses. For other notes see Table 4.3.
equations. These coefficients are positive and significant at the 10 per cent
level for both men and women in 1985. Moreover, in the pooled regression
for 1985, the coefficient of DISP is significant at the 5 per cent level.23
The coefficient of this variable is also moderately large and of the right
sign for women in 1973, although not statistically significant; for men it is
insignificant and of the wrong sign.
These regression results suggest—though the evidence is not really very
strong—that wage dispersion has acquired some independent influence over
employment in recent years. At any given level of centralization, employment
for both men and women is likely to be higher in countries with relatively
high wage dispersion than in their more egalitarian counterparts. Thus,
inequalities in the wage structure do seem to have encouraged employment
creation since 1973, though not to any great extent previously. However, this
does not appear to be the main reason why highly decentralized economies
have performed relatively well in terms of employment in recent years. If we
control for wage dispersion, their performance still looks fairly good by
international standards. This suggests that, in recent times at least, extreme
decentralization in bargaining has had a beneficial impact on employment
in its own right, over and above any effect resulting from high wage dispersion.
None of this, however, alters the fundamental conclusion of this chapter,
that the best labour market performance of all has been in the Nordic
countries, especially Sweden, which have managed to achieve high levels of
employment without resorting to the kind of inequalities and wage disper-
sion observed in Japan and North America.
4.6 WAGE DIFFERENTIALS AND ECONOMIC DYNAMISM
The solidaristic wages policy in Nordic countries is often considered in

purely social terms as an egalitarian device for reducing income inequality.
However, the original advocates of this policy also stressed its potential
economic virtues as a mechanism for promoting modernization and struc-
tural change. They reasoned as follows. Centralized bargaining would equalize
wages across the economy, reducing differentials to the minimum required
to compensate workers for arduous work or long hours or for any sacrifices
involved in acquiring skills. As the economy developed, this equalization
would put continuous pressure on backward firms to modernize and also
accelerate the contraction of declining sectors. Simultaneously, the gov-
ernment would take active measures to avoid unemployment by encour-
aging the creation of new jobs elsewhere in the economy. The resulting
23
Similar results are obtained when employment is measured in full-time equivalents. Note
that the coefficients of DISP in 1985 are much smaller and statistically insignificant if the
dummy for Austria is excluded.
114 BOB ROWTHORN
combination of high employment and uniform wages would further ac-

celerate structural change by reducing worker opposition to plant closures
and rationalization. With equally well-paid jobs available elsewhere, workers
would have less reason for resisting redundancy than would be the case if the
only alternatives were unemployment or a large reduction in pay. Finally,
the average wage level would be set at a level compatible with this pro-
gramme for continuous restructuring and for the achievement of the trade
union movement’s wider economic and social goals. Thus, the solidaristic
wages policy was to be one arm of a comprehensive policy for modernization
and structural change. The policy would accelerate the transformation of the
economy by increasing the pressure on firms to adapt, whilst simultaneously
reducing worker resistance to such adaptation.
The above argument was most elaborately developed in Sweden and is
sometimes referred to as the Swedish, or Rehn-Meidner, model.24 In this
model, the financial cost of acquiring skills was to be borne mainly by the
government or employers, with little or no contribution by the worker
concerned. Moreover, government intervention would ensure that mobility
was relatively easy, and the costs of moving would again be borne mainly by
the government or employers. As a result, only relatively small wage differen-
tials would be required to induce workers to acquire skills or move between
jobs, and a highly egalitarian wages structure could therefore be achieved
without significant economic costs and with clear benefits in terms of
allocative efficiency, technological dynamism, and structural change.
The impact of solidaristic bargaining on the wages structure in Sweden
has been extraordinary. Fig. 4.4 provides information on wages and labour
productivity in Swedish industry in 1985. In this diagram, firms are ranked
according to the amount of value-added they produce per worker. On the
left are the firms with very high productivity and then, moving to the right,
are firms with progressively lower productivity. The graph also shows the
average wage level in each firm. As can be seen when comparing firms, there
is no relationship whatsoever between average wages and labour produc-
tivity. Some firms pay slightly above the national average and some slightly
below, but the variations are very small. Moreover, the firms with the highest
value-added per worker in Sweden pay just about the same wages as do firms
with the lowest value-added. Under the solidaristic policy, this is just what
we should expect. Moreover, given the assumptions underlying the Swedish
model, such uniformity in wages is efficient, since it implies a refusal to
subsidize backward firms or penalize their more dynamic counterparts. The
situation shown in Fig. 4.4 is just what a perfectly competitive labour market
24
The original version of this model was produced in 1951 by two economists, Gösta Rehn
and Rudolf Meidner, who were employed by the Swedish trade union federation LO. For an
English translation see LO (1953), and for a good survey of subsequent developments in this
area in Sweden see Martin (1984).
FIG. 4.4 Inter-firm productivity and wages in Swedish industry, 1985

Source: Standing (1988).
would produce in a situation where the costs of labour mobility, education,

and training are socialized, as they are in Sweden. Through a combination
of centralized solidaristic wages policy and vigorous government interven-
tion in the labour market, Sweden has effectively mimicked the Walrasian
ideal of perfect competition.
The case of Austria provides an interesting contrast to Sweden. As we have
seen, wage differentials in Austria are extremely high, and there is evidence
that this factor has had a conservative effect on the country’s industrial
structure. For example, the low pay of workers, especially women, in the
textile and clothing industries has discouraged innovation and subsidized
activities which should have been phased out. Conversely, the high rates of
pay in older, nationalized industries, like steel and chemicals, intensified the
opposition of workers to rationalization during the severe structural crises of
the early 1980s. If the Austrian trade union movement had pursued a more
116 BOB ROWTHORN
egalitarian wages policy the pace of structural change would almost certainly
have been faster. This would have created new problems, of course, and a
vigorous labour market policy along Swedish lines would have been required
to ensure the absorption of workers displaced from declining industries. As
it was, Austria had neither an egalitarian wages policy, nor a Swedish-style
labour market policy to facilitate the absorption of displaced industrial
workers. The absence of this combination helps to explain the structural
problems which have become visible in Austria during the 1980s.25
4.6.1 Some criticisms

In recent years the Swedish model has been criticized for ignoring certain
important aspects of worker motivation.26 The model tacitly assumes that
labour productivity is determined primarily by the skill of workers, the
quality of management, and the quantity and quality of capital stock, and
the kind of output produced. All except the first are quite outside of the
worker’s control and require no wage incentive to modify them. Skill, of
course, is an attribute of the worker and some financial inducement may be
required to encourage workers to acquire the requisite skills. However, since
the cost of acquiring skills is largely socialized in this model, the financial
inducement needed for workers to become skilled is very small. Thus, within
the model, there is simply no reason for large differentials in pay or for
significant local wage flexibility.
In this story, it is taken for granted that workers will perform their jobs in
a broadly similar fashion, and that only minimal financial incentives will be
required to induce this performance. Such an assumption, the critics admit,
might have been justified in the past, when jobs were fairly standardized and
easily monitored, so there was no problem in ensuring a uniform level of
performance. Nowadays, they claim, the situation is different. Production
methods have changed and a new relationship between workers, their job,
and their employers is required.27 Under the old methods, sometimes
described as Fordist, jobs were standardized and workers were closely
monitored, enjoying little discretion in their work. As a result, financial
incentives played only a minor role in their performance. Under the new
production methods, however, which are sometimes described as post-
Fordist, work is less standardized, monitoring is often impractical, and
workers enjoy considerable discretion. Moreover, there is merging of
25
The case of Austria is discussed at greater length in the essay by Alois Guger in the present
volume.
26
For a comprehensive attack on solidaristic bargaining see Jonsson and Siven (1986).
27
For a general theoretical statement in support of this view, inspired by the Japanese
experience, see Aoki (1988). Further support can be found in the ‘efficiency wage’ literature,
of which some of the most important articles are reprinted in Akerlof and Yellen (1986).
functions, so the dividing line between workers and managers is increasingly
blurred, and workers now play a far more active role in shaping the destiny
of their firms than was formerly the case. As a result of these changes, it is
claimed, the performance of firms now depends crucially on the active
involvement of workers in their jobs, and to induce this kind of involvement
a different kind of wages system is required. Genuine career structures must
be established within firms and the pay of workers must be linked to the
performance of their own firm, or some aspect thereof. By their nature, such
firm-specific incentives conflict with the basic principle of solidaristic bar-
gaining, which seeks to establish a uniform rate for the job across the whole
economy, independently of the situation in particular firms.
How important is this problem in practice? There is obviously something
in it, for otherwise it is difficult to explain the proliferation of profit-sharing
and similar firm-specific incentive schemes in recent years in Sweden (and
also many other countries). The proportion of metal workers in Sweden paid
‘according to results’ has increased from 5 per cent in 1980 to almost 50 per
cent in 1986. Over the same period the proportion involved in profit-sharing
schemes has risen from almost zero to more than 20 per cent.28 An avowed
purpose of the new payment systems is to motivate workers by providing
incentives which are supposedly lacking under the solidaristic wages policy.29
From the individual employer’s point of view, these schemes have an obvious
attraction. However, for the employers as a whole, they pose an equally
obvious danger. If they get out of hand, such schemes make centralized
bargaining impossible and may lead to an inflationary competition between
different employers, thereby undermining macro-economic stability and
growth. What is rational for the individual firm may be counter-productive
for the collectivity of firms taken as a whole.30
This point is recognized in a recent book by a tripartite group of Swedish
economists, one from each of the two major trade union federations and one
from the central employers’ confederation (Faxen et al. 1988). After describ-
ing the dangers of fragmentation, they explicitly call for a strengthening of
the centralized bargaining system, which has been seriously eroded in
28
These figures are derived from information contained in Standing (1988).
29
In Sweden, as in many other countries, there is also something of a ‘hidden agenda’ behind
these schemes. Their purpose is not always narrowly economic, in the sense of providing a
financial incentive to motivate workers to work. They are often part of a more general strategy
for weakening trade unions by undermining collective solidarity either within the enterprise itself
or between workers from different firms. It is hardly coincidental that employers in Sweden have
become so keen on profit-sharing schemes and the like just at a time when the unions are
acquiring new collective powers of control over private enterprise through the recently estab-
lished wage earners’ funds. These powers are not yet very extensive, but the employers are
concerned about the longer-term threat which they represent (see Pontusson 1987).
30
See Ahlen (1989) for a good analysis of the forces undermining centralized bargaining in
Sweden in recent years. These forces include inter-union rivalry and the growing importance
of public sector workers in wage negotiations.
118 BOB ROWTHORN
Sweden during the 1980s. It is interesting to note that their call has been
heeded, and in 1989, for the first time in six years, there were comprehensive,
centralized wage negotiations in Sweden covering most of the private sector.
However, the powerful metal workers’ union has remained outside these
negotiations, so there is still a serious question mark over the future of
centralized bargaining in Sweden.
Critics of the solidaristic wages policy have complained that it discourages
labour mobility and the acquisition of skills. These criticisms have been
especially strident in Sweden, where the process is allegedly most advanced
(Jonsson and Siven 1986). In fact, the evidence for Sweden provides little
support for these claims. By international standards, Swedish workers are
highly skilled and their financial incentives to acquire skills are adequate, if
not unduly generous (Flanagan 1987). On the question of labour mobility,
which has attracted considerable attention abroad, the picture is clear. It was
originally claimed by the critics that low wage differentials were discouraging
mobility by reducing the incentive for workers to move voluntarily between
jobs. The evidence in support of this contention was a noticeable decline in
mobility between Swedish firms during the 1970s. However, this was only
part of the story. Alongside the decline in ‘external’ mobility between firms,
there was a parallel rise in ‘internal’ mobility within firms. When the latter
is taken into account, the decline in total mobility in Sweden in the 1970s—
internal and external combined—was quite small. Moreover, even the
decline in external mobility during the 1970s cannot be ascribed to the
solidaristic wage policy. It was a perfectly normal response to the economic
recession and slack labour markets afflicting Sweden, like other countries,
at the time. Workers did not leave their jobs voluntarily to go to other jobs,
because there were not that many jobs to go to. When the Swedish economy
picked up in the 1980s, external mobility rose sharply and internal mobility
continued to rise. As a result, total mobility is now back to the level it was in
the 1960s.31
Thus, of the various arguments against the solidaristic wages policy, those
concerning skills and mobility are unconvincing. However, there does seem
to be something in the motivational argument, and the critics of solidaristic
bargaining are expressing a concern about incentives which is nowadays
common in most advanced economies. On the other hand, the social and
macro-economic advantages of centralized bargaining and the solidaristic
wages policy are considerable. A major challenge in the coming years will be
to see how this system can be made more responsive to motivational
requirements without destroying its essential character.
The problem of motivation is not, of course, simply a question of money
or material reward, but is often concerned with the nature of work itself. The
For evidence on this point see Faxen et al. (1988).

absenteeism, excessive labour turnover, and lack of commitment which
employers complain about frequently have their origin in some discontent
about work, such as lack of stimulation or authoritarian work relationships.
Workers in advanced countries are increasingly unwilling to accept these
conditions, and it is unrealistic to believe that merely rejigging the system of
financial rewards will in itself be sufficient to resolve the current difficulties.
The subject of work and its future is obviously of great importance, but we
shall not consider it here.
4.7 LABOUR MARKET SOLIDARITY
To conclude, let us locate the questions we have been discussing against a

wider background. Throughout this chapter die discussion of wages and
employment has been informed by a more general concept, which, for want
of a better name, I shall call ‘labour market solidarity’. This concept
embodies the idea that the costs and benefits of economic performance
should be distributed widely throughout the population. In particular, the
costs arising from recession or structural change should be shared equally
and not off-loaded on to a weak or vulnerable section of the work-force.
Thus, labour market solidarity is concerned with citizenship and entitle-
ments, with who should share in the benefits of economic performance and
who should bear the costs. The concept has a number of dimensions, of
which the following are of particular importance: employment, wage disper-
sion, unemployment benefits, and other welfare entitlements. The record of
the corporatist economies in these areas can be summarized as follows.
4.7.1 Employmen t
Norway and Sweden have a strong commitment to full employment for
both men and women, although in the case of Norway this commitment has
come under severe strain in the late 1980s following a collapse in the price
of North Sea oil. The commitment to full employment is somewhat weaker
in Denmark and Finland, though both have performed quite well by inter-
national standards. Finally, Austria’s record in this area has been rather poor,
especially for women, for whom employment opportunities are severely
limited.
4.7.2 Unemployment benefits

The financial cost of being unemployed varies enormously internationally.
Most corporatist countries provide relatively generous benefits to those
officially classified as unemployed. As a result, even where there is a great
120 BOB ROWTHORN
deal of open unemployment, as in Denmark, its financial impact is frequently

cushioned. Sometimes, as in Sweden, extremely high benefits are accompanied
by considerable pressure on the unemployed to find a job or enrol in a
government training scheme or work programme. In other countries, such
as Denmark, the attitude is more relaxed. Some evidence on the generosity
of unemployment benefits is provided in Table 4.8, which shows how these
benefits compare to the average pre-tax wage.32
4.7.3 Wage dispersion

As documented previously, wage dispersion in the Nordic countries is
relatively low by international standards. They have all avoided the emer-
gence of a large underclass of ‘working poor’, which is such a prominent
feature of the American labour market. In Austria wage dispersion is very
high and similar to that observed in the USA. This situation, it must be said,
is partly mitigated by the welfare state, which provides even low-paid workers
in Austria with a fairly high ‘social wage’ in the form of health services,
pensions, and the like.
4.7.4 Other welfare entitlements

The role of the welfare state as an equalizer can be gauged from Table 4.8,
which presents some indices of ‘decommodification of labour power’, as
calculated by Esping-Andersen (1989). These indices measure the extent to
which the entitlement of workers to pensions and sickness benefit is inde-
pendent of their monetary income from employment. Of all the countries
examined by Esping-Andersen, Austria and the Nordic countries exhibit the
greatest independence, according to his index, of pensions and sickness benefits
from employment income. Such unconditional entitlement to welfare
benefits is a major factor helping to mitigate inequalities arising from
unemployment or low wages.
The above is by no means a complete list of what is covered by the term
labour market solidarity. One should also include items like job security,
employee rights at work, holidays, parental leave for child-care, and the like.
All of these influence the extent to which various groups share in the fruits
of economic growth and are shielded from discrimination during periods of
recession or structural change. In the Nordic countries progress on these
fronts has been impressive even during the troubled years since 1973.
Progress in Austria has been less impressive.
32
Note that the figures shown in Table 4.8 provide a composite measure reflecting the level,
coverage, and duration of unemployment benefits. Thus, the low figure for the USA is due
mainly to the limited coverage and duration of benefits; for those actually receiving benefits in
the USA the level is quite high. For further evidence on unemployment benefits see Burda
(1988), Burtless (1987), and OECD (1988*).
TABLE 4.8 Labour market performance and welfare benefits in the 1980s in 17 OECD countries
Unemployment Employment Wage Pensionand Unemployment
rate (%) rate(%) dispersion sickness benefits benefits (%)
(U) (ER) (DISP) (PENSICK) UBEN UBEN*
ustria 5 63 21 24 34 40
Norway 2 75 13 29 35 35
Sweden 2 80 8 32 63 72
Denmark 8 74 9 30 58 86
Finland 5 73 16 24 34 57
Other Europe 9 59 13 20 32 37
Japan 3 71 29 17 27 27
USA 6 68 22 7 15 15
Other OECD 7 64 20 12 33 33
Notes: U, UBEN, and UBEN* refer to 1987, ER and DISP to 1985, and PENSICK to 1980. ER, DISP = as in Table 4.7; PENSICK = index
indicating how far pensions and sickness benefits are delinked from employment income (source: Esping-Andersen 1989); UBEN = average
unemployment benefit per worker officially classified as unemployment (including those not receiving benefits) as a proportion of average pre-tax
wages (source: OECD 1988a); UBEN* = as UBEN but including special government assistance for early retirement; Other Europe = Germany,
The Netherlands, Belgium, France, the United Kingdom, Italy, and Switzerland; Other OECD = Australia, New Zealand, and Canada.
122 BOB ROWTHORN
When these various aspects of labour market solidarity are taken into
account—employment, wage dispersion, welfare benefits, etc.—the record of
the Nordic countries is outstandingly good by international standards,
although not without its blemishes.33 In all of these countries the costs and
benefits of good and bad economic performance have been widely shared
throughout the population, either in the form of wage income or welfare
benefits of some kind. In none of them has there emerged a large underclass
of poor people existing on low wages or miserly benefits. They have all
managed to avoid the creation of a ‘two-thirds/one-third society’, of the type
found in much of North America and Western Europe (i.e. two-thirds doing
well, one-third in poverty). The record of Austria in this respect is mixed.
Wage dispersion is high and employment opportunities for women are
limited, but this situation is partially mitigated by the existence of a welfare
state which is reasonably generous by international standards.
4.7.5 Correlations
It is interesting to ask at this point whether there exists any systematic
relationship between the various indicators of labour market solidarity just
described? Table 4.9 shows the correlation coefficients between some of
these indicators. It also shows how these indicators are correlated with the
measure of decentralization, DECEN. In this table, the figures above the
diagonal are computed for all seventeen countries in the sample, whilst for
those below the diagonal Italy is excluded. A justification for excluding Italy
is that the country is an extreme outlier in a number of dimensions, and
besides some of the Italian data are either suspect or misleading.34
The situation revealed by Table 4.9 can be summarized as follows. There
is no correlation between DECEN and the employment variable, ER. This
is to be expected since, as we have seen above, there is a U-shaped relation-
ship between these variables, which is unlikely to be picked up in a simple
correlation. There is, however, a clear negative correlation between DECEN
and the variables PENSICK and UBEN*, indicating that welfare entitle-
ments of all kinds tend to be much lower in decentralized economies. This
is probably not a result of decentralization as such, but of the fact that Labour
is politically weak in decentralized economies. On this and related points see
the interesting paper by Paloheimo (1988). There is also a clear positive
correlation between DECEN and DISP, indicating that wage dispersion is
normally high in decentralized economies. Finally, there is a negative
33
For example, there is still a fair amount of poverty in the Nordic countries amongst single
mothers who are either unemployed or unable to work full time (see Allen 1989). Such poverty
is least in Finland where even single mothers commonly have full-time jobs.
34
In the case of Italy, the variables DECEN, UBEN, UBEN*, and DISP may be misleading
or of questionable accuracy. (For DECEN see n.9.)
TABLE 4.9 Labour market performance and welfare benefits in

17 OECD countries
Correlation matrix for the 1980s
DECEN ER DISP PENSICK UBEN*

DECEN 1.00 –0.22 0.44 –0.73 –0.61
ER –0.15 1.00 0.15 0.34 0.41
DISP 0.57 0.01 1.00 –0.57 –0.45
PENSICK –0.74 0.35 –0.62 1.00 0.73
UBEN* –0.58 0.29 –0.70 0.78 1.00
Notes: Correlations above the diagonal cover all 17 countries; those below the diagonal
exclude Italy. DECEN = Index of decentralization in wage bargaining (= Calmfors–Driffill
rank); other variables as in Table 4.8.
correlation between DISP and the variables PENSICK and UBEN*, in-
dicating that high wage dispersion is associated with low welfare entitle-
ments. This negative correlation is especially strong when Italy is excluded
from the sample.
Since we know from previous evidence that employment levels are rel-
atively high in decentralized economies, we can tell the following rather
convincing story. In such economies, labour tends to be politically weak,
welfare entitlements of all kinds are typically low, and as a result many people
are forced by economic pressures to accept poorly paid jobs. This helps to
explain why decentralized economies normally exhibit both a high level of
employment and high wage dispersion.
The situation is quite the opposite in the centralized Nordic countries,
where the labour movement is strong and welfare entitlements of all kinds,
including unemployment benefits, are generous. The generosity of these
entitlements helps to squeeze wage differentials by reducing the pressure on
workers to accept low-paid employment, thereby reinforcing the equalizing
effects of solidaristic bargaining by the trade unions. Finally, since the unions
are centralized, they consciously set average wages low enough to permit the
simultaneous achievement of a high level of employment and an egalitarian
wage structure. This is, of course, an idealized picture of the situation in the
Nordic countries, but it does contain a significant element of truth. It shows
how the economic and political strength of organized Labour can function
in a complementary fashion to produce mutually reinforcing, and egalitarian,
outcomes in the labour market and the welfare state. Conversely, the
example of the United States, and certain other decentralized economies,
illustrates how the economic and political weakness of organized labour may
conspire to produce just the opposite result of mutally reinforcing, but
inegalitarian, outcomes in both arenas.
124 BOB ROWTHORN
4.7.6 Austria
Austria provides an interesting special case. The labour movement in Austria
has considerable economic and political strength, yet wage dispersion is
extremely high and there is a large substratum of low-paid workers. These
inequalities reflect the balance of power within the Austrian labour move-
ment and its resulting priorities and objectives. Unlike its Nordic counter-
parts, the labour movement in Austria has not pursued a solidaristic wage
policy designed to reduce differentials. On the contrary, it has accepted,
indeed maintained, an inegalitarian wages structure which is partly inherited
from the past, and is partly an outcome of competitive forces currently
operating in the country’s segmented labour market.35 This situation is
especially harmful to women, who have only limited access to more highly
paid jobs and, if they work at all, are crowded into the bottom end of the
labour market.
Such inequalities are reinforced by the unemployment benefit system,
whose character also reflects the priorities of the Austrian labour movement.
As can be seen from Table 4.8, this system is on average quite generous.
However, the system discriminates heavily against low-paid workers, for
whom the replacement rate (unemployment benefit as a percentage of net
earnings) is often considerably less than for workers in the middle range of
earnings. Benefits are especially meagre for low-paid workers without
children, and when such workers become unemployed, they are under great
pressure to accept whatever job is available, no matter how badly paid it may
be (see Table 4.A6). This makes it difficult for the low-paid to escape from
their situation, and helps to keep down wages in certain traditionally low-
paid occupations. Since women earn much less than men on average, the
benefit system is thus highly discriminatory in its impact.
Another factor helping to depress wages in Austria is the large reserve of
women who are potentially available for employment, but are not officially
classified as unemployed and receive no unemployment benefit. The official
unemployment rate for women in Austria during the 1980s has never been
more than a few per cent, yet the statistics indicate that approximately 50
per cent of women of working age are without paid employment of any kind
(see Appendix Tables 4.A4 and 4.A5). The existence of such a huge reserve
of potential workers, who receive no unemployment benefit, must inevitably
exert a downward pressure on the earnings of women, and thereby on the
earnings of those male workers who compete with them.
Thus, the high wage dispersion in Austria, and the low pay of women and
certain male workers, are the result of two factors, both of which reflect the
35
On this point see the chapter by Alois Guger in the present volume. Very recent
developments suggest that this situation is beginning to change. There is increasing awareness
both inside and outside the trade unions of the extent of inequality in Austria and of the need
to do something about it.
balance of power within the Austrian labour movement and its resulting
priorities. At an economic level, this movement has pursued a bargaining
strategy which favours certain historically powerful groups of male workers,
whilst at a political level it has helped to shape an unemployment benefit
system which favours such workers, and reinforces their advantages in the
labour market at the expense of women and other male workers.
4.7.7 Two types of corporatism?

How do we explain the striking contrast between Austria and the Nordic
countries? They are all ‘corporatist’ economies, with strong and centralized
labour market institutions, yet the priorities and objectives of their labour
movements seem very different. One clue is provided by Göran Therborn in
his article in the present volume, where he comments on the different forms
of ‘industrial order’ in capitalist countries. He draws a distinction between
systems organized around the principle of conflict and those organized
around the principle of consensus. For historical reasons, he argues, the
industrial order in the Nordic countries is organized around the institu-
tionalization of conflict, whereas in Austria the dominant principle is con-
sensus. Although he does not explore this issue, the institutionalization of
conflict is associated with, indeed is mainly the result of, a historical struggle
by labour for democracy and equality. This is certainly true in the Nordic
countries. The result is that the labour movement in these countries has had
an inclusive and egalitarian thrust, which is reflected in commitment to wage
equality and genuinely full employment, for women as well as men. In
Austria, on the other hand, for reasons which Therborn mentions briefly,
the accommodation between labour and capital has been of a far more
conservative kind, establishing consensus by freezing the status quo of the
immediate post-war period. This may help to explain why women receive
such a low priority in wages and employment, and why earnings differentials
in general are so high in Austria. The conclusion seems to be that industrial
orders in which conflict is institutionalized have an inherent dynamic
towards equality, whereas those based on consensus are more conservative
and tend to preserve the power relationships and inequalities in existence at
the time these orders were established. Both kinds of order may be charac-
terized by centralized institutions, both may exhibit certain features in
common which derive from these centralized institutions, and both may be
reasonably described as ‘corporatist’, but in other respects they may be very
different. This may serve as a warning against trying to squeeze too much
out of concepts like corporatism or labour market centralization.
126 BOB ROWTHORN
APPENDIX 4.A.
SUPPLEMENTARY TABLES
TABLE 4.A1 Calmfors–Driffill Index of centralization in wage bargaining

Rank Co-ordination level Existence of parallel Total
within central central organizations score
organizations and their co-operation
1 Austria 3 3 6
2 Norway 3 2 5
3 Sweden 3 2 5
4 Denmark 3– 2 5–
5 Finland 3– 2 5–
6 Germany 2– 3– 5–
7 The Netherlands 2 2+ 4+
8 Belgium 2 2 4
9 New Zealand 1 3 4
10 Australia 1 3 4
11 France 1+ 2 3+
12 UK 0+ 3 3+
13 Italy 1+ 2 3+
14 Japan 1 2 3
15 Switzerland 1 2 3
16 USA 1 1 2
17 Canada 1 1 2
Source: Calmfors and Driffill (1988).
TABLE 4.A2 Labour market performance 1985 (males)
Country Unemploy- Employ- Employ- Wage INDEED

ment rate ment rate ment rate disper- Indexb
(FT equi- sion3
valent)
Austria 4.6 77.7 77.1 12.5 64.6
Norway 2.1 84.2 79.3 11.3 68.0
Sweden 2.7 83.4 80.4 6.6 73.7
Denmark 6.6 79.9 77.3 6.0 71.3
Finland 5.4 75.1 73.4 10.2 63.2
Germany 7.8 71.5 70.9 6.2 64.7
The Netherlands 13.4 65.9 63.6 7.8 55.8
Belgium 9.4 66.6 65.7 11.7 54.0
New Zealand 2.3 80.2 78.1 14.4 63.7
Australia 7.9 77.8 75.4 14.3 61.1
TABLE 4.A2 (cont.)

(FT equi- sion3
valent)
France 8.6 67.2 66.3 9.0 57.3
UK 13.6 74.9 73.7 11.4 62.3
Italy 6.8 71.7 70.8 5.5 65.4
Japan 2.6 85.5 83.4 12.0 71.5
Switzerland 0.8 88.6 85.6 9.7 75.9
USA 7.0 76.3 73.4 15.1 58.3
Canada 10.3 75.4 72.5 12.8 59.7
a
Employment weighted and refers to INDUSTRY (manufacturing, mining, and
construction).
b
Employment rate (FT equivalent) - Wage dispersion.
Note: All statistics are measured in percentages.
TABLE 4.A3 Change in labour market performance 1973–1985 (males)

(FT equi- sion3
valent)
Austria 4.0 –7.7 –7.7 1.9 –9.6
Norway 1.1 0.3 –1.0 0.9 –1.9
Sweden 0.4 –2.7 –4.2 –2.3 –1.9
Denmark 5.9 –3.1 –4.9 –1.9 –3.1
Finland 3.2 –8.5 –9.6 –2.0 –7.6
Germany 6.9 –14.2 –14.4 0.8 –15.2
The Netherlands 10.9 –15.0 –16.8 2.1 –19.0
Belgium 7.5 –13.4 –14.1 –2.4 –11.7
New Zealand 2.0 –7.5 –7.5 3.5 –11.0
Australia 6.1 –10.4 –11.3 7.2 –18.5
France 7.0 –13.3 –13.6 0.2 –13.8
UK 10.6 –13.3 –13.7 3.2 –16.9
Italy 2.5 –6.4 –6.4 –5.4 –1.0
Japan 1.3 –3.4 –3.3 0.1 –3.4
Switzerland 0.8 –12.0 –11.9 3.5 –15.4
USA 2.8 –3.0 –3.0 –4.4 1.4
Canada 5.4 –5.4 –6.2 –3.6 –2.6
Note: The figures shown here are absolute changes in the magnitudes given in Table 4.A2.
128 BOB ROWTHORN
TABLE 4.A4 Labour market performance 1985 (females)
Country Unem- Employ- Employ- Wage F/M INDEED

ployment ment rate ment rate disper- ratiob Indexc
rate (FT equi- siona
valent)
Austria 3.6 48.8 44.0 15.9 70.2 19.7

Norway 3.0 66.3 48.1 11.6 81.6 29.8
Sweden 2.9 75.9 58.4 7.1 87.2 44.7
Denmark 8.2 68.4 53.1 6.5 84.7 39.5
Finland 4.4 70.3 65.9 10.6 75.8 41.9
Germany 9.5 45.6 38.8 8.1 73.1 22.4
The Netherlands 12.2 36.2 27.1 5.3 76.7 16.7
Belgium 16.5 42.1 38.0 9.0 77.0 22.3
New Zealand 7.5 44.1 37.9 9.5 72.5 20.5
Australia 8.7 49.9 40.9 9.3 75.4 23.9
France 12.8 47.9 43.1 7.4 81.5 29.1
UK 8.8 54.7 43.1 12.3 73.0 22.5
Italy 16.2 34.2 32.6 6.5 85.6 22.4
Japan 2.7 55.7 49.8 14.9 48.9 17.1
Switzerland 1.1 52.7 44.8 11.3 70.3 23.6
USA 7.4 58.9 52.0 17.5 68.0 23.5
Canada 10.7 55.6 48.3 21.0 63.7 17.4
a
construction).
b
F/M ratio = ratio of average female to average male hourly earnings.
c
(Employment rate (FT equivalent) - Wage dispersion) x F/M ratio.
Note: All statistics are measured in percentages.
TABLE 4.A5 Change in labour market performance 1973–1985 (females)

Country Unem- Employ- Employ- Wage F/M INDEED
ployment ment rate ment rate disper- ratiob Indexc
rate (FT equi- sion3
valent)
Austria 2.0 –2.3 –3.1 0.3 2.7 –1.6
Norway 0.6 17.0 10.6 2.9 7.8 8.5
Sweden 0.1 15.0 9.3 –0.1 7.6 11.4
Denmark 7.1 3.6 1.4 –2.6 5.3 5.6
Finland 2.2 8.0 5.7 2.4 8.2 6.8
Germany 8.3 –3.3 –5.3 1.0 1.4 –4.0
The Netherlands 10.4 7.5 0.6 1.1 –0.8 –0.6
Belgium 13.2 1.0 –1.5 –2.9 11.2 4.2
New Zealand 7.2 5.0 3.1 3.8 4.0 0.7
Australia 5.4 3.8 1.1 3.0 7.6 1.1
France 8.3 0.3 –1.8 1.8 3.2 –1.7
UK 7.9 2.1 0.6 3.7 10.1 1.2
Italy 4.8 4.2 3.9 –7.1 11.8 11.2
Japan 1.5 2.2 1.0 0.6 –3.9 –1.2
Switzerland 1.1 –1.4 –9.3 6.0 3.1 –9.2
USA 1.4 10.9 10.0 3.9 6.6 6.0
Canada 4.0 11.5 8.7 2.2 6.3 5.4
a
construction).
b
F/M ratio = ratio of average female to average male hourly earnings.
c
{Employment rate (FT equivalent) - Wage dispersion} x F/M ratio.
Note: The figures shown here are absolute changes in the magnitudes given in Table 4.A4.
TABLE 4.A6 Replacement rates in Austria in 1983: Some illustrative calculations

No children Two children
I II III I II III
1. Gross income 18.9 11.6 6.6 18.9 11.6 6.6
2. Net income 13.4 8.9 5.6 15.6 11.1 7.8
3. Unemployment benefits etc. 9.4 6.0 2.7 13.1 9.7 6.3
4. Replacement rate % 70.5 67.5 47.3 83.9 87.3 80.5
Note: The figures refer to monthly income in thousands of schillings for someone who is
unemployed for three months; the benefits shown include tax refunds; both net income and
benefits include all transfer payments. Line (4) = 100 x line (3)/line (2).
Source: G. Fischer and M. Wagner (cited in Walterskirchen 1987: 227).
130 BOB ROWTHORN
REFERENCES
AHLEN, K. (1989), ‘Swedish collective bargaining under pressure: inter-union rivalry

and incomes policy’, British Journal of Industrial Relations, Nov.: 330–46.
AKERLOF, G. A., and YELLEN, J. L. (1986), Efficiency Wage Models of the Labour
Market, Cambridge: Cambridge University Press.
ALLEN, T. (1989), ‘Feminization of poverty and its economic determinants’, paper
presented at the International Economic Association 9th World Congress, Athens.
AOKI, M. (1988), Information, Incentives and Bargaining in the Japanese Economy,
London: Cambridge University Press.
BELL, L., and FREEMAN, R. B. (1985), ‘Does a flexible industry wage structure
increase employment? The US experience’, NBER Working Paper No. 1604.
BRAUN, A. R. (1986), Wage Determination and Incomes Policy in Open Economies,
Washington, DC: International Monetary Fund.
Basil Blackwell.
BURDA, M. (1988), ‘Wait unemployment in Europe’, Economic Policy, 7: 393–425.
BURTLESS, G. (1987), ‘Jobless pay and high European unemployment’, in R.
Lawrence and C. Schultze (eds.), Barriers to European Economic Growth,
Washington, DC: Brookings Institution.
DEWATRIPONT, C. (1988), ‘Industry wage flexibility and employment in the EEC,
Cahiers economiques de Bruxelles, 120: 1–27.
ESPING-ANDERSEN, G. (1989), The Three Worlds of Welfare Capitalism, Cambridge:
Polity Press.
FAXEN, K. O., ODHNER, C.–E., and SPANT, R. (1988), ‘Wage formation and the
economy in the nineties’ (mimeo).
FLANAGAN, R. J. (1987), ‘Efficiency and equality in Swedish labour markets’, in B.
P. Bosworth and A. M. Rivlin (eds.), The Swedish Economy, Washington, DC:
Brookings Institution.
FREEMAN, R. (1988), ‘Labour market institutions and economic performance’,
HARRISON, B., and BLUESTONE, B. (1988), The Great U-Turn: Corporate Restructuring
and the Polarisation of America, New York: Basic Books.
----------- (1989), ‘Wage polarisation in the US and the “flexibility” debate’ (mimeo).
JONSSON, L., and SlVEN, C.-H. (1986), Why Wage Differentials?, Stockholm: Swedish
Employers’ Confederation.
KATZENSTEIN, P. J. (1984), Corporatism and Change: Austria, Switzerland and the
Politics of Industry, Ithaca, NY: Cornell University Press.
KLAU, F., and MITTELSTADT, A. (1986), ‘Labour market flexibility’, OECD
Economic Studies, 6 (Paris): 8–45.
LO (1953), Trade Unions and Full Employment, Stockholm: LO.
MARTIN, A. (1984), ‘Trade unions in Sweden, strategic responses to change and
crisis’, in P. Gourevitch et al. (eds.), Unions and Economic Crisis, London: Allen &
Unwin.
OECD (1985), ‘Relative wages, industrial structure and employment performance’,
Employment Outlook, Sept. (Paris): 83–98.
----- (1987), ‘Occupational differentials in earnings and labour demand’, Employ-
ment Outlook, Sept. (Paris): 67–96.
----- (1988a), ‘Profiles of labour market budgets 1985–87’, Employment Outlook,
Sept. (Paris): 84–91.
----- (19886), ‘Description of unemployment benefit systems in OECD countries’,
Employment Outlook, Sept. (Paris): 115–28.
PALOHEIMO, H. (1988), ‘The effect of trade unions and governments on economic
performance’ (mimeo), paper prepared for a workshop on ‘Coping with the crisis:
lessons from the corporatist experiments’, Helsinki: WIDER Institute.
POLLAN, W. (1980), ‘Wage rigidity and the structure of the Austrian manufacturing
industry: an econometric analysis of relative wages’, Weltwirtschaftliches Archiv,
116:697–728.
PONTUSSONj. (1987), ‘The story of wage-earner funds’, New Left Review, 165: 5–33.
PRYOR, F. L. (1988), ‘Corporatism as an economic system: a review essay’, Journal
of Comparative Economics, 12: 317–44.
ROWTHORN, R. E. (1990), ‘Wage dispersion and employment: theories and
evidence’, Department of Applied Economics Working Paper, University of
Cambridge.
------and GLYN, A. (1980), ‘The diversity of unemployment experience since 1973’
in S. Marglin and J. B. Schor (eds.), The Golden Age of Capitalism, Oxford: Oxford
University Press.
STANDING, G. (1988), Labour Market Flexibility in Sweden, Geneva: ILO.
THERBORN, G. (1986), Why Some People Are More Unemployed Than Others, London:
Verso.
WALTERSKIRCHEN, E. (1987),’Verteilungswirkungen derArbeitslosenversicherung’.
in A. Guger (ed.), Umverteilung durch ojfentliche Haushalte in Osterreich, Vienna:
WIFO.
WILLIAMSON, P. J. (1989), Corporatism in Perspective, London: Sage.
5
Corporatism, Patterns of Employment, and
Access to Consumption
Andrew Glyn
5.1 INTRODUCTION
A striking feature of the period of slower growth since 1973 is the diversity
of employment performance between OECD countries. As noted in earlier
studies (e.g. Bruno and Sachs 1985) there is little relationship between the
degree to which output slowed down and the rise in unemployment. Many
factors connected with the growth of the labour force have played a role (see
Rowthorn and Glyn 1990), but there still remain wide differences in the
extent to which the slow-down in output growth led to slower growth of
employment.1 This has important, but little explored, implications for the
fundamental question of who bears the costs of slower growth.
A deterioration in employment performance unloads a disproportionate
share of the costs on those who fail to get jobs. Conversely maintaining em-
ployment growth spreads the costs of slow output growth by extending more
widely the access to consumption provided by paid work. Such ‘employment
spreading’ may be achieved through conscious policies to subsidize work in
the market sector. But even if the volume of work required in the market sec-
tor declines, employment may be maintained through the expansion of jobs
in the state sector (providing welfare services etc.) or through cuts in the
working week. So employment spreading includes policies both to maintain
work in the market sector and to increase work in the state sector. Of course
the extent of the output slow-down in response to oil shocks, slower growth
of world trade, higher real interest rates, and so forth is not independent of
the policies pursued. But to analyse the diversity of employment spreading
we take output growth as given.
Employment spreading widens access to consumption, but within the con-
fines of slower output growth such spreading of consumption must be paid
I would like especially to thank Bob Rowthorn for his help and Wendy Carlin and Mustaq
Khan for comments on earlier drafts.
1
The extent to which output slowed down in a country in die period 1973–86 (as compared
to 1965–73) is not significantly correlated with the extent to which the growth of employment
slowed down, in the sample of 15 OECD countries analysed in this chapter. Even if the extreme
case of Japan is left out the correlation is barely significant for the period 1973–9, and less than
half of the variance of the change in employment growth is accounted for by the change in output
growth for the whole period 1973–86.
EMPLOYMENT AND ACCESS TO CONSUMPTION 133
for. In the short run at least consumption out of property incomes may be
squeezed through taxation, extra resources may be borrowed from abroad,
or investment may be cut. But these sources have fairly narrow limits and/or
adverse longer-term consequences. Ultimately the costs of employment spread-
ing must be borne by the mass of wage and salary earners.
Some aspects of employment spreading bring direct benefits to those who
would have kept their jobs anyway, in the form of better welfare services,
more leisure as working hours fall, better possibilities for a second family
member to find a job, for example. But for the most part the policies—of the
state and of the collective organizations of labour—required to implement
or initiate employment spreading have a strong egalitarian thrust. This is
because the benefits are received by broader sections of the population than
contribute to their financing. The implementation of such policies demands,
therefore, a high degree of social solidarity on the part of those with secure jobs.
But there is a more fundamental problem, that of free riding. It is quite
possible for such policies of employment spreading to be advocated, but
when it comes to wage bargaining their costs may implicitly be shrugged off
(compare Calmfors and Driffill 1988). Suppose compensation is achieved
in wage claims for tax increases used to finance an expansion of the welfare
services. This tends to nullify the increase in state employment through
forcing faster rationalization and employment shedding in the market sector.
Indeed the situation may be even worse to the extent that the higher real cost
of labour reduces market sector output (as marginal exporters close, for
example). Cuts in working time provoke similar effects if hourly earnings are
increased to maintain weekly wages (see also Dreze 1986). So in addition to
social solidarity the labour movement has to show strong collective discipline
if employment spreading is to be successful.
The benefit system is another route through which consumption spread-
ing occurs. The extent to which unemployment concentrates the costs of
slower growth on the unemployed depends on the degree to which the benefit
system ‘replaces’ the income lost, and the comprehensiveness with which it
covers all those who cannot get jobs (see Burda 1988). But again these
benefits have to be paid for. Higher unemployment benefits would lead to
more people being unemployed if those with jobs resisted the implied cut in
their consumption. The neo-classical route would have the actual real cost
of labour rising leading to rationalization of production; alternatively in a
conflicting claims framework (e.g. Rowthorn 1977) the level of unemploy-
ment required to suppress conflict over real incomes (the NAIRU) would
increase. Again collective discipline in bearing the costs is necessary to pay
for higher benefits, in addition to the social solidarity required to initiate them.
Similar remarks apply to other state benefits (pensions for example) which
also represent ‘consumption spreading’, in that entitlements to consump-
tion are extended beyond those in paid work. Maintaining the generosity
134 ANDREW GLYN
of state benefits ensures that the costs of slower growth are not met dis-
proportionately by those dependent on them. But again it is necessary that
there should be collective discipline amongst the employed to accept their
share of the costs. In the short run the problem may be concealed by rising
state deficits, reduced investment, and so forth. But eventually tax rates have
to rise, and if the implied cut in (the growth of) living standards is not
accepted, the inevitable result is that the costs are pushed on to those made
unemployed. Consumption spreading, as well as being an implication of
effective employment spreading, is a broader category as it covers the whole
operation of the benefit system. The common element is the necessity for
both social solidarity and collective discipline if wide spreading of the costs
of slower growth is to be a durable pattern.
Social solidarity and collective discipline are precisely the virtues claimed
for at least some of the so-called corporatist countries—that is the Scan-
dinavian countries and Austria (see Rowthorn, Pohjola, Pekkarinen this
volume). Deliberate and successful spreading of employment and consump-
tion, through the policies described above, might be expected, therefore, to
characterize successful corporatist countries. Such an approach is certainly
the antithesis to the market solution to restoring high employment under
conditions of slower growth. In ideal form this consists of relying on the
impersonal pressure of the reserve army of labour to drive down wages and
productivity in the market sector until the natural rate of unemployment is
restored—in effect there is forced consumption and employment spreading.
Whilst there is a common element in the outcomes (slower growth of
consumption per worker) the overall configurations are different—for ex-
ample greater hours reductions and increases in state employment in the
interventionist/collective response to slower growth.
Attempts to implement market solutions where institutional structures (e.g.
trades unions) resist lead (via nominal or real wage rigidities for example) to
a heavy concentration of the costs of slower growth on particular groups.
Those most likely to suffer would be those without jobs (who would also
tend to face less than generous benefit systems) and more generally all those
dependent on social welfare, cuts in which are usually an integral part of
programmes to restore ‘market forces’. Resistance has a similar effect (in
terms of the concentration of costs on the unemployed) where deliberate
policies of employment spreading are opposed by groups not prepared to
bear a share of the costs.
The aim of this chapter is to use this general framework of employment
and consumption spreading to explore the evolution of the corporatist
countries since 1973 and to assess whether they diverge systematically from
other OECD countries, especially those within the European core which
provide the obvious frame for comparison. Sections 5.2 and 5.3 examine
how the generally superior employment record of the corporatist countries
(see Rowthorn and Pohjola this volume) has been maintained, and in par-
ticular the respective roles of the state and the market sector. Section 5.4 re-
cords how the various patterns of employment and consumption spreading
have been reflected, in the longer run if not always in the short run, in growth
of consumption out of earned incomes for the average worker. This allows
an assessment to be made of the degree to which policies for employment
and consumption spreading have been carried out in various countries, of
their contribution to differences in employment performance, and of the
degree to which either collective restraint or market pressures had to bear
down on the growth of consumption.
5.2 DEMARKETIZATION AND THE TREND IN EMPLOYMENT
Our analysis of trends in participation in paid employment begins with the

availability of work in the market sector.2 The total number of hours worked
in the market sector depends on the balance between the growth of output
and the growth of hourly labour productivity. In order to control for differ-
ences over time and across countries in the potential labour supply the best
measure of work in the market sector is total hours worked per head of the
population of working age.3 This in turn depends on the balance between
the growth of output per head of the population and the growth of labour
productivity. Where the former is the faster an increasing number of hours
are required in the market sector.4 In the reverse case the amount of work in
the market sector is declining, a process which we term ‘demarketization’ by
(rough) analogy with the well-known phenomenon of deindustrialization.
Demarketization does not necessarily imply a falling employment rate (ratio
of employment to population) for the economy as a whole, since the labour
time ‘released’ from the market sector may be absorbed by increases in
employment in the state sector and by reductions in the average number of
hours worked.
Table 5.1 presents a simple decomposition for the employment rate,
showing how its trend depends on the balance between demarketization on
the one hand and rising state employment or reduced working time on the
other.5 In order to focus on comparisons between the corporatist countries
(CCs) and other European countries data is shown for the five CCs
2
The market sector excludes that part of GDP produced within the ‘general government’
sector; thus it includes public enterprises producing commodities for sale.
3
For conciseness ‘population’ is used henceforth as shorthand for population of working age
when considering access to work; clearly the availability of work should be judged in relation to
the latter rather than to the whole population.
4
This follows from the identify MHIP - O/P x MH/O, where MH is total hours worked in
the market sector, P is population of working age, and O is market sector output.
5
The decomposition on which the analysis of the employment rate is based is set out on p.
162 of Appendix 5.A.
TABLE 5.1 Demarketization and employment, 1965–1986 (average annual percentage changes)
Corporatist countries European core

1965–73 1973–9 1979–86 1965–73 1973–9 1979–86
1. Market output per

head of population 3.7 2.1 1.9 4.1 1.6 0.6
2. - Productivity –5.9 –3.8 –2.6 –5.5 –3.5 –2.8
3. = Hours in market
–2.2 –1.8 –0.7 –1.4 –1.9 –2.2
sector/population
4. + Share of state
0.9 0.9 0.5 0.3 0.6 0.3
employment
5. + Fall in average
1.3 0.7 0.1 0.7 0.6 0.5
hours worked (FT)
6. + Share of part-time n/a 0.3 0.1 n/a 0.2 0.2
7. = Employment/population:
(3)+(4)+(5)+(6) 0.0 0.2 0.0 –0.3 –0.6 –1.1
Sources and notes: For definitions, sources of variables, and the decomposition underlying the table see Appendix 5. A; population is of working age
(15–64).
–1.1
combined (Austria, Denmark, Finland, Norway, and Sweden) and for five
countries in the ‘European core’ (EC) combined (Belgium, France,
Germany, The Netherlands, and the UK) .6 Tables 5. A1–A3 in the Appendix
provide the same information for each corporatist country, and for Belgium
and The Netherlands which are perhaps the most closely comparable
countries in the European core.7 These tables form the basis for comments
in the text on specific features of particular countries. The appendix tables
also provide the data for Switzerland, which is the subject of another chap-
ter in this volume and often taken to have the most liberal economic policies
of the small European countries, and for the USA and Japan whose more
thoroughgoing market orientation is often contrasted with that of Europe.
5.2.1 The golden age

In the period 1965–73 (the end of the ‘golden age’ pattern of post-war
growth—see Glyn et al. 1990) the outstanding feature of the CCs taken as
a group was their faster rate of demarketization (line 3). All the European
countries considered showed a decline in total work done in the market
sector per head of the population of working age, but in the CCs it
proceeded nearly 1 per cent per year faster on average than in the EC
countries. The magnitude of the CCs’ 2.2 per cent per year decline can be
judged by the fact that had it continued until the present then the amount
of market work undertaken per head of the population would now be half
the level of 1965!
Such very rapid potential release of labour from the market sector did not
lead to a declining employment rate in the CCs (line 7). On the contrary,
rises in state employment and cuts in average hours worked absorbed all the
reduced availability of work in the market sector and so the employment rate
was maintained.8 In the EC the rate of demarketization was distinctly slower
(reflecting both faster output growth and slower productivity growth); but
the extent of employment spreading via rises in state employment, and
especially through cuts in working time, was considerably less pronounced
6
Data for the groups is a simple unweighted average of the data for the individual countries,
unweighted in order to suggest ‘typical’ performance. Data limitations prevented the analysis
of a broader sample of European countries; Italy, whose national accounts have recently been
hugely, but only incompletely, revised is the most obvious omission. As explained in the
Appendix in some instances it is necessary to use quite broad approximations to particular
variables.
7
Rankings of countries according to the degree of centralization of bargaining tend to show
The Netherlands and Belgium just below the Scandinavian countries and Austria; see Calmfors
and Driffill (1988: 18). Their size and openness also make them especially comparable.
8
Regression analysis, reported in more detail in the next section, which includes dummy
variables for the CCs, confirms that in the golden age they displayed 1.1% per year faster
demarketization (equation 14) than other OECD countries with the same productivity growth,
but this was precisely offset by faster increases in the share of state employment (equation 27)
and reductions in average hours worked (equation 30).
138 ANDREW GLYN
than in the CCs so that even in the golden age the employment rate declined
a little.
A rapid rate of demarketization characterizes every one of the CCs (Table
5.A1). But in both Austria and Finland the experience was closer to the EC,
in that the extent of employment spreading was insufficient to maintain the
employment rate. This supports the idea (Rowthorn, Pekkarinen this
volume) that Austria represented at that time a narrow or ‘exclusive’ form
of corporatism and that Finland was only in the process of evolving towards
the typical corporatist pattern. By contrast, in Belgium and The Netherlands
demarketization was as rapid as in the CCs, but in The Netherlands in
particular this was associated with a very poor employment performance—
the employment rate falling by almost 1 per cent per year. This was an early
example of a much more widespread EC phenomenon after 1973. It suggests
difficulties in securing the consumption restraint which would have been
necessary better to preserve market sector work or to finance employment
spreading at CC rates.
The data for demarketization during the golden age confirm casual
impressions as to variations in economies’ degree of market orientation. In
the golden age two countries, Australia and the USA, experienced ‘market-
ization’ (increased total hours worked in the market sector per head of the
population) and two more (Switzerland and Canada) demarketization of
around 0.5 per cent per year. Cuts in working time and increases in state
jobs were correspondingly small. The EC countries appear in general to lie
between the extreme demarketization of the CCs and the most market-
orientated economies mainly outside Europe.
Table 5.A4 provides some rough indicators of the level which the proces-
ses of demarketization had reached in the various groups of countries at the
end of the golden age when the contrast between the CCs and the rest was
at its sharpest. The extremely high shares of state employment in Sweden
and Denmark in 1973 stand out, together with substantially shorter hours
of work in Sweden than elsewhere.9 In terms of state employment Austria is
close to the EC average. Finland’s relatively low level of development was
reflected in less state employment and apparently longer hours of work than
the EC countries. In general, however, rapid golden age employment spread-
ing in the CCs was not simply a process of catching up with the general
European trend. Whilst the CCs are not an entirely homogeneous group in
this respect, in Sweden and Denmark the trend towards increasing state
employment and declines in working hours had proceeded further by the
time that growth slowed down.
9
It is likely that the rapid reductions in hours worked in Japan in the golden age, which were
the counterpart of substantial demarketization there, were a response to the extremely long
hours of work there—see notes to Table 5.A4.
5.2.2 The intershock period

The intershock period, between 1973 and 1979, is characterized by slower
growth of market sector output. But in relation to demarketization the
contrast between the CCs and the EC shifts. The CCs show a reduction in
the rate of demarketization as productivity growth fell more than output
growth. In the EC, by contrast, wifh the decline in output being nearly 1 per
cent per year steeper, productivity growth declined less than output growth
and the rate of demarketization increased. With the trends in employment
spreading (state employment and average hours) continuing at golden age
rates in both groups, slower demarketization meant that the trend of the
employment rate actually improved in the CCs in the intershock period. This
remarkable result was not repeated in the EC; on the contrary the speed-up
of demarketization was associated with a deterioration of the employment
trend there.10
It is important to emphasize that the slower average decline in output
growth in the CCs is not due just to the peculiar case of Norway, where the
exploitation of North Sea oil led to market output growing faster after 1973.
For in both Sweden and Denmark output growth slowed down less than in
the EEC countries. Moreover in both these countries productivity growth
slowed down much more than in the EEC. This strongly suggests that
deliberate policies to support market sector jobs were effective—perhaps in
underpinning output as well as preventing redundancies (see Landesmann
this volume).11
The contrast in employment patterns between the groups is epitomized
by the comparison of The Netherlands with Sweden; with very similar
growth of market output Sweden achieved an extraordinary increase of the
employment rate of nearly 1 per cent per year, whereas The Netherlands
suffered a decline of more than 1 per cent per year. The accounting explana-
tion is slower productivity growth, faster declines in working hours for
full-timers, and much faster increases in state employment in Sweden.
Underlying the Swedish pattern lay a whole range of interventionist policies
(see Bosworth and Rivlin 1987).
Once again Austria and Finland seem less typical of the CC pattern, with
demarketization continuing at the previous rate during the intershock
10
Regressions confirm that the CCs no longer displayed significantly faster rates of hours
reductions after 1973 (equations 31 and 32); only in terms of state employment growth did they
remain exceptional, but the extent of this faster growth was halved after 1979 as compared to
the golden age (equations 27–9).
11
For a given fall in output growth market work was better preserved after 1973 in the CCs
since the fall in productivity growth was some 0.9% per year more (and thus job loss less) than
elsewhere (equations 22 and 23). The exceptional case of Norway, where output and produc-
tivity growth were maintained, tends to mask this important tendency for more productivity
slow-down in the CCs.
140 ANDREW GLYN
period, and in Finland there was insufficient employment spreading to

prevent a sharp deterioration in the employment position.
5.2.3 After 1979

The contrasting trends for demarketization which emerged in the intershock
period were consolidated in the period after 1979, which includes both the
oil and ‘monetarist’ shocks (see Glyn et al. 1990) and the subsequent
recovery (although this is only partially captured in our data which reach up
to 1986). The rate of demarketization slowed down further in the CCs, as
output growth was practically maintained and productivity growth fell back.
Conversely in the EC demarketization speeded up again as output fell
sharply and by more than productivity growth. A comparison with the golden
age brings out the extent of the role reversal—before 1973 demarketization
was proceeding nearly 1 per cent per year faster in the CCs, after 1979 it was
nearly 1.5 per cent per year faster in the EC.
In both blocks the rate of growth of state employment fell off radically after
1979. In the CCs the rate of decline of working time also slowed, but weaker
employment spreading simply counterbalanced slower demarketization,
without adverse effects on employment. In the EC, however, reduced
employment spreading coincided with the speed-up of demarketization and
thus contributed to a drastic (more than 1 per cent per year) deterioration
in the employment rate. The pressures to reduce employment spreading,
because of its impact on the consumption of those at work, were common
to both groups of countries. What seems to distinguish the CCs as a group
was the combination of state intervention and consumption restraint which
held down the rate of decline of market sector work.
Once again Austria is the exception to the CC picture, with very poor
output growth leading to rapid demarketization and to sharp declines in the
employment rate. Its overall pattern is rather similar to that of Belgium and
suggests a common inability to deal with the restructuring of uncompetitive
heavy industry. When The Netherlands is again compared to Sweden the
continuing huge discrepancy in employment performance is now due, in a
proximate sense, to lack of output growth in The Netherlands rather than to
lack of employment spreading as in the intershock period. Finland, however,
decisively escapes its ‘transitional’ status and turns in the most dynamic
performance amongst the CCs, with good output performance, rapid
productivity growth, and with moderate demarketization fully absorbed by
increases in state employment and hours cuts.
The first, and unsurprising, lesson of this preliminary analysis of employ-
ment patterns is that the performance of the corporatist countries since 1973
is not homogeneous. Austria’s deteriorating employment performance is
typical more of the EC than of the corporatist pattern; its relatively good
unemployment figures owing much to labour supply factors, especially the

exclusion of many immigrant and women workers (see Rowthorn and Glyn
1990). Finland only achieved corporatist ‘best practice’ of limited de-
marketization and adequate employment spreading after 1979. Norway had
the special assistance of the North Sea to lubricate its employment main-
tenance policies. With Denmark’s good employment record tarnished by the
substantial rise in unemployment, this leaves Sweden as the outstanding
corporatist performer.
Despite these reservations, there is an undeniable contrast between the
employment record of the Scandinavian countries, especially since 1979,
and the disastrous performance of the European core countries. Fig. 5.1
summarizes this section by highlighting the strikingly different trends in
demarketization in the two groups of countries, and the failure of employ-
ment spreading to absorb the labour time released from the market sector in
the EC where demarketization increased after 1973. It seems, therefore, that
the character of demarketization changed after 1973 and that particular sets
of policies and institutional responses were required to hold employment up.
It is to a further examination of this issue that we turn now, by analysing the
shifting patterns in a more precise way.
FIG. 5.1 Demarketization and employment

Notes: CC refers to a simple average of the 5 corporatist countries (Austria, Denmark,
Finland, Norway, Sweden); EC refers to a simple average of Belgium, France, Germany,
The Netherlands, and the UK; MH refers to total hours worked in the market sector per
head of the population of working age; EMP refers to the ratio of employment to population
of working age.
142 ANDREW GLYN
5.3 POSITIVE AND NEGATIVE DEMARKETIZATION
Rowthorn and Wells (1987) use the term ‘positive deindustrialisation’ to

indicate a situation where rapid productivity growth in industry releases
workers for employment in the services sector and where demand for
industry and services grows fast enough (and in parallel) to ensure full
employment. We may use the term ‘positive demarketization’ to indicate a
similar process of release of labour time from the market sector for absorp-
tion in non-market work or as reduced working time, whilst total access to
paid work—the employment rate—is maintained. Demarketization is ‘negat-
ive’ when the labour time released by the speed of productivity growth in
relation to market output is not absorbed by rises in state employment and
falls in working time and thus leads to a deterioration in employment
performance.
Positive demarketization reflects the role of state policies and labour’s
bargaining objectives in permitting the benefits of rapid productivity growth
in the market sector to be used partly to expand state services and to cut
hours. Negative demarketization results where the market sector fails to
generate sufficient work to provide jobs for the working-age population and
the state and union movement do not offset this by spreading employment
sufficiently.
Two variants of negative demarketization may be distinguished. The first,
‘failed interventionism’, would occur where extensive employment spread-
ing was implemented (state employment, hours cuts, supporting market
sector jobs), but given the circumstances (which could include a particularly
weak market sector) there was insufficient support for the implied degree of
consumption restraint necessary to maintain the overall employment rate.
The second case, ‘market failure’, would describe a case where policy
emphasis was on encouragement of market forces, but they were insuffi-
ciently powerful to impose the consumption spreading necessary to maintain
employment within the market sector.
The patterns of demarketization and employment spreading are demon-
strated in a series of cross-country regressions reported in detail in Table
5.A6. These cover our sample of eleven European countries, plus the USA,
Japan, Australia, and Canada (the other OECD countries which had the
relevant data). Each regression aims to discover whether a variable (such as
the change in the employment rate) is significantly correlated with a relevant
structural feature (for example the growth of total hours worked in the
market sector per head of the population).
The shift towards increasingly negative demarketization for the OECD
countries as a whole is most clearly signalled by the relationship between the
trend in market work (total hours per head of the population) and the trend
in the employment rate. Prior to 1973 there was only a very weak tendency
for a faster rate of demarketization to be associated with worse employment

performance. After 1973, however, each 1 per cent faster rate of demarket-
ization was associated with a 0.6 per cent per year fall in the employment
rate; after 1979 the relationship was extremely tight (Table 5.A6 equations
1–5).12 Demarketization was strongly negative after 1973 in the sense that it
was associated with slower employment growth, whereas previously it was
practically counterbalanced by more employment spreading. The contrast
between Figs. 5.2 and 5.3 illustrates this very important structural change in
the functioning of the capitalist economies.
Fig. 5.2, showing the golden age patterns of employment performance and
market hours, suggests a U-shaped relation. Two groups of countries dis-
played exceptional employment performance. The Nordic CCs (apart from
Finland) achieved this through rapid demarketization combined with extens-
ive employment spreading. Others (the USA, Canada, Australia, all regarded
as liberal and with rather decentralized bargaining) showed increases in the
employment rate through rising work in the market sector. Countries in the
EC (together with Japan, Austria, and Finland) displayed more or less
substantial demarketization with somewhat declining employment. The idea
FIG. 5.2 Employment and market hours, 1965–1973
12
Regressions (4)–(5) for changes in the growth rates of these variables confirm this picture,
with a larger and more significant coefficient for the final period. To avoid repetition we will
only comment on the regressions for changes in the growth rates where they seem particularly
helpful in clarifying the patterns.
144
FIG. 5.3 Employment and market hours, 1979–1986
that there were two routes to outstanding employment performance before

1973 is an interesting precursor of the much discussed post-1973 relation-
ship between employment and bargaining (see Calmfors and Driffill 1988
and Rowthorn and Pohjola this volume).13
These two routes to employment success were still followed in the inter-
shock period. But after 1979 Fig. 5.3 shows a very different situation. Almost
all the countries where employment rates were maintained showed little or
no demarketization; only Finland represented a partial example of the earlier
CC pattern of substantial demarketization offset by increases in state
employment and hours cuts.
The proximate determinants of the availability of market work also
changed in a very significant way. Before 1979 there was no relationship
between the growth rate of market output (per head of the population) and
the growth rate of market work (equations 6–8). After 1979 however there
was a strong and significant correlation implying that 1 per cent faster growth
13
Golden age differences in the trend of the employment rate have to be interpreted carefully
since they are even less closely associated with changes in unemployment rates than was the
case after 1973 (see Rowthorn and Glyn 1990). But even if these differences wholly reflected
variations in the trend in women’s participation rates, for example, there may well have been
distinct economic forces contributing to these changes in the two groups of countries with big
increases in employment rates—expansion of state employment providing part-time jobs for
married women in the CCs and the economic pressure of slow growth in living standards
encouraging two-earner families in the more market-orientated economies.
of market output was associated with 0.7 per cent faster growth (slower fall)
of market work.14 Fig. 5.4 shows the strong relationship after 1979 when the
trend of the overall employment rate, and not just market work, was sig-
nificantly increased by a faster growth rate of market output (equations
11–13). This establishes that the basis of negative demarketization (with
associated inability to maintain the employment rate) was failure to expand
market output. But Fig. 5.4 also shows that the strong positive relationship
between market work and market output was heavily dependent on the
exceptional output performance of a few countries (especially Finland,
Norway, and Japan with rapid output growth). A mass of countries with
output growth around 1 per cent per year showed a tremendous range of
demarketization (by definition depending on productivity growth).
Indeed rapid productivity growth was consistently significantly and posit-
ively associated with demarketization (equations 14–18); but here too the
impact of a 1 per cent faster rise in productivity in reducing market work
increased (from 0.27 per cent before 1973 to 0.72 per cent after 1979). This
suggests that those countries which increased productivity rapidly after 1979
tended to achieve this by rationalization in the context of slow growth of
FIG. 5.4 Market work and market output, 1979–86
14
The regressions for changes in growth rates confirm that it was ability to maintain market
output growth after 1979 that was associated with reduced demarketization (equations 9
and 10).
146 ANDREW GLYN
production, rather than by Verdoorn-type exploitation of dynamic eco-

nomies of scale characteristic of rapid output growth in the golden age.15
Thus productivity growth was increasingly associated with loss of market
sector work. It was not, however, strongly associated with a declining
employment rate (equations 24–6). This is because there was a significant
and increasing tendency for rapid productivity growth to be associated with
the rate of hours reduction (equations 27–9), and after 1979 a significant
tendency for the share of state employment to rise faster in countries with
faster market sector productivity growth (equations 30–2). Employment
spreading after 1979 seems, therefore, to have been a response to the loss of
market sector work implied by rationalization and productivity growth in the
context of stagnant output.
Before 1979 there was no relationship whatever between employment
performance and total employment spreading16 (see equations 33–4), con-
firming the different routes to good employment performance in that period.
After 1979, however, those economies with the most employment spreading
experienced worse employment trends (equation 35).17 So employment
spreading was quite inadequate to combat loss of market sector work (see
Fig. 5.5), and may of course have increased it to the extent that it required
further consumption restraint. France showed the greatest degree of employ-
ment spreading after 1979—the early policies of the Mitterrand government
included hours cuts and increases in state employment with explicit employ-
ment objectives (Sachs and Wyplosz 1986)—but disastrous employment
performance in the years up to 1986. But the European core also exhibited
cases of’market failure’ as well as ‘failed interventionism’ (see p. 134 above);
for in The Netherlands, where employment performance was disastrous,
little employment spreading was attempted.
The analysis of this and the previous section yields some interesting con-
clusions about the superior employment performance of (most of) the CCs.
The crucial role of the market sector in maintaining the momentum of employ-
ment growth in these countries, through both limiting the decline in output
growth and moderating the rate of productivity growth, forms the starkest
comparison with the EC countries. Both interventionist state policy and
bargaining restraint contributed to this outcome. Maintaining market sector
work allowed some winding down after 1979 of the employment spreading
15
Equations 19–21 confirm the breakdown after 1979 of the very close golden age relation-
ship between productivity and market output growth.
16
Our employment spreading variable (the sum of the absolute values of changes in the share
of state employment and of falls in average hours worked) unfortunately cannot include jobs
supported in the market sector (by subsidies etc.), for which there is no comprehensive
comparative data, so it reflects just hours reductions and rises in state employment.
17
To some extent this effect may be relatively automatic; large-scale loss of market sector
work will tend to increase the share of state employment without deliberate employment-
creating policies in the state sector and cuts in overtime will reduce hours of work without
deliberate action on the part of unions.
147
FIG. 5.5 Employment spreading, 1979–86
mechanisms which had been such a feature of the CCs in the golden age.
Employment spreading at very high rates was no longer necessary to main-
tain employment performance in the CCs, whereas in the EC failure in the
market sector could only have been offset by employment spreading on an
increasing, and presumably unfmanceable, scale. The slow productivity
growth, and the extensive employment spreading in the intershock period,
had already constrained the growth of consumption per worker in the CCs.
The next section examines the implications of consumption spreading for
workers’ living standards, and shows how the squeeze was severely com-
pounded by rising transfer payments.
5.4 EMPLOYMENT, TRANSFERS, AND CONSUMPTION FROM

EARNINGS
Employment spreading through cuts in working time and increases in the

share of state employment also represents the spreading of resources poten-
tially available for consumption. The growth of output per worker (state and
market sectors) is reduced below that of hourly market productivity growth.
But the pressure on ‘consumption from earnings’18 and state transfers may
18
It is difficult to find an appropriate name for the trend of consumption out of earned
incomes. Its designation as ‘consumption from earnings’, rather than the more obvious
148 ANDREW GLYN
be eased if other expenditure claims are squeezed. The share of net exports
may be reduced as a result of an improvement in the terms of trade or
financed by a deterioration in the current account. The share of private
investment or the share of government purchases of market output (medical
supplies or new hospitals for example) may be squeezed. Finally the share
of consumption out of property incomes could be reduced (by taxation or
falling real interest rates for example). Conversely that part of consumption
financed from earnings will have to be further restrained if the share of
transfers rises; taxes on earnings must rise to transfer claims on consumption
to those not engaged in paid work.
The main issue here is to trace the course of consumption from earnings
over the period of slower growth, isolating the influences of employment
spreading, of the trend in transfers, and of the ‘raiding’ of other expenditure
claims. We comment first on the data for our two groups of countries—the
CCs and the EC—presented in summary in Table 5.2, whilst drawing on
the more detailed data for the individual countries in Table 5.A5.
5.4.1 The golden age

The first part of Table 5.2 shows the transition from market sector produc-
tivity to consumption from earnings per worker. First of all the effect of
employment spreading is shown in the move from market sector hourly
productivity (line 1) to market output per worker employed in market and
state sectors (line 2). Since employment spreading was stronger in die CCs,
as already described, the differential between productivity and output per
worker was correspondingly greater in die CCs in the golden age (some 2
per cent per year as against 1 per cent in the EC).
The transition from output per worker to adjusted output per worker (line
2 to line 3) shows the effect of other expenditure claims in either squeezing
or boosting the growth rate of what was available for consumption from
‘workers’ consumption’, is chosen to emphasize that it is a somewhat narrower category. First

of all workers may own some property from which income is derived and consumption financed
(the most obvious example being imputed rent on owner-occupied houses and associated
imputed consumption). Comprehensive data to analyse this do not exist, which is unfortunate
inasmuch as owner occupation has been increasing in a number of countries. Secondly workers’
families receive cash benefits from the state which are not included in our estimate of earned
incomes (after deduction of direct taxation, social security payments, and interest on consumer
debt). Fortunately comprehensive data on these transfers (essentially various forms of child
benefit) have been published by OECD since 1973 (1989). These show that in 1973 cash
transfers, for a family with two children, varied between 0% (the USA, Japan, and Germany)
and 16% (Belgium) of the average post-tax earnings of production workers. The average for the
five CCs was 8.5%, whereas for the five EC countries the share averaged 8.2%. Most important
for our analysis, by 1985 these averages had increased to 12.0% and 10.9% respectively. Thus
our calculations for consumption from work represent some underestimation after 1973 of
average growth of workers’ earnings, but the effect is generally not very large.
TABLE 5.2 spreading, 1965–1986 (average annual percentage changes)
EMPLOYMENT AND ACCESS TO CONSUMPTION

European core
Corpora tist countries
1965–73 1973–79 1979–86 1965–73 1973–79 1979–86
1. Hourly productivity 5.9 3.8 2.6 5.5 3.5 2.8
2. Market output
per worker 3.7 1.9 1.9 4.4 2.2 1.7
3. Adjusted market
output per worker 3.3 2.0 1.2 4.5 3.3 1.1
4. Consumption from
earnings per worker 2.4 0.4 0.6 3.5 1.6 0.1
Percentages of market output
(average change per year)
5. Net exports 0.1 0.5 0.3 0.2 0.0 0.5
6. Market investment 0.1 –0.7 0.0 0.0 –0.5 –0.4
7. Consumption from property 0.0 0.0 0.3 –0.1 –0.1 0.4
8. Consumption from transfers 0.4 0.7 0.1 0.4 0.8 0.2
9. Consumption from earnings –0.6 –0.6 –0.5 –0.4 –0.2 –0.6
Notes: Productivity is market output per hour worked in the market sector. The series for market output per worker (line 2) grows more slowly than
hourly productivity (line 1) because of falls in average hours worked and rises in the share of state employment (per worker refers to all workers
employed in market and state sectors). Adjusted market output means that the shares of net exports, market investment, government purchases, and
consumption from property are subtracted from market output at constant prices. This series (line 3), equalling consumption from transfers and
earnings, grows more slowly than total market output (line 2) if the sum of the other expenditure shares rises. Consumption from earnings is real
149
employee compensation per worker less estimates of taxes and social security payments, interest on debts, and savings. This series (line 4) grows
more slowly than adjusted market output per worker (line 3) if the share of transfers grows relative to consumption from earnings.
Sources: See Appendix 5.A for sources and detailed descriptions.

150 ANDREW GLYN
earnings and from transfers. The very small discrepancies between lines 2
and 3 imply that in total other expenditure shares did not change much in
either the CCs or the EC in the golden age.19 The second half of Table 5.2
confirms that the individual expenditure shares also were very constant
during the golden age.20
Finally the move from adjusted output per worker to consumption from
earnings (lines 3 and 4) shows the impact of the rising share of transfer
payments. In the case of both groups of countries the growth rate of
consumption from earnings in the golden age is reduced a further 1 per cent
per year as a result of the increased importance of transfers.21
The combined impact of these three effects—employment spreading, other
expenditure shares, and consumption spreading via transfers—is reflected in
the comparison between hourly productivity growth and the rise in con-
sumption from earnings (lines 1 and 4). In the golden age consumption from
earnings grew some 3.5 per cent per year more slowly than productivity in
the CCs and 2 per cent more slowly in the EC.22 In the CCs in particular
well over half of the benefit of (hourly market sector) productivity growth
was reflected in cuts in hours of work, increased employment in the state
19
As explained in the Appendix 5.A the calculation of consumption from property depends
on pro rata allocations of taxation, and interest payments between the categories of personal
income. The inevitably crude nature of many of these attributions (no taxes on state benefits,
attribution to the self-employed of an average wage so that only their incomes above the average
wage count as income from property, the prorating of income tax and interest payments between
the non-benefit sources of incomes, inclusion in some cases of imputed rent on owner-occupied
property, inability to include real capital gains as well as recorded nominal income flows,
assumption that savings ratios move in parallel for earned and unearned income) all make our
estimates of consumption from property very rough indicators. Accordingly not too much
weight should be placed on our finding that the CCs had only half as much consumption from
property (as a share of market output) as did the EC countries (6% against 12% in 1973).
20
The share of state purchases of market output (for consumption or investment) is omitted
from the table since it remains rather steady throughout the whole period. Note that our
categories of state purchases and state employment include the military. For most countries
military spending is recorded as quite a small proportion of GDP (around 3%), and is generally
stable but with some tendency to decline (OECD 19856: 48–52). In general therefore it is social
spending (particularly health, education, and social security) which is responsible for most
changes in state expenditure. The USA is the obvious exception. Military spending declined
from 7.5% of GDP in 1970to4.5%in 1979, but rose to 6.9% in 1986 and dominated the trend
in government purchases of marketed output.
21
The difference between lines 3 and 4 depends on the relative behaviour of the share of
consumption from transfers and the share of consumption from earnings (lines 8 and 9) since
adjusted MO per worker is the sum of the two categories of consumption. Thus even if the share
of consumption from transfers is constant, if other expenditure shares are rising, and thus the
share of consumption from earnings is falling, then the growth rate of consumption from
earnings is below the growth rate of adjusted MO, i.e. given the falling share of consumption
out of transfers and earnings combined a constant share of transfers squeezes the growth rate
of consumption from earnings. The magnitudes work out so that the difference between lines
3 and 4 is in fact roughly proportional to the sum of the absolute values of lines 8 and 9, a useful
rule of thumb when interpreting the table.
22
Regression 36 in Table 5.A7 confirms that workers’ consumption in the CCs grew 1.5%
per year slower, relative to productivity growth, than in other OECD countries.
sector, and an increased share of transfers.23 This employment and con-

sumption sharing represents a profound process of what may be termed
‘decommodification’, in the sense that potential growth of consumption is
taken in the form of increased leisure, of an increased share of employment
in state services, and in increased consumption from transfers allocated on
criteria other than participation in paid work (cf. Esping-Andersen 1990).
Such decommodification has strong egalitarian implications in the sense that
welfare is spread beyond those deriving income from work (via enjoyment
of the state services and receipt of transfers) and inasmuch as reduced hours
of work spread access to paid employment.
The wedge between the growth of productivity and the growth of con-
sumption from earnings was very large in the golden age in all the Scan-
dinavian countries, but the meagre growth of transfers in Austria helps to
put it nearer to the EC pattern. The Netherlands, where the increase in
transfers was fastest at the end of the golden age, shows a rate of decom-
modification equal to any of the corporatist countries other than Sweden.
But as already noted The Netherlands also had the worst employment
performance at that time. This suggests that whilst there was the political
support for income maintenance through transfers there was not the com-
bination of interventionist policies from the state and consumption restraint
from organized labour required to maintain employment levels. The Nether-
lands seems to have been displaying early symptoms of the kind of ‘failed
interventionism’ characteristic of many of the EC countries during the
period after 1973.
The contrast between the corporatist pattern of distribution and the
opposite market-dominated pattern is exemplified by comparing the USA
and Sweden. Between 1965 and 1973 they displayed an identical (1.2
per cent per year) growth of consumption from earnings. In the USA this
was just below the 1.9 per cent per year growth of hourly productivity—
employment spreading through cuts in working time and increases in state
employment, and consumption spreading through rising shares of transfers
were all very modest. In Sweden, by contrast, productivity grew at more than
5 per cent per year, but the great majority of the growth was swallowed up
in employment and consumption spreading leaving consumption from earn-
ings rising no faster than in the USA.
23
All forms of employment spreading have the same effect in reducing consumption from
earnings per worker. However cuts in average hours worked achieve this without any reduction
in consumption from earnings per hour, whilst increases in state employment and deliberate job
maintenance in the market sector require proportionate cuts in consumption from earnings per
hour. This maymake the spreading of employment through hours cuts more acceptable to those
bearing the costs although the frequency of demands for a ‘35–hour week with no loss of pay’
suggests that this may not always be true. It seems that the best single indicator of the effects of
employment (and consumption) spreading is consumption from earnings per worker.
152 ANDREW GLYN
5.4.2 The intershock period

As we have already seen employment spreading in the European countries
was broadly maintained in the intershock period. In addition, as Table 5.2
shows, the growth rate of the share of transfers practically doubled as
extension of the coverage and generosity of welfare benefits continued nearly
unabated despite the slower growth (OECD 1985a). This tended to magnify
the effect of the slow-down in productivity on workers’ consumption. Indeed
in the CCs consumption from earnings grew by less than half of 1 per cent
per year in the intershock period. In the EC countries, however, there was
extensive ‘raiding’ of non-consumption claims which allowed consumption
from earnings to grow about 1 per cent per year faster; whilst the share of
investment fell rapidly in both groups, in the CCs part of the resources
released were devoted to greater net exports whereas in the EC they were all
consumed.24
The extraordinary pressure on consumption from earnings involved in the
corporatist pattern of decommodification is exemplified by the fall in con-
sumption from earnings in Sweden of nearly 2 per cent per year in the
intershock period (see Table 5.A6 where the data is given for the individual
countries). There is a striking contrast with Switzerland whose ‘exclus-
iveness’ in employment performance (see Blaas this volume) is paralleled by
relatively weak consumption spreading. Productivity growth was nearly 2 per
cent faster in Sweden, but consumption from earnings grew 2 per cent per
year less! Belgium provides a different contrast with Sweden. As in Switzer-
land consumption from earnings grew much faster, relative to productivity
growth, than was the case in Sweden. But Belgium continued with extensive
employment and consumption spreading, so the explanation of faster growth
of consumption from earnings lay in greater consumption raiding in Belgium
as the fall in real net exports was greater.25 Of the CCs, raiding was much
the greatest in Denmark where both the share of private investment and that
of consumption from property fell more than in any of the other CCs; this
cushioned consumption from earnings which continued to grow at 1 per cent
per year in the intershock period (or about 3 per cent per year faster than in
24
On average the 5 EC countries moved from a 1.5% of GDP current account surplus in
1973 to a 0.7% deficit in 1979; the CCs moved from a 0.8% of GDP average deficit in 1973 to
a 2.2% of GDP deficit in 1979. So about half of inferior real net export performance of the EC
countries represented a bigger average deterioration of the balance of payments than in the CCs
(the other half representing a better trend of the terms of trade). Whilst Norway had much the
largest diversion of real resources into net exports, Sweden was the only CC where net exports
declined as a share (and only marginally).
25
Belgium showed the greatest deterioration of our 5 EC countries in the current account
of the balance of payments between 1973 and 1979 (from 2.9% of GDP surplus to 2.7% deficit).
Sweden also showed the biggest deterioration of the CC group (from 2.7% of GDP surplus to
2.2% deficit). Thus the better performance of net exports in Sweden was almost all absorbed
by a bigger deterioration in the terms of trade, rather than reflected in a better current account
performance.
Sweden). Despite the case of Denmark, the CC group continued to show

significantly slower growth of consumption from earnings (about 1.5 per
cent per year less than productivity growth) in the intershock period (Table
5.A7 equation 37). Fig. 5.6 confirms Sweden’s place as the extreme con-
sumption spreader.26
FIG. 5.6 Consumption from earnings and productivity, 1973–79
5.4.3 After 1979

As already described in the first section, the rate of reduction of hours of
work and the rate of increase in state employment both declined sharply after
1979. It is entirely consistent with this winding down of employment
spreading that there was simultaneously a virtual halting of the upward trend
in the share of transfers, despite the larger rise in unemployment.27 All these
trends away from consumption spreading tended to relax the pressure on
earnings from work. But contrary factors were also in operation. Both groups
of countries experienced a substantial rise in net exports as most of the
26
The only other case of consumption from earnings falling was Austria where a sharp rise
in our estimate of consumption from property was substantially responsible—this is hardly
consumption spreading!
27
Regressions (not reported) show that there was no significant difference between the CCs
and EC countries in the rate of transfer growth in any of the subperiods; nor was there a
significant relationship between the growth rate of the transfer share and that of market output.
154 ANDREW GLYN
individual countries had by 1986 moved into current account surplus.28

There was also a substantial and unprecedented rise in the share of consump-
tion out of property incomes (reflecting higher interest rates and a profit
recovery in a number of countries). One significant difference between the
two groups of countries was that the decline in the share of private invest-
ment continued in the EC, but not in the CCs. Nevertheless the extent of
employment spreading, net export increase, and transfer increase were all
bigger in the EC than in the CCs after 1979. The result was that while
consumption from earnings continued creeping up at about half a per cent
per year in the CCs, in the EC it fell to virtually nothing.29 The UK was
exceptional amongst the EC, keeping consumption from earnings growing
at 1.6 per cent per year between 1979 and 1986; employment spreading was
weak and the balance of payments was allowed to deteriorate.30
The most extreme case of decline in the growth of consumption from
earnings was The Netherlands. Whereas in the intershock period adjusted
output per worker had grown nearly 2 per cent per year faster than actual
output per worker (as net exports and investment were squeezed), after 1979
it grew nearly 2 per cent per year slower (as the shares of net exports and
consumption from property rose sharply). Combined with slower growth of
productivity this meant that consumption from earnings declined by 2 per
cent per year after 1979. Belgium and Denmark also sustained falls in the
level of consumption from earnings, as raiding of other expenditure claims
(private investment and net exports) was sharply reversed. In Sweden
consumption from earnings continued to fall after 1979, but less fast than in
the intershock period. In both Norway and Finland consumption from work
grew at a respectable 2 per cent per year after 1979. In the case of Norway
this was achieved by minimal spreading of employment and consumption,
and relatively little shift of resources away from consumption. Finland is
somewhat unique amongst the CCs in that productivity growth continued
at a rapid rate, and there was substantial spreading of employment and
consumption through state employment, transfers, etc. The weak link was
the continued decline in the investment share.
Fig. 5.7 charts the Swedish pattern for each year since 1965, bringing out
the extraordinary fact that average consumption from earnings in 1986 was
below the level of twenty-one years earlier, with a sharp fall from the
28
The exceptions were Denmark and Norway, which had current account deficits of 5.5%
and 6.4% of GDP respectively in 1986. In both cases deteriorations in the terms of trade (oil
prices in the case of Norway) absorbed all and more of the increase in net exports.
29
Regression 38 (Table 5.A7) shows that there was no significant difference after 1979 in
consumption from earnings in the two groups (allowing for productivity growth).
30
The explanation for continuing growth of consumption from earnings was not more
dynamic productivity growth; only in manufacturing did productivity rise fast and this was offset
by slower increases in services productivity.
FIG. 5.7 Sweden: consumption and employment
mid-1970s to the early 1980s.31 This represents an extreme case of the

process of decommodification described above where all workers’ share of
increased productivity over a twenty-year period is used to finance cuts in
hours, increased state employment, and rises in transfers. Sweden is unique
in this respect, but Austria and Denmark have displayed stagnant consump-
tion from work over the whole period since 1973 (Figs. 5.8 and 5.9).
Denmark is an interesting case because consumption from earnings con-
tinued to grow in the intershock period and then fell back sharply. The rise
in transfers, concentrated in the early 1980s as unemployment rose sharply,
slashed consumption from earnings. Only Finland and Norway (with the aid
of North Sea oil) maintain a significant growth in consumption from
earnings over the post-1973 period (Figs. 5.10 and 5.11). The very large
gaps in all these charts between the growth of productivity and the growth
31
Between 1973 and 1985 Swedish workers with children benefited from a small increase
in transfers; relative to net earnings receipts by a family with two children rose from 11.8%
to
13.9%. This represents only a small offset to the decline in consumption from earnings (see n. 18).
156 ANDREW GLYN
FIG. 5.8 Austria: consumption and employment
of consumption from earnings confirm the importance of decommodifica-

tion over the period as a whole.
Both The Netherlands and Belgium displayed the same pattern of delayed
adjustment of consumption from earnings already noted for Denmark (Figs.
5.12 and 5.13). Indeed The Netherlands represents the pattern in extreme
form, with a huge dip in consumption from earnings in the early 1980s. In
these three cases it seems that the necessary adjustment of workers’ living
standards to the conditions of slower growth of output and productivity had
to await the onset of mass unemployment. They stand in contrast to Sweden
in particular where even more extensive employment and consumption
spreading in the intershock period was absorbed by reductions in consump-
tion from earnings, without the market discipline of mass unemployment.
The charts show that all the CCs, except Austria, plus Belgium and The
Netherlands managed a definite recovery in consumption from work in the
mid-1980s. For the years since 1986 the data are not available for making
the full set of calculations; yet the figures for aggregate consumption and for
employment suggest substantial growth of consumption and employment in
FIG. 5.9 Denmark: consumption and employment
Sweden and Austria and also in Finland, The Netherlands, and Belgium
where unemployment fell strongly. It seems most probable that much more
moderate consumption spreading has continued so that these trends imply
definite growth of consumption from earnings. Denmark and Norway ex-
hibit the opposite trends of declining employment and falling consumption.
Whilst Norway’s performance has been strongly influenced by the vagaries
of the oil sector, the fact that unemployment had risen to over 5 per cent in
1989 suggests weakness in the effectiveness of the corporatist arrangements
there, for it is precisely in the context of difficult circumstances that the
solidity of these arrangements is tested. The rise in Danish unemployment
back to over 9 per cent in 1989 (only equalled or exceeded by France and
Italy in the European core) also highlights the weaknesses of corporatism
there. The recovery, and consequent fall in unemployment by about one-
third, in The Netherlands and Belgium still leaves unemployment rates not
far below 10 per cent. So there is as yet no clear evidence that the EMS for
example has replaced mass unemployment as an effective guarantee of
consumption restraint. With Austria’s record tarnished by poor employment,
158 ANDREW GLYN
FIG. 5.10 Finland: consumption and employment
Finland and Sweden are confirmed in their positions of outstanding per-

formers amongst the European countries (balance-of-payments difficulties
notwithstanding).
5.5 CONCLUSIONS
This chapter has analysed some important respects in which the corporatist
countries have performed exceptionally well in the years since 1965. In the
golden age years before 1973 they exemplified a process of positive
demarketization whereby a substantial portion of the quite rapid productivity
growth in the market sector was utilized to increase leisure and to increase
employment in (and thus provision of) state services. Allied to a rapid
increase in the share of transfers, which was a much more generalized feature
of the OECD countries, these trends implied that consumption from earn-
ings grew much more slowly than productivity growth.
In the intershock period the CCs managed to reduce the rate of
FIG. 5.11 Norway: consumption and employment
demarketization a little by limiting the decline in output growth and accept-

ing a substantial decline in productivity growth. Such deliberate policies of
job maintenance in the market sector prevented the deterioration in employ-
ment performance which characterized the ‘negative demarketization’ of
some but not all of the EC countries. Consumption and employment spread-
ing were maintained at a very high rate in the CCs, with the implication of
stagnant consumption from earnings; in Sweden the extent of solidaristic
redistribution implied an extended period of decline in consumption from
earnings (though the impact on consumption for many working households
was moderated by a second income as women’s participation rates rose and
living standards in the broader sense benefited from more leisure and
improved welfare services).
In the period after 1979 the limits to politically acceptable redistributions
were reached in the CCs; the seemingly inexorable rise in consumption and
employment spreading was substantially slowed and consumption from
earnings began to rise again in the mid-1980s. What is most striking in
Sweden especially is that the transition to a period of less redistributive
160 ANDREW GLYN
FIG. 5.12 The Netherlands: consumption and employment
growth required neither mass unemployment nor major cuts in the welfare
state to hold back or satisfy pent-up demands for faster growth of consump-
tion by workers. In this sense the solidarity and collective discipline has not
been decisively reversed, despite the talk in the early 1980s of the ‘crisis’ of
the Swedish model (see e.g. Lundberg 1985).
If Sweden exemplifies solidaristic redistribution and the limits it may
reach, the contrast between Finland and Denmark underlines the import-
ance of dynamism in the market sector. The much slower growth of output
and productivity in Denmark after 1979, together with the necessity of
reversing the unsustainable squeeze on investment and the balance of
payments, imposed a very heavy pressure on consumption from earnings in
that country. In Finland by contrast faster growth of productivity and output
allowed more redistribution, good employment performance, and quite
rapid growth of consumption from earnings.
There are obvious limits to the extent to which jobs can be preserved in
the market sector at the cost of very low productivity growth. Such ‘defensive
egalitarianism’ puts great strain on the growth of consumption from earnings.
FIG. 5.13 Belgium: consumption and employment
Rapid growth of productivity, even in the context of constraints on market

sector output growth, at least allows slow growth of consumption from
earnings to be compensated with improved social services or reduced hours
of work. Whatever the undoubted difficulties of maintaining the social
solidarity ahd collective discipline necessary for such a pattern it has great
attractions over the alternative of purely market-orientated rationalization of
the market sector. In the absence of perfect market-enforced flexibility such
a strategy allows average consumption from earnings to grow faster than with
egalitarian policies of redistributing work and consumption. But such negat-
ive demarketization unloads the costs of slower growth on those who lose
their jobs and on those especially reliant on state services. It is the singular
achievement of ‘social’ corporatism, above all in Sweden, that it largely
avoided this.
162 ANDREW GLYN
APPENDIX 5.A
Data Sources and Methods
The following list gives the basic sources and methods for the calculations presented
in the tables (ordered by line in the tables). Where these sources had to be supple-
mented from other data this is indicated in the country notes at the end.
Table 5.1
Lines (l)-(3) are connected by the following decomposition:
( 5.A1)
and lines (3) to (7) by the following decomposition:
( 5.A2)
where MH = hours worked in market sector;
POP = population of working age;
MO = market output;
EMP = total employment;
MSH = hours worked in market and state sectors;
FTE = full-time equivalent employment (total employment adjusted
for part-timers).
Thus from (5.A1) the growth rate of hours worked in the market sector per head of
the population equals the growth rate of market output per head of the population
less the growth rate of hourly productivity; and from (5.A2) the growth rate of the
employment population ratio (the ‘employment rate’) equals the growth rate of
market hours per head of population plus the growth rate of the ratio of total hours
worked to market hours worked (the share of state employment) less the growth rate
of average hours worked per full-time equivalent employee (or plus the rate of decline
of average hours worked) plus the growth rate of the ratio of total employment to
full-time equivalent employment (the effect of part-time employment).
1. Market output per head of the population: Market output is GDP at constant market
prices less the output of producers of government services (OECD 1988a: ii, tables
1 and 12 of each country section and comparable tables of OECD (1987) and OECD
(1983); for all national account (NA) series these three volumes are used for the
periods 1979–86, 1973–9, and 1965–73 respectively with appropriate linking).
Population is of age 15–64 (OECD 1988*: table 2).
2. Productivity: Market output per hour worked in the market sector. For a few
countries (Finland, Sweden, the USA) hours worked in the market sector are
available directly (NA table 15). For the rest they are estimated from total employ-
ment in the market sector (NA table 15) and the OECD’s index of average hours
worked (OECD Employment Outlook, various issues).
3. Hours in the market sector per head of the population: See lines (1) and (2).
4. Share of state employment: Measured as the ratio of total hours worked in the state
and market sectors to hours worked in the market sector. Except for the three
countries listed in the note for line (2) it has to be assumed that average hours worked
in the state sector are the same as for the market sector, so that the share of state
employment is the ratio of total employment to market employment (NA table 15).
5. Fall in average hours worked of full-time workers: Measured as the ratio of total
hours worked in the market and state sectors to full-time equivalent employment.
The latter is estimated assuming that a person in part-time employment (from OECD
Employment Outlook, various issues) is the equivalent of half a person in full-time
employment.
6. Share of part-time employment: Measured as ratio of total employment to full-time
equivalent employment (weighting part-time work as half full-time work). Share of
part-time work from OECD Employment Outlook, various issues.
7. Employment/population ratio: Total employment (NA table 15) divided by popu-
lation aged 15–65 (see line 1).
Table 5.2
2. Market output per worker. Number of workers is total employment, both em-
ployees and self-employed (NA table 15).
3. Adjusted market output per worker: Market output less shares of net exports,
market investment, government purchases (NA table 1), and consumption from
property (see below). Government purchases are government current expenditure on
goods and services (NA table 1) plus investment by producers of government services
(NA table 3) less output of government sector.
4. Consumption from earnings per worker: Consumption from earnings is total private
consumers’ expenditure (NA table 1) less consumption out of transfers (see line 8
below) and less consumption out of property incomes (see line 7 below).
5. Net exports: Net exports are exports of goods and services less imports of goods
and services both at constant market prices (NA table 1).
6. Market investment: Market investment is total gross fixed capital formation plus
increase in stocks, at constant market prices (NA table 1) less gross capital formation
by producers of government services (NA table 3). It should be noted that it includes
investment by government enterprises; but it is not ‘business investment’ since it
includes private housing.
7. Consumption from property: Post-tax income from property is calculated as total
income of the household sector (NA table 8) less transfers received (see line 8 below)
and less disposable income from work. The latter is calculated first by calculating the
post-tax wage of employees as employee compensation (NA table 8) less the average
tax rate (calculated as direct taxes as a percentage of household receipts of income
164 ANDREW GLYN
less government grants received—NA table 8) less the average rate of social security
contributions (social security contributions as a percentage of employee compensa-
tion plus the imputed gross wage of the self-employed—NA table 8) less (after 1973)
the average amount of interest paid from employee compensation (interest payments
as a percentage of household receipts of income less government grants—NA table
8). The disposable income from work of the self-employed is assumed equal to that
of employees (numbers of employees and self-employed are from NA table 15).
Disposable income from property is therefore (in principle) equal to excess of income
from unincorporated enterprises over and above the average post-tax wage imputed
to proprietors plus property and entrepreneurial income (rent, dividends, and inter-
est) less estimates of tax paid (assumed to be at the average rate) and interest
payments made out of that income (again assumed to be at an average rate). Personal
savings are allocated pro rata between disposable income from property and dis-
posable income from work, so that any changes in the savings ratio are assumed to
affect consumption from income from work and from property proportionately.
8. Consumption from transfers: Transfers are social security benefits and social
assistance grants (NA table 6); these are deflated by the price index of consumers
expenditure (implicit in NA table 1) and then expressed as a share of marketed
output. It is implicitly assumed that there is no taxation levied or savings made out
of these state benefits.
9. Consumption from earnings: See line 4 above.
Additional data sources and gaps for specific countries
Australia: The national account data for output and employment of ‘producers of
government services’ refer only to public administration and defence (less than 5%
of total employment). An alternative series for government employment (as a per-
centage of total employment) is available in OECD Historical Statistics, tables 2.13
and 1.13; this is said to be a ‘close approximation to the institutional sector “general
government” ‘, and represents more than one quarter of total employment. This
series is used for state employment (in line 15). It would be desirable to derive a series
for the output of these workers which could be subtracted from GDP to reach
marketed output. But our attempts to do so gave results which seemed inconsistent
with the national account data for government current and capital spending (the
government employment series being broader than ‘producers of government
services’). Accordingly it seemed best to abandon the attempt to estimate government
output and instead use the growth of GDP as the indicator of the growth rate of
marketed output; in that the OECD series for government employment stays very
steady as a percentage of total employment it is likely that the growth rate of marketed
output (and productivity) is substantially underestimated (by as much as one quarter)
on account of the slower (if any) growth of productivity in the government sector.
Total employment and employees: from OECD Labour Force Statistics, Australia
section, various issues.
Government investment: pre-1973 no constant price data were available, so es-
timated assuming price index is the same as for total fixed investment.
Average hours worked: average of series for men and women, average hours paid
for in non-agricultural sectors, ILO Yearbook, various issues.
Austria: Total employment and employees: OECD Labour Force Statistics.
Government employment: OECD Historical Statistics.
Government investment: no data before 1973 so that for these years government
purchases of marketed output refer only to current expenditure. After 1973 no
constant price data are available, so estimated assuming price index is the same as for
total fixed investment.
Average hours worked: US Bureau of Labor Statistics data for manufacturing sec-
tor only.
Belgium: Government employment: before 1970 OECD Historical Statistics.
Average hours worked: non-agricultural sector; ILO Yearbook.
Property income paid by personal sector: no data.
Canada: Total employment and employees: OECD Labour Force Statistics.
Denmark: There are no income and expenditure data for the government sector
prior to 1971 and no household sector data at all. But the availability of government
sector data on social security grants paid and taxes and social security contributions
received (NA table 6) made it possible to derive an approximation to the household
sector data from 1973 onwards (and a very rough calculation was performed for
1965). The main problem was estimating property income received by the household
sector. Especially since the focus is on changes in variables a reasonable approxima-
tion could be derived as net operating surplus (NA table 1; in effect the excess of
operating surplus over and above the average wage was used) plus net payments by
the general government of property income and operating surplus (NA table 6) plus
net receipts from overseas of property income (NA table 1). This is essentially the
proper category except that it includes the undistributed (net) profits of corporations,
but it may be noted that there is anyway wide variation between countries on the
extent of corporate business so that the calculation of consumption out of property
income should only be regarded as broadly indicative of relevant trends.
France: Total employment and employees: before 1970 from OECD Labour Force
Statistics.
Government employment: before 1970 from OECD Historical Statistics.
Japan: Government employment: for period prior to 1965 estimated from data on
employee compensation in the government sector.
The usual method of attributing an average wage to the self-employed left a negative
amount of property income from that sector, due to the low level of self-employment
incomes on average. Instead all the net operating surplus of the self-employed was
regarded as income from employment leaving zero (rather than negative) property
income. This illustrates in an extreme form that the method of attribution understates
the amount of property income in the unincorporated sector by focusing only on
average levels of income rather than making an assessment, based on the full
distribution.
The Netherlands: All data start in 1968. 1968 employment data from OECD
Historical Statistics.
Norway: The personal sector data begin only in 1975. In order to calculate income
166 ANDREW GLYN
and consumption from property, the data on taxation, social security payments, and
grants are used from the general government account. For 1973 and 1974 the
simplest approach is to make an estimate of personal sector savings (assumed to be
the same percentage of private sector savings as in 1975—NA table 1), which together
with consumption expenditures and tax and social security payments allows total
personal income to be estimated, and thus income from property as a residual after
deducting income from work and from government grants. For the period 1965–73
the same method is used as for Denmark.
Switzerland: Market sector output: there are no data for the output of the govern-
ment sector which can be subtracted from GDP to reach market sector output.
Estimates of this are made as the product of government employment (see below)
and average employee compensation in 1970 (to give a series for real output of
government producers at 1970 prices).
Total employment and employees: OECD Labour Force Statistics; number of
employees only available for selected years so interpolations necessary (constant
percentage assumed after 1980).
Government investment: no data so that government purchases of marketed output
refer only to current expenditure.
Average hours worked: up to 1973 ILO series for manufacturing sector; after 1973
ILO series for non-agricultural sector. This is obviously a particularly unsatisfactory
series.
Property income paid by the personal sector: no data.
TABLE 5.A1 Demarketization and employment in the golden age
(average annual percentage changes, 1965–73)
CCs EC USA Japan The Belgium Austria Denmark Finland Norway Sweden Switzerland
Netherlands
1. Market output
per head of
population 3.7 4.1 2.0 9.0 4.3 4.6 5.4 2.9 4.2 3.2 2.8 2.7
2. – Productivity –5.9 –5.5 –1.9 –10.5 –6.5 –6.7 –7.2 –5.4 –6.5 –5.1 –5.4 –3.3
3. = Hours in
market sector/
population –2.2 –1.4 0.1 –1.6 –2.3 –2.1 –1.7 –2.5 –2.3 –1.9 –2.7 –0.6
4. + Share of state
employment 0.9 0.3 0.0 0.1 0.3 0.3 0.5 1.5 0.5 0.7 1.2 0.2
hours worked 1.3 0.7 0.5 1.2 1.1 1.4 0.9 1.4 1.3 1.3 1.6 0.2
6. = Employment/
population:
(l)+(2)+(4)+(5) 0.0 –0.3 –0.2 0.4 –0.3 0.1 0.2 –0.3
–0.3 0.7 –0.2 –0.8
Source and notes: For definitions, sources of variables, and the decomposition underlying the table see the Appendix, above; population is of working age
(15–64).
TABLE 5.A2 Demarketization and employment in the intershock period
Netherlands
1. Market output
per head of
population 2.1 1.6 0.9 2.7 1.3 1.3 2.4 1.0 1.6 4.2 1.1 –0.5
2. – Productivity –3.8 –3.5 –0.7 –3.8 –3.4 –3.7 –4.1 –2.6 –4.0 –5.6 –2.8 –1.1
3. = Hours in
market sector/
population –1.8 –1.9 0.2 –1.1 –2.1 –2.4 –1.8 –1.7 –2.4 –1.4 –1.6 –1.6
4. + Share of state
5. employment 0.9 0.6 –0.1 0.1 0.3 0.7 0.6 1.3 0.8 0.6 1.2 0.4
+ Fall in average
hours
6. worked (FT) 0.7 0.6 0.5 0.6 0.1 0.8 1.1 0.4 0.5 1.0 0.7 0.3
+ Share of
part-time 0.3 0.2 0.0 0.1 0.6 0.2 0.1 0.1 0.0 0.4 0.6 0.0
7. = Employment/ 0.2 –0.6 0.6 –0.3 –1.1 –0.6 0.0 0.2 –0.8 0.6 0.9 –0.9
population:
(l)+(2)+(4)+(5)
Sources and notes: For definitions, sources of variables, and the decomposition underlying the table see the Appendix, above; population is of working age
(15–64).
TABLE 5.A3 Demarketization and employment after 1979
Netherlands
1. Market output
per head of
population 1.9 0.6 1.5 3.0 0.0 1.1 0.8 1.6 2.8 2.9 1.4 1.0
2. - Productivity –2.6 –2.8 –1.4 –3.2 –1.8 –3.4 –2.4 –1.6 –3.9 –2.8 –2.1 –2.1
3. = Hours in
market sector/
population –0.7 –2.2 0.1 –0.2 –1.9 –2.3 –1.7 0.0 –1.1 0.1 –0.7 –1.1
4. + Share of state
employment 0.5 0.3 –0.1 0.0 0.2 0.4 0.4 0.5 0.5 0.4 0.6 0.1
hours worked 0.1 0.5 0.2 0.0 0.1 0.9 0.3 –0.3 0.5 0.0 0.1 0.5
6. + Share of
part-time 0.1 0.2 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.0 0.0
7. = Employment/
population:
(l)+(2)+(4)+(5) 0.0 –1.1 0.3 –0.1 –1.5 –0.8 –0.9 0.3 0.0 0.5 0.0 –0.5
Sources and notes: For definitions, sources of variables, and the decomposition underlying the table see the Appendix, above; population is of working age
(15–64).
TABLE 5.A4 State employment, hours of work, and transfers in 1973
Netherlands
1. State employ-
ment (% of
total) 18.3 14.7 17.2 6.1 13.0 14.6 15.0 21.0 13.1 18.4 23.8 8.6
2. Av. working
hours annual,
full-time 1,873 1,898 1,954 n/a n/a n/a n/a n/a 1,980 1,930 1,711 n/a
3. Transfers
(% of market
output) 13.5 14.9 9.3 6.6 21.4 16 14.0 13.9 _ 9.4 15.9 14.4 10.5
Sources: State employment: OECD (1986: table 2.13). EC refers to simple average of Belgium, France, Germany, The Netherlands, the UK. Note that
figures for state employment include armed forces; these represented 2.1% of US employment in 1973 (2.7% if forces overseas are included), 2.1% of
employment in the EC and 1.7% on average in Finland and Denmark (the only CCs which publish the data in OECD Labour Force Statistics). Thus the
inclusion of armed forces does not explain the surprisingly high US figure. Hours worked: OECD, Employment Outlook (1989), table L. These data have
been roughly adjusted to hours for ‘full-time’ equivalent workers by assuming that part-time workers work half the hours of full-time workers. Data for
France, Germany, and the USA refer to employees, the rest to all workers. EC refers to simple average of France, Germany, and Italy, CCs to average of
Finland, Norway, and Sweden. Data in Maddison (1987: table A9) are somewhat inconsistent with the above table as they suggest that hours worked in the
USA in 1973 were below those in the EC countries (other than the UK); they also estimate that hours worked in Japan were much higher (20–5%) than in
Europe or the USA. Transfers: See Appendix, above.
TABLE 5.A5 Employment and consumption spreading, 1965–1986

CCs EC USA Japan
1965–73 1973–9 1979–86 1965–73 1973–9 1979–86 1965–73 1973–9 1979–86 1965–73 1973–9 1979–86
Average annual
percentage changes
1. Productivity 5.9 3.8 2.6 5.5 3.5 2.8 1.9 0.7 1.4 10.5 3.8 3.2
2. Market output
3. per worker 3.7 1.9 1.9 4.4 2.2 1.7 1.3 0.3 1.2 9.2 3.0 3.1
3. Adjusted market output
4. perworker 3.3 2.0 1.2 4.5 3.3 1.1 2.0 0.6 1.1 7.8 3.3 1.4
4. Consumption from
5. earnings per worker 2.4 0.4 0.6 3.5 1.6 0.1 1.2 0.2 0.9 7.5 1.8 0.7
5. Net exports 0.1 0.5 0.3 0.2 0.0 0.5 –0.1 0.1 –0.5 –0.2 0.7 1.1
6. Market investment 0.1 –0.7 0.0 0.0 –0.5 –0.4 0.1 0.2 0.0 1.3 –1.0 –0.2
7. Consumption from property 0.0 0.0 0.3 –0.1 –0.1 0.4 –0.2 0.0 0.3 –0.3 0.0 0.2
8. Consumption from transfers 0.4 0.7 0.1 0.4 0.8 0.2 0.5 0.2 0.1 0.1 0.8 0.1
9. Consumption from earnings –0.6 –0.6 –0.5 –0.4 –0.2 –0.6 –0.1 0.0 –0.2 –0.8 –0.6 –1.1
171
TABLE 5.A5 (com.)
172
Austria Denmark Finland Norway
1965–73 1973–9 1979–86 1965–73 1973–9 1979–86 1965–73 1973–9 1979–86 1965–73 1973–9 1979–86
Average annual
percentage changes
1. Productivity 7.2 4.1 2.4 5.4 2.6 1.6 6.5 4.0 3.9 5.1 5.6 2.8
2. Market output
per worker 5.7 2.4 1.7 2.4 0.8 1.3 4.5 2.4 2.8 3.1 3.6 2.4
perworker 5.1 1.6 0.8 2.1 2.5 –0.2 3.7 2.2 3.3 3.3 2.6 1.9
4. Consumption from
earnings per worker 4.7 –0.2 0.1 1.6 1.1 –0.8 2.9 0.9 2.1 1.5 1.5 2.0
ANDREW GLYN
5. Net exports 0.1 0.2 0.1 –0.3 0.4 0.7 0.4 0.6 0.1 –0.4 1.5 –0.2
6. Market investment 0.4 –0.5 0.1 0.3 –0.9 0.4 0.0 –0.6 –0.4 0.3 –0.9 0.2
7. Consumption from property –0.1 0.7 0.5 0.6 –0.6 0.4 0.4 0.1 –0.1 –0.3 0.0 0.3
8. Consumption from transfers 0.1 0.6 0.1 0.1 0.9 0.0 0.3 0.5 0.6 0.8 0.3 –0.1
9. Consumption from earnings –0.4 –1.1 –0.6 –0.7 0.1 –0.9 –0.8 –0.7 –0.3 –0.7 –0.8 –0.1
TABLE 5.A5 {com.)
Sweden Switzerland Belgium The Netherlands
1965–73 1973–9 1979–86 1965–73 1973–9 1979–86 1965–73 1973–9 1979–86 1968–73 1973–9 1979–86

Average annual
percentage changes
1. Productivity 5.4 2.8 2.1 3.3 1.1 2.1 6.7 3.7 3.4 6.5 3.4 1.8
2. Market output
perworker 2.6 0.2 1.4 2.9 0.4 1.5 4.9 1.9 1.9 5.1 2.4 1.5
perworker 2.4 1.2 0.2 2.8 1.4 1.2 4.9 3.7 0.5 5.1 4.1 –0.4
4. Consumption from
earnings per worker 1.2 –1.7 –0.8 1.8 0.4 0.7 3.8 1.9 –0.1 2.5 2.0 –2.2
5. Net exports 0.6 –0.1 0.7 –0.3 –0.3 –0.4 0.1 –0.5 1.2 1.0 –0.2 0.7
6. Market investment –0.5 –0.4 0.0 0.1 –0.6 0.7 –0.2 –0.2 –0.5 –0.1 –0.6 0.0
7. Consumption from property –0.1 –0.2 0.2 0.2 0.2 –0.1 0.0 –0.4 0.5 0.3 –0.1 0.6
8. Consumption from transfers 0.5 1.3 0.1 0.5 0.6 0.2 0.5 1.0 –0.1 0.9 1.1 0.0
9. Consumption from earnings –0.6 –0.8 –0.8 –0.5 0.0 –0.4 –0.5 0.0 –0.8 –0.9 –0.1 –1.1
Notes: Productivity is market output per hour worked in the market sector. The series for market output per worker (line 2) grows more slowly than
hourly productivity (line 1) because of falls in average hours worked and rises in the share of state employment (per worker refers to all workers
employed in market and state sectors). Adjusted market output means that the shares of net exports, market investment, government purchases, and
consumption from property are subtracted from market output at constant prices. This series (line 3), equalling consumption from transfers and
earnings, grows more slowly than total market output (line 2) if the sum of the other expenditure shares rises. Consumption from earnings is real
employee compensation per worker less estimates of taxes and social security payments, interest on debts, and savings. This series (line 4) grows
more slowly than adjusted market output per worker (line 3) if the share of transfers grows relative to consumption from earnings.
Sources: See Appendix, above, for sources and detailed descriptions.
173
174 ANDREW GLYN
TABLE 5.A6 Employment structure regressions for 15 countries

In each equation variables are measured as average annual rates over period defined
except where specified as changes in growth rate as compared to previous period.
Dependent variable Explanatory variables R2
Const CC
Employment/ Total
population 0.19 market hours
1 1965–73 0.17(2.0) 0.23
2 1973–9 0.70 0.56(3.3) 0.45
3 1979–86 0.21 0.58(8.1) 0.83
Changes in growth rates
4 1973–9
(from 1965–73) –0.06 0.40(3.9) 0.52
5 1979–86
(from 1973–9) –0.43 0.48(3.5) 0.48
Total market hours Marked output

6 –0.12 –0.20(1.5) 1.36(2.9) 0.45
7 –1.63 0.12(0.5) –0.39(0.8) 0.06
8 –2.13 0.75(2.7) 0.03(0.1) 0.42
9 1973–9
(from 1965–73) –0.75 –0.08(0.7) 1.05(2.5) 0.35
10 1979–86
(from 1973–9) 0.29 0.35(2.0) 0.87(2.3) 0.44
Emp/pop Market output
11 1965–73 0.34 –0.10(1.8) 0.08(0.4) 0.23
12 1973–9 –0.57 0.14(0.8) 0.48(1.3) 0.22
13 1979–86 –1.16 0.53(3.5) 0.13(0.5) 0.59
Market hours Productivity
14 1965–73 0.42 –0.27(2.9) –1.07(2.7) 0.62
15 1973–9 –0.44 –0.35(2.2) 0.00(0.0) 0.31
16 1979–86 0.46 –0.72(2.5) 0.71(1.5) 0.39
17 1973–9
(from 1965–73)
(excluding Japan) –1.01 –0.22(1.4) 1.00(2.8) 0.51
18 1979–86
(from 1973–9) 0.11 –0.17(0.8) 0.78(1.7) 0.30
TABLE 5.A6 (cont.)

Const CC
Productivity Market output
19 1965–73 0.39 1.16(10.2) 1.21(3.1) 0.90

20 1973–9 1.64 0.88(4.0) 0.36(0.7) 0.62
21 1979–86 2.07 0.28(1.0) –0.05(0.1) 0.09
22 1973–9
23 (from 1965–73) 0.09 0.80(4.3) –0.85(2.0) 0.64
(excluding Japan)
1979–86
(from 1973–9) –0.31 0.67(3.8) –0.84(2.1) 0.60
Emp/pop Market output
24 1965–73 0.49 –0.11(2.7) 0.20(1.1) 0.39
25 1973–9 0.15 –0.18(1.3) 0.73(2.0) 0.28
26 1979–86 0.10 –0.27(1.4) 0.57(1.7) 0.27
State employment
27 1965–73 0.32 –0.01(0.3) 0.64(3.8) 0.54

28 1973–9 0.18 0.07(0.8) 0.47(2.1) 0.38
29 1973–86 –0.10 0.12(2.0) 0.28(2.9) 0.53
Average hours
worked
30 1965–73 0.02 0.13(2.8) 0.51(2.6) 0.61

31 1973–9 0.28 0.11(2.1) 0.03(0.2) 0.30
32 1979–86 –0.34 0.28(3.4) –0.25(1.9) 0.54
Emp/pop Employment
spreading
33 1965–73 –0.03 –0.00(0.00) 0.00
34 1973–9 –0.36 0.15(0.5) 0.02
35 1979–86 0.01 –0.70(2.1) 0.25
Notes: Emp/pop is the ratio of employment to population of working age; Total market
hours is total hours worked in the market sector; CC is dummy variable for the five
corporatist countries; the numbers in parentheses are t-statistics; Employment spreading is
sum of State employment and (absolute value) Average hours worked.
176 ANDREW GLYN
TABLE 5.A7 Consumption regressions: 15 countries

In each equation variables are measured as average annual growth rates over period
defined except where specified as changes in growth rates as compared to previous
period.
Const CC
Consumption from earnings Productivity
36 1965–73 –0.43 0.71(6.2) –1.38(2.8) 0.77

37 1973–9 –0.13 0.51(3.3) –1.49(3.6) 0.60
38 1979–86 –1.30 0.66(1.8) 0.13(0.2) 0.22
39 1973–9
(from 1965–73) –0.36 0.67(3.2) –0.29(0.4) 0.47
40 1973–9
(from 1965–73)
(excluding Japan) –0.69 0.42(0.3) –0.49(0.7) 0.19
41 1979–86
(from 1973–9) –0.70 0.72(2.0) 1.81(2.3) 0.37
Notes: CC is dummy variable for the five social corporatist countries; the numbers in
parentheses are f-statistics.
REFERENCES
BOSWORTH, B., and RIVLIN, A. (1987) (eds.), The Swedish Economy, Washington,
DC: Brookings Institution.
Basil Blackwell.
BURDA, M. (1988), ‘Wait unemployment in Europe’, Economic Policy, 7: 393–425.
DREZE, J. (1986), ‘Work-sharing: some theory and recent European experience ,
ESPING-ANDERSEN, G. (1990) The Three Worlds of Welfare Capitalism, Cambridge:
Polity Press.
GLYN, A., HUGHES, A., LIPIETZ, A., and SINGH, A. (1990), ‘The rise and fall of the
golden age’, in S. A. Marglin and J. B. Schor (eds.), The Golden Age of Capitalism,
LUNDBERG, E. (1985), ‘The rise and fall of the Swedish model’, Journal of Economic
Literature, 23: 1–36.
MADDISON, A. (1987), ‘Growth and slowdown in advanced capitalist economies:

techniques of quantitative assessment’, Journal ofEconomic Literature, 25: 649–98.
OECD (1983), National Accounts 1964–81, Paris.
----- (1985a), The Future of Social Protection, Paris.
----- (1985b), The Role of the Public Sector, by P. Saunders and F. Klau, OECD
Economic Studies, 4 (Spring).
----- (1986), Historical Statistics 1960–84, Paris.
----- (1987), National Accounts 1973–85, Paris.
------(1988a), National Accounts 1974–86, Paris.
------(1988b), Labour Force Statistics 1975–86, Paris.
----- (1989), The Tax-Benefit Position of Production Workers 1985–88, Paris.
ROWTHORN, R. (1977), ‘Conflict, inflation and money’, Cambridge Journal of
Economics, 1: 215–39.
-----and GLYN, A. (1990), ‘The diversity of unemployment experience since 1973’,
in S. A. Marglin and J. B. Schor (eds.), The Golden Age of Capitalism, Oxford:
Oxford University Press.
------and WELLS, J. (1987), Deindustrialisation and Foreign Trade, Cambridge:
SACHS, J., and WYPLOSZ, C. (1986), ‘The economic consequences of President
Mitterrand’, Economic Policy, 2: 261–322.
6
Saving and Economic Growth from a Nordic
Perspective
Katri Kosonen
6.1 INTRODUCTION
The term ‘social corporatism’, invented by Katzenstein (1985), is used in

this volume to denote economies with centralized wage bargaining institu-
tions, a strong and interventionist state, and co-operative mood or consensus
in economic policy decision-making. A high degree of labour market solid-
arity, evidenced by low wage dispersion, high employment, and efficient pro-
tection of individuals against the loss of employment, is seen as one outcome of
social corporatism. According to all these criteria the three Nordic countries—
Sweden, Finland, and Norway—are highly socially corporatist economies; their
labour market performance has been exceptionally good in the post-war era, the
degree of centralization in their labour market organizations is relatively high,
they have low wage dispersion and an egalitarian welfare state, etc. One fun-
damental question in this volume concerns the exact functional relationship
between these factors, i.e. can corporatism really explain macro-economic per-
formance of a country (see chapters by Pekkarinen, Pohjola, and Rowthorn, in
particular, in this volume).
This chapter is not concerned with the labour market performance of the
three Nordic countries, but rather with the pattern of economic growth and
resource allocation that prevailed there in the post-war era. The emphasis is
entirely on the institutional similarities between them, although there are
also differences in areas such as political structure and the conduct of eco-
nomic policy, which are scrutinized more closely by Pekkarinen in this
volume. Particular stress is put on the central impact of the state in all the
Nordic economies in the promotion of social welfare and economic equality
on the one hand, and private investment and growth on the other. The highly
progressive income tax system, deliberately maintained in all the Nordic
countries throughout the post-war period, has made a significant contribu-
tion to the pursuit of both these policy goals. The specific sectoral saving
pattern of the Nordic economies, characterized by persistently very low
household saving and exceptionally high public saving rates, can be per-
ceived as a behavioural outcome of this double role of the state. It facilitated
the maintenance of relatively high economic growth rates up until the
ECONOMIC GROWTH: A NORDIC PERSPECTIVE 179
mid-1970s without excessive recourse to foreign borrowing. Since then,

however, more destabilizing tendencies have shown up, and in Sweden in
particular, current account and public sector deficits have grown rapidly. This
pattern of accumulation, which prevailed in all three Nordic economies in
the post-war period at least until the mid-1970s, even longer in Finland and
Norway, is referred to here as ‘the Nordic model’. It can be seen as one part
or aspect of the social corporatist structure of these economies, largely
conditioned by their social and cultural traditions.
In what follows, the specificity of the Nordic accumulation pattern is first
illustrated in the light of comparative data from the OECD countries. Altern-
ative interpretations of these empirical phenomena are also discussed. Sec-
tions 6.3 and 6.4 depict the characteristics of the Nordic welfare state and tax
systems from a comparative perspective. The nature of state intervention in
production and the financial markets is described in section 6.5, in which an
attempt is also made to analyse the functional relationship between the dif-
ferent activities of the state illustrated in earlier sections. The problems and
threats facing ‘the Nordic model’ today are briefly discussed in the last section.
6.2 INTERNATIONAL COMPARISONS OF SECTORAL

SAVING PATTERNS
In this section, comparative data from OECD countries are used to show
certain empirical similarities in the accumulation patterns of the Nordic
countries in the post-war period. Possible interpretations of these empirical
phenomena are also discussed.
Table 6.1 shows data describing the long-term saving, investment, and
growth rates in various OECD countries for the period 1960–84. All the
figures are annual averages over the whole period. These figures give rise to
a few comments, which have some bearing on what follows.
1. Household saving rates in the three Nordic countries—Sweden, Fin-
land, and Norway—were the lowest in the whole OECD area. This
phenomenon persisted throughout the whole period considered despite the
short-term volatility of household saving rates. Household saving also con-
stituted a smaller share of net national saving here than in any other OECD
country (cf. Figs. 6.1 and 6.2).
2. By contrast, the long-term average gross national saving rates (GS) in
Finland and Norway were among the highest of all the OECD countries,
and were slightly below the OECD average in Sweden. Finland and Norway
maintained a high level of national saving during the whole period, but in
Sweden the saving rate fell by almost 10 percentage points over twenty years,
which explains its relatively low long-term average rate.
3. Countries which devote a large share of their national income to saving
180 KATRI KOSONEN
TABLE 6.1 The average rates of saving, investment, and growth in various
OECD countries, 1960–1984
HS GS GI GDP/capita
% growth
USA 8.5 18.9 18.0 2.1
Japan 18.1 33.7 31.9 5.8
Germany 12.2 24.9 23.2 2.7
France 12.6 23.5 22.3 3.2
UK 6.8 18.8 18.3 1.9
Italy 19.6 22.5 20.5 3.2
Canada 8.8 21.4 22.0 2.8
Austria 9.3 27.0 26.0 3.4
Belgium 15.1 21.3 20.9 3.1
Denmark — 20.4 22.0 2.6
Finland 5.4 25.7 26.3 3.5
The Netherlands 14.1 24.9 22.8 2.4
Norway 5.0a 27.2 28.9 3.5
Portugal 14.2 24.2 25.7 4.2
Spain 9.6 22.1 20.9 3.8
Sweden 4.6 22.0 21.9 2.5
Switzerland 12.3 29.5 25.8 1.8
Total OECD 11.1 22.2 21.3 2.8
a
1974–84.
Note: HS = net household saving as % of disposable household income; GS = gross saving
as % of GDP; GI = gross fixed capital formation as % of GDP.
Source: OECD (1986).
also have, as a general rule, high rates of accumulation and growth (the
Spearman rank correlation across countries between gross national saving
(GS) and gross fixed capital formation (GI) is 0.98 in Table 6.1). Thus, on
average, the economy grew much more rapidly in Finland and Norway than
in Sweden and Denmark in the post-war period. It should be remembered,
however, that the level of GDP per capita was originally much higher in the
two latter countries than in either Finland or Norway, and this was still the
case in the 1960s. Close correlation between national saving and investment
rates has also been found in several other studies (Feldstein and Horioka
1980; Summers 1985; Gintis 1986; Dooley et al. 1987) and is generally
interpreted to be strongly at variance with the assumption of perfectly
integrated international capital markets, favoured in neo-classical macro-
economics literature.
4. On the other hand, there is almost no correlation across countries
between household saving rates and rates of capital accumulation or growth.
Correlation between the first and last columns of the table is in fact slightly
181
FIG. 6.1 Household sector: net saving divided by disposable income

Source: OECD (19856, 19876).
FIG. 6.2 Net household saving divided by net saving of the nation
Source: OECD (19856, 19876).
182 KATRI KOSONEN
negative, which is clearly at variance with the postulates of the Life Cycle
Hypothesis of Saving, as presented in Modigliani (1985), for instance.
Possible reasons for the discrepancy between our observations and those of
Modigliani et al. are discussed in Kosonen (1988).
The apparent contradiction between the last two points may be explained
by the differences in sectoral saving patterns across countries. These are
illustrated in Fig. 6.3. The sectors included are household (households +
non-profit institutions serving households), corporate (corporate + quasi-
corporate enterprises + financial institutions), and government (central and
local government + social security funds). The net saving of all these sectors
is divided by the GDP. The graphs cover the period 1960 onwards for
countries for which data were available. The figures are, however, strictly
comparable only from 1971 because of the differences in national accounting
systems before then.
It is obvious from these figures that, during the whole period, the public
sector has contributed a much larger share of national saving in the NCs than
in other OECD countries. In Norway and Finland, the high public saving
rate helped to maintain the high national saving rate up to the 1980s. In the
1960s Sweden also had a very high rate of public saving, which amounted
to almost 50 per cent of its total saving, but this situation changed rapidly in
the mid-1970s. By the end of the decade, with rapidly rising public spending.,
its whole public sector was in deficit. The public saving rate declined from
the late-1970s level in Finland, too, but nevertheless stayed well above zero
still in the early 1980s. In the other OECD countries, the sectoral distribu-
tion of saving generally displayed a very different pattern; the level of
household saving was well above that of both corporate and public saving.
The latter was, in fact, below zero in most OECD countries in the 1970s and
thus helped to decrease rather than increase the national saving rate. Cross-
country variation in corporate saving levels was much smaller than either
with respect to household or public saving and no specific country patterns
could be found.
Similar cross-country differences in saving patterns are apparent if the
financial saving of sectors is also compared. Financial saving is defined as the
difference between gross saving and gross investment and represents the
share of savings not used inside the sector itself, i.e. net lending to other
sectors within the economy. From the point of view of the supply of financial
resources, it is a more relevant concept than the aggregate saving of a sector.
Fig. 6.4 (from Soderstrom 1986) depicts the financial saving as a percentage
of GDP for the public sector, the private sector (households + corporate),
and for the country overall (current account deficit) for selected countries.
According to these graphs, the Norwegian and Finnish public sectors were
net creditors throughout the whole period, as was the case in Sweden until
the late 1970s. The private sector as a whole, on the other hand, ran up a
183
FIG. 6.3 Household sector, general government, and corporate saving divided
by GDP
Sources: OECD, National Accounts (1975, 1981, 1984, 1987) (saving rates for Norway,
Sweden, and Finland up to 1970 are based on the former SNA system).
FIG. 6.4 Financial saving by sector
Source: Soderstrom (1986).
financial deficit. Even Denmark fitted into the Nordic pattern until the
mid-1970s. The public sector financial position in the other OECD
countries was generally much worse. The latter half of the 1970s, in par-
ticular, saw budget deficits growing rapidly. However, the private sector
showed far less deficit man in the NCs, which was entirely due to the
financial surplus of their household sectors. The corporate sector was a net
debtor in all countries, the implication being that corporate saving was
generally not sufficient to finance productive investment in the private
sector. The deficit was covered partly by the current account deficit and
partly by the surplus generated in the other institutional sectors of the
economy. The public sector had mis role in the Nordic countries, and, in
general, the household sector in the other OECD countries.
There are, of course, alternative ways of explaining such structural dif-
ferences. Both graphs presented above indicate that there is an inverse
relationship across countries between public and private/household saving.
According to the so-called Ricardian Equivalence Principle, this should be
taken as an expression of close substitutability between public and private
saving, resulting from the adjustment of rational individuals to the changes
in government saving and dissaving (cf. Barro 1974). Applied to the situation
prevailing in the Nordic countries, this would imply that consumers an-
ticipate that taxes will be lower in the future and therefore save less in order
to compensate the benefit accruing to the next generation. Low household
saving would then simply be a reaction to government financial surplus,
largely maintained by high income tax rates.
The Ricardian Equivalence is based on assumptions concerning the
far-sightedness and information-processing capacity of individuals tiiat can
hardly be considered plausible. It also requires that there is no uncertainty
about future income and that no consumers are liquidity-constrained. These
assumptions are hardly realistic given the conditions prevailing, for instance,
in the Western European countries today, despite the deregulation of their
financial markets in recent years. Empirically, Ricardian Equivalence has not
generally received any strong support (see e.g. Poterba and Summers 1987).
Another problem with it is the fact that the considerable increases in budget
deficit observed since the late 1970s in most OECD countries were not
generally accompanied by any increases in household saving (Canada being
almost the only exception to this).
In an alternative explanation the causality would be the reverse. It is rather
the public sector that adjusts to the changes in private sector saving be-
haviour. If the private sector as a whole is in financial deficit, as is the case
in all the NCs, and the policies aiming at the encouragement of private saving
are not successful, external balance can be maintained only by a correspond-
ing public sector financial surplus (since the sum of financial savings of the
institutional sectors is equivalent to the current account deficit of the
186 KATRI KOSONEN
country). High public saving in countries where private domestic investment

permanently exceeds private savings should thus be seen as a deliberate
policy to avoid the recurrence of excessive foreign indebtedness. Conversely,
countries with a private sector in financial surplus are less constrained in their
fiscal policies and tend therefore to run budget deficits (see Summers 1985).
This does not mean, of course, that the countries would not have recourse to
foreign borrowing over the short or medium term, but only that foreign
indebtedness is kept within certain limits by appropriate fiscal and monetary
policies.
This view is in accordance with the views expressed, for instance, in the
papers by Summers (1985) and Gintis (1986). Gintis contrasted what he
called the national economy model with the neo-classical view on perfect
international capital markets. This model states simply that, while over the
short or medium term international borrowing may be used to finance
investments, in the long run the level of investment is limited by the
availability of domestic savings. According to Gintis, this is not only because
national governments impose impediments to the free flow of capital or
because transaction costs are associated with foreign borrowing. It is also
because national governments, for reasons that are more political than
economic, are reluctant to tolerate a large outstanding foreign debt for
sustained periods. They therefore tend to use appropriate national economic
policies to achieve external balance over the longer term (see also Summers
1985). Empirical evidence presented above (Table 6.1), as well as in the
studies by Feldstein and Horioka (1980) and Dooley et al. (1987), among
others, clearly supports this view.
The essential implication of the national economy model is that national
savings policies can be efficiently used to foster investment and growth in a
country. In contrast with Ricardian Equivalence, tax increases would be an
efficient means of increasing the national saving rate, at least if government
spending were not altered correspondingly. This requires, of course, that
individuals do not reduce their personal saving in response to tax increases
by the same amount as the increase in public saving. However, for several
reasons that have nothing to do with the ultrarational behaviour of
‘Ricardian’ consumers, the high level of income taxation may have a depress-
ive effect on the household saving rate (see section 6.4 below).
In retrospect, the experiences of the NCs in the 1960s and the 1970s are
well in accordance with the assumptions of the national economy model;
although some ‘crowding-out’ of private saving because of high taxation
probably took place, the national saving rates nevertheless remained at a
relatively high level due to high public saving, which in turn allowed high
rates of investment and growth to be maintained without excessive reliance
on foreign borrowing.
The above does not imply that foreign borrowing would not deliberately
have been used to finance investment in the NCs at certain periods. This
was the case in Finland in particular in the 1950s and 1960s. Rapid in-
dustrialization was in progress and investment rates were at a very high level
by international standards (Vartia and Vuori 1987). In Norway, foreign
borrowing was extensively used in the 1970s to finance investments needed
to exploit North Sea oil and gas fields. Both of these countries, however,
were able to keep their foreign indebtedness under control, and outstanding
foreign debt relative to GDP in fact decreased after the late 1970s. In 1982,
the Norwegian foreign debt was still 26 per cent of GDP, but a huge improve-
ment has taken place in the 1980s. By 1987, Finnish foreign debt had also
slighdy decreased from the 17 per cent of GDP it was in the early 1980s.
Sweden, on the other hand, still had a much higher public saving rate
and also a much better external balance than either Finland or Norway in
the early 1970s. It was still a net creditor in 1973 having net foreign claims
amounting to 7 per cent of its GDP (see Economic Growth in a Nordic
Perspective 1984 and Growth Policies in a Nordic Perspective 1987). After the
mid-1970s, however, its public debt started to grow rapidly, as public
spending increased at a faster rate than taxes and other government revenue.
At the same time, the trade balance deteriorated mainly because of poor
industrial export performance. By 1984, Swedish foreign debt had reached
25 per cent of GDP. In the 1980s one of the primary goals of Swedish
economic policy was to stop the accumulation of foreign debt and restore
external balance.
This section was devoted to a very general description of the sectoral saving
pattern that prevailed in the three Nordic countries—Sweden, Finland, and
Norway—in the 1960s and the 1970s. We have emphasized the importance
of the public sector in financial intermediation and the supply of financial
resources in these countries—a feature that clearly seems to distinguish them
from other Western European welfare states. In the following sections,
government intervention in the fields of social welfare and taxation is
considered. The Nordic countries share certain common characteristics here
too, which are important for the functioning of the ‘Nordic model’.
6.3 THE NORDIC WELFARE STATES: A COMPARATIVE ANALYSIS
The two fundamental goals or dimensions of the welfare state are security
and equality. By security is meant the protection of individuals against loss
of income in case of unexpected events or emergencies (accidents, sick-
ness, unemployment), or in the typical life-cycle phases of unemployment
(maternity/parenthood/childhood, old age, widowhood, etc.). The principal
instruments in pursuing this goal are the direct payment of cash benefits and
the public provision of services in kind (Flora and Heidenheimer 1981).
188 KATRI KOSONEN
Cash benefits are paid through social insurance or income maintenance

schemes, which cover the most standardized situations of need (old age,
unemployment, sickness, disability to work), or in the form of public
assistance. Access to benefits may be either universal, i.e. granted to everyone
regardless of income level, or means-tested, i.e. granted only to those below
the minimum income level judged necessary for subsistence. On the other
hand, benefits may he flat-rate, i.e. represent an equal sum of money to all,
or they may be linked to the previous income level of the beneficiary, in which
case they are said to be earnings-related (cf. Jallade 1988). Income main-
tenance schemes are financed either by general taxation, budget deficits, or
contributions paid by employers or the beneficiaries themselves.
As a policy objective, equality is a more ambiguous notion than security.
Jallade (1988) suggested that at least two different meanings should be
distinguished:
1. equality of opportunity or access to welfare benefits and services;
2. the degree of equalization actually achieved in, for instance, income
distribution (see Jallade 1988: 5).
These two notions of equality are not always congruent; social programmes
that are based on the egalitarian principles in the meaning of the first
definition do not necessarily lead to any reduction in income inequality or
well-being. An example of this is the public welfare services, which may, in
fact, benefit more high- and middle-income households than low-income
households. It is also clear that, as policy goals, security and equality in the
second definition are not entirely complementary, but may contain con-
tradictory elements. For example, the adoption of the earnings-related
principle in social insurance considerably increases the adequacy of benefits
and therefore leads to higher protection and security. At the same time,
however, it may merely replicate the existing structure of earnings and thus
have no redistributive effect.
On the other hand, high redistributive potential is associated with different
ways of financing the welfare state. In particular, progressive income taxa-
tion, while not necessarily egalitarian in the first meaning, is shown to have
a strongly equalizing impact on the post-tax distribution of income (see, for
instance, Sawyer 1982). Various kinds of tax reliefs and deductions, which
generally benefit high-income groups most, could, however, somewhat lessen
this effect. The incidence of indirect taxation (taxes on goods and services)
is also generally considered to be regressive rather than progressive, and may
therefore counteract the equalizing impact of progressive income taxation.
The Nordic countries are often described as ‘institutional welfare states’
with universal and egalitarian income maintenance schemes and extensive
public services financed largely by general taxation (see Alestalo et al 1985).
This should be contrasted with ‘the marginal welfare state’, in which social
security is directed only to those judged to be in need, i.e. it is mainly based

on means-tested benefits and social assistance. In this case the level of public
social expenditure as a whole is low. It has also been suggested that ‘the
Nordic welfare state tradition’—typical of Sweden, Finland, and Norway—
differs in certain basic institutional characteristics from the two other
Western European welfare state traditions, namely ‘the Bismarck tradition’
found, for instance, in Germany, Austria, France, and Italy, and ‘the
Beveridge tradition’ of the UK and Ireland (Jallade 1988). This section
contains a brief, comparative overview of the historical development of the
Nordic welfare states, followed by an analysis of the basic similarities and the
most important differences between them and the other Western, industrial-
ized countries at the present time.
The historical development of the welfare state in the Nordic countries
has not been at all uniform. Denmark was clearly a forerunner and had
already established fairly extensive social insurance schemes along
Bismarckian lines in the early twentieth century. Sweden too had developed
a relatively advanced social security system by the 1930s, largely due to the
strong influence of the Social Democratic Party. In Norway and Finland,
developments were much slower. Indeed, in Finland only a tiny proportion
of the population was covered by any social security system during the
inter-war period. As a whole, none of the Nordic countries, except perhaps
Denmark, was in any sense among the leading welfare states at this time
(Kosonen 1987: 195). It was only in the post-war period that a major
expansion took place which affected the whole area and was then quite rapid
compared with the other Western European countries.
One way to illustrate the development of the welfare state in a country is
index numbers describing the coverage of social security. Alestalo et al. (1985:
193) present such index numbers concerning Western European countries
during the period 1920–75. They indicate the proportion of population
actually covered by major social insurance schemes (old age pensions,
sickness, unemployment, and occupational injuries insurance) during the
time period in question. According to this criterion, there was considerable
diversity in the development of the welfare state in Western Europe up until
the 1960s. The leading countries in the 1950s were Denmark, Norway,
Sweden, the UK, and Switzerland. Several countries caught up in the 1960s
and many actually passed the level reached earlier by these leading nations.
In particular, the position of the UK and Switzerland deteriorated in relative
terms. The retarded development of the welfare state in Finland becomes
evident here; the major extension of the social security system only took place
in the 1960s and it was not until 1965 that Finland reached the same
standard in terms of social security coverage that Denmark was already
enjoying in the 1920s. The welfare state expanded rapidly in Finland after
that time and today its coverage exceeds the Western European average.
190 KATRI KOSONEN
The level of social expenditure (relative to GDP) is another important

indicator of the extension of the welfare state in a country. It reflects the
coverage of social programmes as well as the average level of social benefits
and the size of population concerned. Fig. 6.5 depicts the evolution of social
expenditure in the OECD countries during the period 1971–81 (calculated
from OECD 1985a). Social expenditure here includes public expenditure
on education, health, pensions, and unemployment compensations and
other social expenditure (sickness, maternity, and temporary disablement
benefit, family and child allowances, etc.). It paints quite a good picture of
the general national levels of social services and social security.
These figures also show how rapid the growth of the welfare state was in
Western industrialized countries during the post-war period, although a
slow-down at the end can be clearly perceived. Since the late 1970s, the ‘size’
of the welfare state has been largest in Belgium, The Netherlands, and
Sweden. Standards reached in the ‘Bismarckian’ countries the FRG, France,
Italy, and Austria in the 1980s also exceeded those achieved in Norway and
Finland. According to these figures, the Anglo-Saxon (‘Beveridgian’) coun-
tries have a more modest level of social expenditure than either Nordic or
Continental European countries.
In the light of this evidence the Nordic countries generally seem to rank
higher in terms of social security coverage than in terms of level of expend-
iture. This phenomenon reflects in a way the universalist and solidaristic
tendencies of the Nordic welfare states. It also implies that Nordic citizens
actually covered by social security receive perhaps somewhat lower benefits
than citizens of the Continental European countries, where total expenditure
is higher but the proportion of the population actually covered is slightly
lower.
This is particularly true of Finland, where the average level of welfare
benefits today still lags behind the three other Nordic countries and, in all
probability, the other advanced welfare states of Western Europe as well. For
instance, a comparison of the social insurance schemes in operation in the
four Nordic countries in the early 1980s reveals that compensation levels
(benefits received as a percentage of net income) varied between 70 and 100
per cent in Sweden, Norway, and Denmark, but were generally 20–50
percentage points lower in Finland. The biggest gaps were in unemployment
and sickness benefits and child allowances (Social Security in the Nordic
Countries 1984). The average level of pension benefits has substantially
increased in Finland following the establishment of an occupational pension
insurance system in the early 1960s. Today it exceeds the OECD average,
although it is still lower than in Sweden and several Continental European
countries (France, Italy, The Netherlands, Austria, and Belgium) (Allen
1988).
One important conclusion to be drawn from this comparative evidence is
FIG. 6.5 Total social expenditure divided by GDP, 1971–1981

Source: OECD (1985a).
192 KATRI KOSONEN
that there is no homogeneous Nordic pattern regarding the level of social

security relative to the advanced welfare states of Continental Europe.
Continental European household saving rates, however, are generally much
higher than those of the Nordic countries (cf. Fig. 6.1). This seems to cast
serious doubt on the relevance of ‘the asset substitution hypothesis’ (cf.
Feldstein 1974) or ‘the precautionary savings hypothesis’ (cf. e.g. Kotlikoff
1986 and Skinner 1987) in explaining the low long-term level of Nordic
household saving and, for that matter, on the importance of the life-cycle or
precautionary motives for household saving behaviour in general (see also
Kosonen 1988).
This conclusion does not imply, however, that the notion of the Nordic
welfare state as a distinct Category would be without any empirical ground.
A closer look at their institutional similarities and at how they indeed differ
from other Western societies follows.
The social security system which operated in all the Nordic countries from
the post-war period up to the 1960s had a highly egalitarian character. The
social insurance schemes were compulsory and universal with mostly flat-
rate benefits (Esping-Andersen and Korpi 1984). The financing of the
welfare state relied to a large extent on general tax revenues collected mostly
through progressive income taxation. In the 1960s, however, this system
went through a profound transformation. The most important change was
the adoption of a supplementary, earnings-related pension insurance system,
which took place in all the Nordic countries early in the decade. This resulted
in a two-tier system:
1. the basic pension, which is flat-rate and granted to all regardless of
income level, and which also includes means-tested supplements;
2. the supplementary (or, in Finland, occupational) pension, which is
calculated on the basis of past earnings, financed largely by employers’
(in Finland and Norway also to a small extent by employees’) contribu-
tions and funded (Social Security in the Nordic Countries 1984).
The pension funds thus created remain under government control in
Sweden and Norway. In Sweden, this solution represented a victory for the
Social Democratic Party after a fierce political struggle. A similar system was
adopted somewhat later in a more harmonious political atmosphere in
Norway (Kosonen 1987). These funds made a significant contribution to
the increase in public saving in the 1960s in both countries, and also offered
the governments an efficient means of financing private investment. In
Finland, the occupational pension insurance system was established in 1960
as a result of close co-operation between both workers’ and employers’
labour market organizations in 1960: this event could in fact be considered
as the first step towards a corporatist wage-bargaining system in Finnish
society (Niemela 1988). The pension funds are not integrated into public
administration as in Sweden and Norway, but are run mainly by private

insurance companies. The enterprises which finance these funds are granted
automatic re-lending amounting to two-thirds of the contributions paid at
an interest rate somewhat lower than the prevailing market rate (Niemela
1988).
By the 1970s the earnings-related principle was also extended to un-
employment, sickness, and industrial injury insurance {Social Security in the
Nordic Countries 1984). All these reforms aimed at (and largely succeeded
in) improving the adequacy of welfare benefits of the wage-earning popula-
tion relative to the more privileged civil servants and private salaried
employees, who already had an earnings-related pensions insurance system
(Esping-Andersen and Korpi 1984). Only after these reforms did the Nordic
countries really step into the front rank among the Western welfare states.
On the other hand, these developments could also be seen as a move towards
the ‘Bismarckian’ social insurance system and a step away from the
egalitarian principles on which the Nordic welfare state was originally based.
In the words of Esping-Andersen and Korpi (1984: 21): ‘The honoured ideal
of equal benefits to all gives way to the ideas that citizens should be
guaranteed against income loss, and that benefits should be closely related
to accustomed earnings or income, ideas which already to some extent had
been realised in connection with reforms of sickness insurance.’ The inequal-
izing impact resulting from the adoption of earnings-graduated provisions in
principle was, however, moderated in all the Nordic countries by income-
tested supplementary benefits designed for low-income groups and those
not eligible for occupational pension schemes (Esping-Andersen and Korpi
1984). It seems that the combination of flat-rate and earnings-related ele-
ments, means-tested supplements, and the progressive taxation of benefits
has made the Nordic pension insurance systems as a whole rather progress-
ive. According to a comparative study by Uusitalo, the equality of income
distribution among the elderly, particularly in Sweden but also in Finland
and Norway, is relatively high by international standards (Uusitalo 1988).
The institutional development of the welfare state was thus very similar in
all the four Nordic countries during the post-war period. The social security
systems were developing along much the same lines in other Western
European countries. By the 1970s compulsory, universal pension schemes
consisting of flat-rate and earnings-related parts as well as earnings-related
unemployment, sickness, and industrial injury insurance had been almost
universally adopted. Consequently, despite minor differences ‘impressive
similarity’ characterized the social security system ‘in use in the late 1970’s
in Europe’ (Sawyer 1982: 220).
A few important differences still remain and they emphasize the more
institutional nature of the Nordic welfare states compared with the
Continental European (‘Bismarckian’) countries. One has already been
194 KATRI KOSONEN
mentioned—the more universal coverage and liberal eligibility conditions of

social security in the Nordic countries (see also Esping-Andersen and Korpi
1984). There are two others which should be emphasized.
First, in the Nordic welfare state tradition, a relatively large share of bene-
fits is provided in the form of services in kind, while the Bismarck tradition
stresses more direct cash transfers (Jallade 1988). One indication of this is
that the level of social security transfers (relative to GDP) is lower in the
Nordic countries dian in The Netherlands, Belgium, and all the typical
‘Bismarckian’ countries, which have a higher total level of social expenditure
including public expenditure on healtii and education. The comprehensive
school system and extensive public health and social services are, in fact,
often considered hallmarks of the Nordic welfare state. In the 1970s major
reforms were carried out in these areas, and as a consequence the public
sector as a whole expanded rapidly. The importance of the public sector as
employers also grew. The rate of expansion was more modest in Finland,
however, and the availability of public services still lags behind in certain
respects (Allen 1986). Day care services for small children in particular are
more scarce in Finland than in the three other countries (Alestalo et al.
1985).
The second difference concerns the ways of financing the welfare state.
Government tax revenue has played a dominant role in the Nordic welfare
state tradition and still contributed more than 50 per cent of total finance in
the 1960s. However, employers’ contributions increased substantially fol-
lowing the establishment of supplementary pension insurance systems. By
1981, the share of the state in the total financing was highest in Denmark
(88 per cent) and lowest in Norway (42 per cent). Employees’ social security
contributions made up only a small part of the total financing (although the
figure is 20 per cent in Norway, it is much less in the three other countries)
(Allen 1988). Here again the contrast with the Continental European
countries is clear. In accordance with the Bismarck tradition they rely to a
much larger extent on the contributions of the beneficiaries themselves (cf.
Tables 6.2 and 6.3 in the next section). Personal income taxes are generally
progressive, while social security contribution schemes are either proportion-
al or regressive (Sawyer 1982). Thus the Nordic tax system as a whole is
more progressive than that of the Continental European countries.
This last point is relevant to the second fundamental dimension of the
welfare state, equality. Inter-country comparisons on this level are more com-
plex than regarding achievements in the area of security. A few comparative
studies in which income distributions are compared indicate, however, that
the redistribution achieved through the welfare state is particularly notice-
able in the Nordic countries. In a recent study, Uusitalo (1988) presented
data on the distribution of disposable income (factorial income + income
transfers- taxes) in seven countries in the early 1980s: three Nordic
countries—Sweden, Finland, and Norway—clearly stood out as having the

highest levels of equality. (For instance, Gini coefficients for the seven
countries are: Sweden 20.5, Finland 22.5, Norway 24.3, UK 27.3, Canada
29.9, USA 32.6, and Israel 33.3.) According to the estimates by Uusitalo, the
redistributive effect of income transfers was in fact more important than that
of taxes in the Nordic countries. This implies that the Nordic welfare state
has preserved its egalitarian character very well. Uusitalo also concluded that
the basic explanation for the high equality achieved in the Nordic countries
concerns the size of the welfare state as such; his data show a clear positive
correlation between these two variables.
Unfortunately the reference group in his study only includes countries
with a ‘smaller’ and more marginal welfare state than the Nordic one. In
particular he did not include any of the advanced welfare states of Con-
tinental Europe with social expenditure shares and average benefit levels at
least as good as, if not better than, the Nordic countries. In an earlier study,
Sawyer (1982) also presented the data on the distribution of post-tax income
in nine Western European countries. Here, too, the Nordic countries turned
out to have the lowest levels of inequality: as an example, the percentage of
post-tax income received by the two top income deciles was lowest in
Sweden, Norway, and Finland, and highest in the ‘Bismarckian’ countries
France, Germany, and Italy (37, 37.3, and 38.5 vs. 46.9, 46.1, and 43.9
respectively). These differences must be largely due to the higher progress-
ivity of income taxation in the Nordic countries. It should be added, how-
ever, that the equalizing impact of progressive income taxation could be at
least partly offset by tax reliefs and deductions, which are extensive by inter-
national standards in all the Nordic countries (for example Nielsen 1988).
All the available evidence thus suggests that the degree of income equality
is very high in the Nordic countries compared with almost any Western,
industrialized country and that this is to a large extent due to the redistribut-
ive impact of the welfare state. In this regard, it is interesting to note that in
Finland, where the major expansion of the welfare state took place in the
1960s and the 1970s, its redistributive effects have continuously become
stronger (Uusitalo 1988). During the period 1966–76 some equalization of
factorial incomes also took place. According to Uusitalo this was largely due
to the solidaristic wage policies of that period. The equalization achieved
through income transfers and taxes was, however, much more important and
also continued in the period 1976–81, when wage dispersion started to grow
again. This Finnish evidence suggests that the equalization achieved
through the welfare state is much more significant than what is achieved
through solidaristic wage policies. Unfortunately, Uusitalo made no such
comparisons regarding the other Nordic countries (i.e. the equalization of
factorial vs. disposable income), but it may be assumed that the redistributive
effects of the welfare state in those countries are at least as strong as, if not
196 KATRI KOSONEN
stronger than, in Finland. The high level of labour market solidarity, em-
phasized in particular by Bob Rowthorn in this volume, is thus also con-
firmed by this evidence. The efficient and egalitarian welfare state is indeed
an essential functional part of social corporatist economies.
In conclusion, the distinctiveness of the Nordic welfare states is much
more apparent along the equality dimension than along the security dimen-
sion. The most apparent example of this is the equal distribution of
household disposable income. Another specific Nordic feature is the import-
ance of state intervention in the provision of welfare services and their
extensive allocation among the whole population.
6.4 TAXATION
It was pointed out in the previous section that one of the most conspicuous
differences between the Nordic and Continental European welfare states
today is in the area of financing welfare expenditure. This section considers
the Nordic tax systems and their possible effects on household saving rates
in more detail. Some of the most important differences are described in
Tables 6.2 and 6.3, illustrating the structure of tax revenue and the level of
taxation in OECD countries in the 1980s.
Total tax revenue as a percentage of GDP is the highest of all OECD
countries in Sweden, Norway, and Denmark. In Finland, however, the gross
tax rate is near the OECD average, reflecting the lower level of social
expenditure (Table 6.2, col. 1). Finland, on the other hand, like Sweden and
Denmark, collects a comparatively large share of its taxes directly from
households in the form of personal income tax (Table 6.2, col. 2), while the
share of taxes paid directly by corporations is small (Table 6.2, col. 3).
(Norway deviates from this Nordic pattern mainly because of the tax revenue
received by the government from oil companies.) Social security contribu-
tions in the Nordic countries are generally lower than in the other OECD
countries (col. 4), while taxes on goods and services constitute a relatively
important source of government tax revenues (col. 5).
The most conspicuous differences between the Nordic and other OECD
countries are found, however, in the level of personal income taxation. This
is illustrated by the figures in column 1 of Table 6.3. Personal income tax
rates in Finland, Denmark, and Sweden, measured at the income level of an
average productive worker, are the highest of all the OECD countries, only
Norway having a somewhat lower figure. The effect of high direct income
tax on disposable income is partly compensated by the generally lower level
of social security contributions paid by employees themselves (Table 6.3,
col. 2) compared with other Western European economies. (The bulk of
social security contributions are paid by employers in Norway and Sweden,
TABLE 6.2 The rate and the structure of taxation in various

OECD countries in 1985
Total tax Personal Corporate Social Tax on
revenue/ tax direct security goods and
GDP tax tax services
(1) (2) (3) (4) (5)
Finland 37.1 46.3 4.4 9.0 36.9
Denmark 49.4 51.6 5.0 3.8 34.5
Norway 47.8 22.5 17.0 20.7 37.6
Sweden 50.6 38.4 3.2 25.1 26.3
USAa 29.0 35.3 7.1 29.1 18.2
Japan3 27.4 24.5 21.1 29.7 15.1
Germany 38 28.6 6.0 36.7 25.6
France 45.6 12.8 4.3 43.7 29.4
UK 38.6 26.5 12.2 18.0 31.1
Italy3 41.2 26.3 9.8 33.9 26.1
Canada 34.2 34.1 9.8 12.9 32.5
Austria 42.3 23.1 3.2 31.7 32.6
Belgium 46.7 34.7 6.2 32.3 24.9
Greece3 35.2 14.6 2.6 35.2 43.4
The Netherlands 44.8 20.1 6.8 43.7 25.7
Switzerland 32.0 34.7 5.9 32.1 19.0
3
1984
Ncnes: Personal tax = the share of personal income tax in total tax revenue; Corporate direct
tax = the share of corporate direct tax in total tax revenue; Social security tax = the share of
social security tax in total tax revenue; Tax on goods and services = the share of tax on goods
and services in total tax revenue.
Source: Hagemann, et al. (1987).
which explains the figures in col. 4 of Table 6.2.) Even taking this into
account, the total tax burden of the household sector in Denmark, Sweden,
and Finland is one of the highest in the OECD area and it is also above the
average in Norway (see OECD 1987).
The high progressivity of Nordic income taxation was emphasized in the
previous section. This is evident in the above tables in the relatively high
marginal tax rates on average and top wages (Table 6.3, cols. 3 and 4
respectively), although these figures of course do not display the progressivity
of the whole tax schedule. Direct income distribution comparison gives more
reliable information in this respect (see above and also Sawyer 1982).
The third feature that needs to be emphasized is the importance of tax
deductions. The tax-deductibility of interest on loans is exceptionally liberal
in all the Nordic countries (see Hagemann et al. 1987 and Skjaeveland and
Stokstad 1988). In Sweden, Norway, and Denmark, interest on both hous-
ing loans and consumer credit is totally deductible. This is subject to a ceiling
198 KATRI KOSONEN
TABLE 6.3 Average and marginal income tax rates in various OECD
countries in 1986
Personal Social Marginal Top marginal
income security tax rate tax rate
tax contributions
(1) (2) (3) (4)
Finland 30.3 3.4 53.2 68
Denmark 40.4 6.3 62.4 68
Norway 22.7 11.1 60.1 56
Sweden 36.1 6.0 62.0 77
USA 19.9 7.2 40.9 38
Japan 9.2 7.0 31.9 88
Germany 18.1 17.2 62.7 56
France 7.2 15.8 51.2 58
UK 21.6 9.0 43.9 60
Italy 16.8 8.8 57.8 62
Canada 20.0 4.0 34.1 52
Austria 10.8 16.4 54.5 62
Belgium 24.4 12.1 63.2 87
Greece 5.1 13.3 40.1 63
The Netherlands 11.5 26.2 61.9 72
Switzerland 11.4 10.3 39.4 46
Notes: Personal income tax = personal income tax as a percentage of gross earnings at the
income level of an average productive worker (APW), single people, year 1986; Social
security contributions = employees’ social security contributions as a percentage of gross
earnings at the income level of an APW, single people, year 1986; Marginal tax rate =
marginal tax rate on average wage under the present tax system, year 1986; Top marginal tax
rate = top marginal tax rate under the present tax system, year 1986.
Sources: OECD (1987a); Hagemann et al. (1987).
in Finland (FIM 25,000 and FIM 10,000 respectively in 1988). Only a few
other countries (The Netherlands, Switzerland) allow unlimited, or any,
deductions on consumer credits, and deductions on housing loans are also
often restricted in various ways. The main effect of this, combined with high
marginal tax rates, is a decrease in the real cost of household borrowing. In
fact, according to estimates made by Skjaeveland and Stokstad (1988), the
real after-tax interest rate on both mortgage and consumer credit was, on
average, negative over the period 1975–85 in all the Nordic countries, and
lower than in any other OECD country included in their calculations. On
the other hand, the real rates of return on financial assets were generally low
due partly to the progressive taxation of property income (not in Finland),
and partly to low bank deposit rates and regulated financial markets in
general. This obviously constitutes a powerful incentive for borrowing and
disincentive for saving in Nordic households.
These considerations suggest that the low household saving rates in the
Nordic countries are a result of their particular tax systems rather than the
social security systems or income growth rates. A highly progressive income
taxation typical to the Nordic countries could in fact have a depressive effect
on the aggregate household saving rate for at least three different reasons:
(1) because it reduces the wedge between after-tax borrowing and lending
rates, as indicated above, (2) because it equalizes after-tax distribution of
income, and (3) because it reduces unexpected variation in income. The
second reason prevails if the saving rate tends to rise with a rise in household
income, as the Keynesian theory suggests. The third explanation in turn
applies if the propensity to save out of unexpected income change is high, as
the Permanent Income Hypothesis assumes.
The hypothesis that progressive income taxation as such has a significant
negative impact on the household saving rate is also supported by empirical
evidence. In cross-country saving equation regressions run on data from
eighteen OECD countries (including the three Nordic countries), the
average personal tax rate was by far the most powerful single explanatory
variable and had a significantly negative coefficient in all model specifica-
tions. Our interpretation of this is that the personal income tax rate variable
is in these regressions a proxy to the progressivity of the tax system as a whole
in the country. (The model specifications and possible interpretations are
explained in more detail in Kosonen 1988.) Empirical distinction of the three
alternative explanations of the apparent negative correlation between the
income tax rate and the household saving rate is unfortunately not possible
on the basis of these regressions.
6.5 NORDIC GROWTH POLICIES
In this section, an attempt is made to develop a more comprehensive view

of the growth process of the three Nordic countries—Sweden, Finland, and
Norway—in the post-war period. A brief overview of the growth policies
practised is presented first in order to produce a fuller picture of the role the
state has played in the accumulation processes. The section ends with a
‘pulling together of strings’ and a look at how the institutional factors
analysed in this chapter have contributed to the functioning of the ‘Nordic
model of accumulation’.
One important function of the state in the NCs right from the early phase
of industrialization has been to support, but also direct and control, the
process of accumulation. This was true in the pre-war period, when they
were still in the stage of transition from agrarian to industrial societies —Sweden
being clearly a forerunner with respect to both Finland and Norway. In the
post-war period, too, the promotion of investment and growth was highly
ranked as a target of economic policy in all the three NCs. The priority given
200 KATRI KOSONEN
to policy targets varied in the countries according to the individual national

economic policy model. The ‘Swedish model’ clearly gave highest priority
to full employment and stabilization, but ‘equality’ (welfare and solidaristic
wage policy) also ranked highly and accumulation was considered com-
plementary to these. In Norway and Finland, on the other hand, investment
and growth were given absolute priority in the formulation of economic
policy. This involved more direct state interventionist forms in Norway than
in Finland, where, in turn, the competitiveness of export industry was more
emphasized at the expense of employment and welfare (the Nordic national
economic policy models were analysed in detail in Mjoset 1986; Kosonen
1987; see also Pekkarinen in this volume).
Although direct state intervention in the allocation of productive capital
was important both in Finland and Norway after the war, the bulk of
industrial production remained the responsibility of private enterprises. A
crucial role of public investments was above all to strengthen the industrial
base and create favourable conditions for the accumulation of private capital.
State-owned enterprises were established in both countries in certain basic
industrial sectors, where the magnitude of capital input needed was large and
investments were therefore too risky for private capital. In Finland, these
included basic metal and chemical industries, and to some extent forestry,
which traditionally had been the main export sector (see Vartia and Vuori
1987). Energy production in Finland is also largely dominated by public
enterprises. In Norway, state enterprises were established immediately after
the war in the traditional export sectors, mining and metallurgy, which had
earlier been dominated by foreign capital (Mjoset 1986 and Kosonen 1987).
Later on, in the 1960s and the 1970s, the exploitation of North Sea oil and
gas fields was carried out with close collaboration between the state and
multinational oil companies. The companies, which are totally or partly
state-owned, now dominate Norwegian petroleum production; this has
become Norway’s most important export commodity (and the most import-
ant source of tax revenues for the government). Despite the importance of
state-owned companies in a few key industrial sectors, their share of total
employment or industrial production has never reached a high level in either
country. In the early 1980s, state-owned companies contributed less than 10
per cent of total industry value-added in Norway and about 18 per cent in
Finland (see Vartia and Vuori 1987 and Economic Growth in a Nordic
Perspective 1984).
In Sweden, direct state intervention in production and the allocation of
investments has been far less important than in the two other NCs. There
are no big state enterprises in the Swedish export industries, which are
dominated by strong and highly centralized national capital. The level of
public investment relative to GDP in the 1960s was somewhat higher in
Sweden than in either Finland or Norway, but this was largely due to her
housing construction programme, which was financed to a great extent by

public funds. In the 1970s public investments remained in the range of 4–5
per cent of GDP in all the three NCs (see Kosonen 1987).
However, growth policies were not limited to the activities of the public
sector alone. The financing of private investments was greatly influenced in all
the three countries by the state both directly by loans, grants, and subsidies,
and indirectly by the regulation of financial markets.
The quantitative importance of subsidies is rather difficult to estimate. If
public expenditure figures are compared, direct transfers to producers (the
bulk of which are subsidies) seem to have been at a particularly high level in
Finland and Norway in the post-war period (see Kosonen 1987: 198).
Subsidies in Finland and Norway, however, were not used principally to
promote industrialization, but to support the population still engaged in
primary production (agriculture and fishing), which at least partly explains
the difference in figures with respect to Sweden. On the other hand,
according to estimates made by Carlsson (1984), Sweden and Norway have
had industrial subsidy programmes since the late 1970s which are substantial
by international standards and much larger than in Finland. The growing
importance of subsidies, at least in Sweden, may rather be seen as an effort
to maintain full employment in the worsening economic climate than as an
element of traditional growth policy. The role of subsidies in the Nordic
countries is more closely analysed by Landesmann elsewhere in this volume.
Besides direct financial support for investments, the Nordic governments’
monetary policies, and in fact the functioning of the whole financial system,
have been rather strongly orientated towards financing productive invest-
ments.
All the three NCs practised policies of credit rationing combined with low
interest rates up to the end of the 1970s. In Norway the state directly
controlled the lending volume of the state banks, although the activities of
the private banks were seen as an integral part of the state-regulated credit
system. In Finland and Sweden, the central banks had a strong influence on
the credit market. Private banks were indebted to the CB, which could
control the outstanding volume of bank credit by regulating the cost of
central bank borrowing. Quantitative regulations (liquidity and cash quotas,
lending limits, and investment quotas imposed on other financial institu-
tions) were also extensively used in Sweden (Horvei et al. 1987: 264). In
both countries the CBs also made informal agreements with private banks,
the purpose of which was to control the total amount of credit and to
influence its allocation in favour of productive investments. Despite general
credit rationing, large enterprises were granted priority access to bank credits.
In Finland, in particular, big corporations operating in the export sector were
strongly favoured in the banks’ lending policy (cf. Vartia and Vuori 1987).
The interest rates on bank credits were tightly controlled by the state (or
202 KATRI KOSONEN
the CB) in Finland and Norway up to the 1980s. (In Finland, control of
average bank lending rates was relaxed only in 1986.) Sweden has had more
flexible interest rates since 1965, but within the limits set by the system of
credit rationing (see Mjoset 1986). Since nominal interest rates have been
more or less fixed, real interest rates have been at very low levels in periods
of accelerated inflation. This was the case in the 1970s in particular, when
real interest rates were negative for several years (Economic Growth in a Nordic
Perspective 1984). The combination of low interest rates and the tax-
deductibility of interest payments made bank credit a very cheap source of
finance for enterprises. The central banks’ tight control of capital flows made
it possible to maintain the gap between domestic and foreign interest rates.
In conclusion, although the bulk of production and investment decisions
in the NCs were left to private enterprises operating freely on the market,
almost all forms of financing investment were strongly regulated by the state.
In Norway, where regulation took the most extreme forms, the term ‘credit
socialism’ has been used to describe the system (cf. MJ0set 1986). This may
be too much of an exaggeration for Finland and Sweden, although state
influence on the conditions of finance was nevertheless very important.
Selective credit rationing and the regulation of interest rates provided
enterprises with cheap funds to finance their investments. Corporate
retained earnings therefore probably played in the NCs a less important role
as a source of finance than they are normally thought to play in capitalist
economies (cf. e.g. Glyn etal. 1986). This does not mean of course that profit
expectations were not an important determinant of investment decisions also
for the Nordic enterprises.
From this perspective, the government welfare and tax policies have made
an important contribution to macro-economic equilibrium in the following
sense. The high general level of taxation which has been maintained in all
the NCs throughout the whole post-war period can be seen as a deliberate
choice of policy motivated by the desire to keep up the high rate of capital
accumulation. Part of wage increases won by workers at the bargaining table
(the corporatist wage formation process practised in the NCs is described in
other studies of this volume) was taxed away through progressive income
taxation, which contributed to increase the public sector financial surplus.
This in turn helped to finance productive investment by a corporate sector
in financial deficit. As A. Glyn (1986: 7) pointed out: ‘If capitalists’ access
to credits frees their investments from a tight dependence on declining
profits, it can only do so by taking away from workers what they gained or
thought they gained at the bargaining table.’ This point assumes particular
relevance in the light of the steadily rising share of wages in value-added in
all the three Nordic countries from the 1950s until the mid-1970s (see
Mjoset 1986). This happened despite the fact that corporatist incomes
policies were used at least partly to moderate wage increases, and exchange
rate policy was actively and deliberately used to change the functional
distribution of income in favour of profits and to maintain the profitability
of export industries.
This policy could be called forced saving through taxation adjustment (cf.
Taylor 1983). In economies which operate at near to full capacity and where
excessive foreign indebtedness is not considered feasible politically or other-
wise, a high public saving rate is almost the only viable way of closing the
saving-investment gap (see also e.g. Kalecki 1970). This is in fact what
Kaldor recommended for ‘backward countries’, where it is impossible to
mobilize household saving on a voluntary basis; in such conditions, accord-
ing to him, the advancement of capital accumulation by fiscal means ‘con-
stitutes the only alternative to inflation, which is a bad and inefficient way to
mobilize resources’ (quoted in Benoit 1985: 278). In the light of the analysis
presented in this chapter, this policy alternative also seems to apply to mature
capitalist economies with a high degree of social corporatism, such as the
Nordic countries; failure to mobilize private saving is not due to ‘back-
wardness’, but rather to a highly developed welfare state which has institu-
tionalized certain important functions of household saving and also to the
tax system, which strongly favours borrowing at the expense of saving for
both households and enterprises. In such economies, the high public saving
rate achieved by fiscal means allows for a higher non-inflationary growth rate
to be maintained (see also Kurihara 1970). On the other hand, to be
successful, such policy requires a strong and ‘hegemonic’ state power, which
was also an essential part of the social corporatism of the Nordic economies
in the post-war period (see also Pekkarinen in this volume).
Finally, it should be emphasized that the Nordic governments did not
use the tax revenues collected from households only to promote capital
accumulation. They also aimed to increase the welfare and equality of their
citizens, as mentioned previously. These policies, which were highly valued
as such in terms of the national model of each country, contributed to the
enlargement of the mass consumption base. In fact, steadly increasing
domestic consumption constituted one important factor in the economic
growth of all the Nordic countries in the post-war period (see also Mjoset
1986). On the other hand, by considerably decreasing the economic in-
equality between the citizens of each country, they also helped to create
political conditions, an ‘atmosphere of consensus or solidarity’, which
allowed the state to maintain high average and marginal tax rates without
too strong popular resistance. As Kaldor (1955: 184) pointed out in his
discussion of the efficiency of British tax policy: ‘With greater equality in
the standards of living, the power of the Government to retard consumption
through taxation would be considerably greater than it is now . . . people
are far more ready to accept sacrifices in a national cause when they feel
the burden tfiey are asked to bear is fairly distributed’ (cf. also Glyn 1986).
204 KATRI KOSONEN
6.6 PRESENT TRENDS: IS THE NMA FALLING APART?
In this final section some problems and contradictions facing the Nordic
economies today are considered. In a way the same institutional charac-
teristics which made an essential contribution to the success of the ‘Nordic
model’ in the post-war period also seem to have potential to assist in its
dissolution. Two trends in particular are worth mentioning.
First, it has become increasingly difficult to maintain the high saving rate
and financial surplus of the public sector. Second, in the early 1980s all the
NCs went through a profound transformation of their financial market
conditions. This radically decreased government ability to control the supply
and allocation of credit. Both of these factors have in turn helped to increase
the instability of the Nordic economies and aggravated the problem of
external balance. These problems are particularly pronounced in Sweden
and Denmark, but the same trends are also visible in Finland and Norway.
The first of these destabilizing tendencies appeared in the latter half of the
1970s, when the financial position of the public sector started to deteriorate
in all the NCs. The public sector financial surplus, which was still, at best,
3–5 per cent of GDP in the early 1970s, shrank and turned to deficit towards
the end of the decade. Deficits grew continuously in Sweden and Denmark
until the early 1980s, and by 1985 gross state debt had reached a level of
more than 70 per cent of GDP in Denmark and 60 per cent in Sweden. In
Finland and Norway, however, public deficit was a more temporary phenom-
enon; Finnish public accounts were in relatively good balance (small deficits
and surpluses) in the first half of the 1980s and the Norwegian public sector
in fact developed huge surpluses during the same period mainly because of
high oil prices (see Growth Policies in a Nordic Perspective 1987). In the latter
half of the 1980s, the balance of public accounts was steadily improving in
Sweden and Denmark, while in Finland and Norway opposite tendencies
were again apparent.
There are various possible reasons for these developments. First, the
commitment of the Nordic governments to the development of the welfare
state contributed to the growth of public expenditure. This continued at a
relatively high level throughout the 1970s in all the NCs, despite deteriorat-
ing economic conditions towards the end of the decade. It was particularly
rapid in Sweden and Denmark partly because social reforms introduced
during the 1960s and the 1970s demanded increasing amounts of financial
resources, but also because general price and wage increases swelled the
expenses of the public sector (Allen 1986). In addition, full-employment
policies, practised above all in Sweden, required a considerable financial
effort from the state.
On the other hand, progressive income tax schedules were not adjusted to
the high inflation rates of the 1970s and consequently the average and
marginal tax rates on wage and salary incomes rose by the end of the decade
to levels considered politically implausible. As an example, at the beginning
of the 1980s the highest marginal tax rate in Sweden was 85 per cent and 43
per cent of income earners had marginal tax rates exceeding 50 per cent
(Nielsen 1988). Despite the considerations presented in the previous sec-
tion, popular resistance to high income tax rates grew in strength in all the
NCs. Tax evasion and other distortions created by high marginal tax rates
also became more widespread. Generous tax-deduction provisions allowed
high income earners in particular to reduce their effective tax rate by, for
instance, borrowing against their wealth. This made the progressivity of
taxation much lower than that dictated by official tax schedules (Nielsen
1988). For all these reasons, the need to lower the level of direct taxation
was unanimously felt among the Nordic political parties and decision
makers. In any case, the financing of public spending by further tax increases
was no longer considered politically plausible.
The tax reforms proposed in all the NCs in the 1980s could be seen as
one attempt to find a solution to the political impossibility of increasing
income tax rates on the one hand, and essentially cutting public expenditure
on the other. (The first tax reforms were implemented in Sweden during
1983–5, in Norway in 1988, and in Finland only in 1989 (Nielsen 1988).)
The common aim of all these reforms is to lower average and marginal tax
rates on wage income to some extent without, however, essentially reducing
government tax revenue. The method of doing this varies somewhat from
one country to another, but the general idea seems to be to abolish or limit
rights concerning tax deductions and also to enlarge the tax base more in the
direction of property and entrepreneurial income. This would avoid the
harmful effects associated with high marginal tax rates and liberal tax
deduction on interest payments. However, a considerable reduction in
marginal tax rates could also lead to inequalities in the post-tax distribution
of income and would therefore counteract the egalitarian principles upon
which the Nordic societies have been built.
The deregulation and liberalization of financial markets which occurred
in all the Nordic countries in the early 1980s were quite rapid. There were
several background factors in this process: the growing instability of the
world economy after the first oil crisis, the increased integration of the
international financial system, the increasing internationalization of Nordic
productive enterprises, and the worsening financial balance of the public
sector (Horvei et al. 1987: 238). The short-term money market emerged—
first as an unofficial grey market between firms—in Sweden, Norway, and
Finland during the 1970s, and was quickly broadened in the 1980s after
several deregulative measures were undertaken by the central banks. These
developments soon made the traditional policy tools, the regulation of
interest rates and credit rationing, ineffective, and their importance has been
206 KATRI KOSONEN
gradually diminished in all three countries. Today, monetary policy is mostly

conducted through market-orientated measures, i.e. intervention in short-
term money markets and foreign exchange forward markets. At the same
time, the bond market has been deregulated and vitalized in Sweden and
Norway. This has never played any significant role in Finland. Stock markets
have experienced very fast and steady growth in trading activities and cor-
porate values since the end of the 1970s in all the NCs. Nevertheless, the
banks and other financial institutions still dominate the long-term financing
of investments in Sweden, Finland, and Norway to a large extent. Denmark’s
financial system has traditionally been different, with bond markets occupy-
ing the central position in overall credit stock (Horvei et al. 1987: 242).
Foreign exchange controls have been progressively relaxed, and by the 1990s
both short-term and long-term capital movements are almost completely
liberalized in all the three NCs.
Naturally, these developments have strongly influenced the behaviour
of both the household and corporate sectors. As already mentioned, loan
interest rates were kept at artificially low levels throughout the 1970s in
Sweden, Finland, and Norway. In addition, the tax systems in these
countries made the financing of investments by credit rather than from own
savings attractive for both households and business enterprises. This kept
the demand for credit continuously high in all the Nordic economies. Credit
rationing, which particularly affected households, nevertheless limited their
borrowing opportunities and kept their saving rate at a low, albeit clearly
positive, level. Furthermore, the housing finance system in Finland, with its
very high down-payment requirements and short loan repayment periods,
required considerable saving efforts by people who were planning to pur-
chase an own home or had already done so. Because of the serious shortage
of rental accommodation and the lack of other alternatives, the number of
owner occupiers is relatively large in Finland (see Kosonen 1985).
Under these circumstances, the loosening of credit rationing in the 1980s
led to an outstanding increase in the indebtedness of the household sector
in all the three previously regulated Nordic economies. Although the interest
rates also rose markedly with the effect of maintaining clearly positive real
interest rates throughout the whole decade, households increasingly bor-
rowed to finance not only housing but also other durable purchases and, to
some extent, non-durable consumption (e.g. travelling). It is thus not
surprising that the 1980s saw a sharp decline in household saving rates from
levels which were already low by international standards. By the end of the
1980s the household sector saving rate as a whole turned negative in Sweden
and Norway. In Finland it followed a very similar trend, but was still slightly
positive (around 0.5 per cent) in the latter half of the decade.
For the corporate sector, the end of credit rationing and low interest rate
policy meant a marked increase in real interest rates compared with the real
rates of return on physical investments (Growth Policies in a Nordic Perspective

1987: 27). The decline of investment rates in manufacturing, experienced
in particular in Sweden and Norway in the first half of the 1980s, could be
at least partly due to this phenomenon. In addition, the vitalization of capital
and equity markets may have opened up wider possibilities for the realization
of short-term profits on operations in these markets and therefore attracted
private capital away from productive investment. Easy access to credit
facilitated the financing of these transactions largely by borrowed funds.
Whether this will indeed constitute a growth problem for the Nordic
economies in the future remains to be seen. In any case, it is obvious that
the balance-of-payment considerations will not allow any reductions of real
interest rates in the years to come.
The improved liquidity that has followed from the deregulation and liber-
alization of financial markets has been mainly used to finance consumption
and asset market transactions. This has led to rapid price increases in equity
and housing markets. The housing boom has been particularly strong in Fin-
land, and the consequences of this mainly affect ordinary households. Despite
serious imbalances in the housing market, the supply of rental accommoda-
tion has not been sufficiently increased, and many households are still forced
to purchase their own home at a relatively early age. Although access to credit
has been improved, the very high housing price level and high real interest
rates have put these households into an economically difficult situation. The
inequality, which stemmed from the discrimination between banks’ customers
in the regulated system, is being replaced in the ‘market’ economy by inequal-
ity between those who can afford to borrow at high interest rates and are able
to liquidize capital gains and those who cannot. On the other hand, rapidly
expanding household consumption, as well as the shrinking financial surplus
of the public sector, are today considered a serious threat to the external bal-
ances of the nordic countries and thus constrain the conduct of their eco-
nomic policies to a much larger extent than earlier.
REFERENCES
ALESTALO, M., FLORA, P., and UUSITALO, H. (1985), ‘Structure and politics in the
making of the welfare state: Finland in comparative perspective’, in R. Alapuro et
at. (eds.), Small States in Comparative Perspective, Oslo: Norwegian University Press.
ALLEN, T. (1986), ‘Hyvinvointivaltio ja julkisen sektorin kasvu Pohjoismaissa’,
Labour Institute for Economic Research Reports 41, Helsinki.
---- (1988), ‘Suomalaisen hyvinvointivaltion kehityspiirteita — koyhainhoidosta
hyvinvointiin ja tasa-arvoon’, TTTKatsaus, 16: 32–50.
BARRO, R. J. (1974), ‘Are government bonds net wealth?’, Journal of Political
Economy, 82: 1095–118.
208 KATRI KOSONEN
BENOIT, J. P. V. (1985), ‘Taux d’interet intemationaux, endettement exterieur des

pays en developpement et aide etrangere’, in D. Kessler and P.-A. Ullmo (eds.),
Epargne et developpement, Paris: Economica.
CARLSSON, B. (1984), ‘Industrial subsidies in the Nordic countries’, in Economic
Growth in a Nordic Perspective, Helsinki: DOR, ETLA, IFF, IUI, IOI.
DOOLEY, M., FRANKEL, J., and MATHIESON, D. (1987), ‘International capital
mobility: what do the saving-investment correlations tell us?’, Staff Papers, Inter-
national Monetary Fund, 34: 503–30.
ESPING-ANDERSEN, G., and KORPI, W. (1984), ‘From poor relief towards institu-
tional welfare state: the development of Scandinavian social policy’, Institutet for
social forskning, Medelande 5/1984.
Economic Growth in a Nordic Perspective (1984), Helsinki: DOR, ETLA, IFF, IUI, IOI.
FELDSTEIN, M. (1974), ‘Social security, induced retirement and aggregate capital
accumulation’, Journal of Political Economy, 82: 905–26.
----- and HORIOKA, CH. (1980), ‘Domestic saving and international capital flows’,
Economic Journal, 90: 314–29.
FLORA, P., and ALBER, J. (1981), ‘Modernization, democratization and the develop-
ment of welfare states in Western Europe’, in P. Flora and A. J. Heidenheimer
(eds.), The Development of Welfare States in Europe and America, London: Transac-
tion Books.
----- and HEIDENHEIMER, A. J. (1981), ‘The historical core and changing boundaries
of the welfare state’, in P. Flora and A. J. Heidenheimer (eds.), The Development of
Welfare States in Europe and America, London: Transaction Books.
GINTIS, H. (1986), ‘International capital markets and the validity of national macro-
economic models’, Helsinki: WIDER (mimeo).
GLYN, A. (1986), ‘Golden age in historical perspective: some issues’, Helsinki:
WIDER (mimeo).
----- HUGHES, A., LIPIETZ, A., and SINGH, A. (1986), “The rise and fall of the golden
age: an historical analysis of post war capitalism in the developed market
economies’, Helsinki: WIDER (mimeo).
Growth Policies in a Nordic Perspective (1987), Helsinki: ETLA, IFF, IUI, IOI.
HAGEMANN, R. P., JONES, B. R., and MONTADOR, R. B. (1987), ‘Tax reform in
OECD countries: economic rationale and consequences’, OECD Department of
Economics and Statistics Working Paper No. 40.
HORVEI, T., JAGREN, L., MITEK, L., and SALO, S. (1987), ‘Financial markets in the
Nordic countries’, in Growth Policies in a Nordic Perspective, Helsinki: ETLA, IFF,
IUI, IOI.
JALLADE, J.-P. (1988), ‘The redistributive efficiency of European welfare states: basic
issues’, in J.-P. Jallade (ed.), The Crisis of Redistribution in European Welfare States,
Stoke-on-Trent: Trentham Books.
KALDOR, N. (1955), An Expenditure Tax, London: Allen & Unwin.
KALECKI, M. (1970), ‘Problems of financing economic development in a mixed
economy’, in W. A. Eltis, M. F. G. Scott, and J. N. Wolfe (eds.), Induction, Growth
and Trade: Essays in Honour of Sir Roy Harrod, Oxford: Clarendon Press.
KOSONEN, K. (1985), ‘Housing financing and saving in an imperfect capital market:
a cross-section analysis of the saving behaviour of Finnish households’, Licentiate’s
thesis, University of Helsinki (in Finnish).
----- (1988), ‘Why are household saving rates so low in the Nordic countries?’,
Discussion Paper No. 269, Department of Economics, University of Helsinki.
KOSONEN, P. (1987), Hyvinvointivaltion haasteet ja pohjoismaiset mallit, Tampere:
Vastapaino.
KOTLIKOFF, L. J. (1986), ‘Health expenditures and precautionary savings’, NBER
Working Paper No. 2008.
KURIHARA, K. K. (1970), ‘The gap between actual and potential output in growing
advanced economies’, in W. A. Eltis, M. F. G. Scott, and J. N. Wolfe (eds.), Induc-
tion, Growth and Trade: Essays in Honour of Sir Roy Harrod, Oxford: Clarendon Press.
MJ0SET, L. (1986), Norden dagen derpd: De nordiske okonomisk-politiske modellene og
deres problemer pa 70– og 80–tallet, Oslo: Universitets förlaget.
MODIGLIANI, F. (1985), ‘Life cycle, individual thrift, and The Wealth of Nations’,
American Economic Review, 76: 297–313.
NIELSEN,N. C.N. (1988) (ed.), ‘Beskattningssystemerogskattereform-vervejelseri
Norden’, Nordiska Ministerr&det Skriftserie No. 1988: 31.
NlEMELA, H. (1988), Suomen kokonaiseldkejdrjestelmdn muotoutuminen, Helsinki:
Sosiaaliturvan tutkimuslaitos.
OECD (1985a), Social Expenditure 1960–1980: Problems of Growth andControl, Paris.
----- (1986), Historical Statistics 1960–1984, Paris.
----- (1987a), The Tax-Benefit Position of Production Workers 1983–1986, Paris.
POTERBA, J. M., and SUMMERS, L. H. (1987), ‘Finite lifetimes and the effects of
budget deficits on national saving’, Journal of Monetary Economics, 20: 369–91.
SAWYER, M. (1982), ‘Income distribution and the welfare state’, in A. Boltho (ed.),
The European Economy: Growth and Crisis, Oxford: Oxford University Press.
SKINNER, J. (1987), ‘Risky income, life cycle consumption, and precautionary
savings’, NBER Working Paper No. 2339.
SKJAEVELAND, A., and STOKSTAD, P. E. (1988), ‘Why are Norwegian households
low savers?’, Norges Bank Economic Bulletin, 69: 135–48.
Social Security in the Nordic Countries: Scope, Expenditure and Financing 1981 (1984),
44, Statistical Reports of the Nordic Countries, Helsinki: Nordic Statistical
Secretariat.
SODERSTROM, H. T. (1986), ‘Sectoral saving and investment patterns in 16 OECD
countries 1965–1982’, SNS Occasional Paper.
SUMMERS, L. H. (1985), ‘Issues in national savings policy’, NBER Working Paper
No. 1710.
TAYLOR, L. (1983), Structuralist Macroeconomics, New York: Basic Books.
UUSITALO, H. (1988), Muuttuva tulonjako: Hyvinvointivaltion jayhteiskunnan raken-
nemuutosten vaikutukset tulonjakoon 1966–1985, Tutkimuksia 148, Helsinki: Tilas-
tokeskus.
VARTIA, P., and VUORI, S. (1987), ‘Development and technological transformation:
the country study for Finland’, Helsinki (mimeo).
7
Social Corporatism and Long-term Economic
Performance
Michael Landesmann andjuhana Vartiainen
7.1 INTRODUCTION
The focus of the economic literature on social partnership institutions, or

social corporatism, has so far mostly been on the relevance of such institu-
tions for short-term economic management and stabilization policies. This
is not surprising, for two reasons. Firstly, there has recently been interest by
economic theorists in issues of centralization and decentralization; their
work, however, has been mostly concerned with static one-shot analysis and
with comparisons of different equilibria without dynamics (Pohjola 1984,
Ploeg 1987 are exceptions). Secondly, the focus on bargaining institutions
arose from an interest by macro-economists in the experiences of a number
of ‘social corporatist’ (henceforth abbreviated to SC) economies such as
Austria, Norway, Sweden, and Finland. These countries seemed to cope
with the economic slow-down of the 1970s relatively successfully. The
distinct feature picked up in these macro-economic analyses of social cor-
poratist economies was the particular pattern of institutionalized and central-
ized wage bargaining typical for these economies (see Calmfors and Driffill
1988; Freeman 1988). In this chapter we would like to broaden the interest
in the experiences of these economies away from the narrow focus on wage
bargaining and from short-term macro-economic analysis.1
We shall argue that institutions of social partnership and more generally
of ways of mediating social conflicts inherent in capitalist economies are very
1
The short-term approach adopted in most economists’ discussions of the experiences of
social corporatist economies seems to us to be seriously flawed. To illustrate this point, consider
the fact that in discussions about centralization of wage bargaining (see e.g. Pohjola in this
volume), the benefits of corporatism are almost without exception seen to accrue through the
moderation of wages, with a higher employment level and equilibrium capital stock as a result.
The focus of the discussion is whether economies like Sweden in which the growth of real wages
was efficiently checked in the 1970s were able to recover more rapidly. However, if that was all
that mattered in corporatism, we should expect social corporatist economies gradually to
become low-wage economies relying on labour-intensive, backward techniques. We know that
this is not the case, and wage moderation accompanying short-run demand management is not
all that there is to corporatism. The corporatist economies of Sweden and Norway have high
labour costs and rely on technologically advanced industries. One essential ingredient of the
corporatist strategy may indeed be the reliance on technological progress and the socialization
of the costs of readjustment.
LONG-TERM ECONOMIC PERFORMANCE 211
much linked to the long-term performance of an economy. Economic policy
which is based on social partnership institutions rather than simply on de-
centralized market co-ordination can have profound effects on the long-run
evolution of the economy. From a longer-run developmental point of view,
corporatism might be attractive to many economies for tackling structural
change at particular stages of their economic development. This is especially
true for small and open economies which are highly dependent upon the
world market (see Katzenstein 1985). Structural and technical change are
an important domain in which social corporatist arrangements have an im-
pact. Their relationship to structural change is, however, an ambivalent one.
Structural changes, particularly in periods of dramatic industrial restructur-
ing, pose challenges that may or may not make solutions based on broad social
agreement attractive to some parties. Furthermore, the issue of institutional
adaptation is crucial for the analysis of social corporatist economies’ long-run
performance.
In this essay, we do not claim that we can present a grand theory of social
corporatism and long-term economic performance, but by investigating the
issue from various angles we hope to highlight some relationships that have
not received due attention in the discussions of social corporatism.
The chapter is organized as follows. In section 7.2 we discuss a few
features which allow an endogenous economic analysis of corporatist struc-
tures. Section 7.3 summarizes theoretical arguments concerning corporat-
ism and accumulation. We consider the discussion about the centralization
of wage bargaining and investment, and also present a few new arguments,
related to Sen’s (1961, 1967) isolation paradox. We then consider social
corporatist features other than centralization of bargaining that may be of
importance for an economy’s capacity to develop effectively. Section 7.4
discusses the conflict of interests inherent in any growth policies and analyses
the evolution and breakdown of trust in social corporatist arrangements.
Section 7.5 tries to single out those institutional and structural factors that
are specific to corporatist economies’ growth policies, the overall theme
being that small corporatist countries have predominantly rejected pure
market solutions as guarantors of growth, although concrete policies have
taken quite different forms in different countries. This section includes a
comparison of different corporatist countries’ growth strategies. In Austria,
Sweden, Norway, and Finland, growth and accumulation were definitely
tackled in a very interventionist way: in this sense all have been very cor-
poratist. However, the actual content of economic policies and the institu-
tions responsible for carrying out the policy strategies have been different.
Section 7.6 attempts to corroborate at a more empirical level some of the
theoretical predictions worked out in the preceding sections. We ask whether
economic variables related to growth behave in a distinctive way in corpor-
atist economies.
212 MICHAEL LANDESMANN AND JUHANA VARTIAINEN
7.2 ENDOGENIZING CORPORATISM
7.2.1 Structural change in small open economies
It has been powerfully argued by Katzenstein (1985) that some degree of

corporatism is a necessary characteristic of small open economies, which are
crucially dependent on the world economy and trade, but unable to influence
that market in a strategic way. Katzenstein identifies corporatism with small-
ness in general, but his concrete analysis of—what he calls the social, in
contrast to liberal—corporatism shows that the former notion very much
coincides with the notion of corporatism used in this paper.
Katzenstein develops much of his analysis around the political implications
of smallness and openness. A more economical way of reaching similar con-
clusions would be to stress structural factors related to industrial organization
and openness. The export industries of small European economies have
necessarily been strongly specialized: instead of building a small-scale replica
of larger economies, small economies have specialized in particular industries
and in product niches. Many arguments, starting from increasing returns to
scale and the role of fixed costs to the dynamics of research and development
and training, support the notion that some industries, in order to be viable,
cannot be perfectly divisible. Instead, there may be a minimum size required
for them to be competitive. This leads to the specific market structures that
one can in fact observe in the small open economies under consideration.
On the business side, this means specialization in a few fairly specific
branches of industry instead of over the wide spectrum of industries typical
for large economies. This in itself makes it easier for political and organ-
izational links to develop between the business community and the state.
Finland is an extreme example: the very founders of the Finnish state were
bankers and industrialists and their representatives. On labour’s side, this
specialization and concentration leads to the emergence of few powerful
industrial branch unions. This is a characteristic that makes it potentially
easier for the parties of the labour market to enter social contracts with one
another or for such contracts to be mediated and/or enforced by the state.
With a few large industrial unions dominating the trade union movement, it
is easier to establish centralized bargaining structures (for the emergence of
corporatism, see also Therborn’s chapter in this volume).
The above structural characteristics also have implications for the social
arrangements needed to tackle economic change. Restructuring and real-
location of resources are a continuous process in market economies, both
small and large; but in large economies where there is a greater number of
industries, it is more likely that the consequences of structural change
‘average out’: when employment opportunities decrease somewhere in the
economy, there will probably be new ones emerging somewhere else. With
more specialized industrial structures, small open economies are more vul-
nerable to restructuring which may not ‘average out’ to the same degree as
in the bigger economies. This emphasizes a need to devise social agreements
in order to contain the social malaise and conflict that might otherwise emerge.
The two sides of the labour market might, however, have quite different
conceptions about the ‘need’ for concerted action. If firms want to carry out
heavy restructuring, any joint action may be felt by them as a constraint if
they are in a favourable bargaining position, and they may also doubt the
labour organizations’ ability to legitimate the restructuring in the eyes of their
membership. But if social corporatist joint action means that the state is
willing to assume some of the costs of adjustment in the form of, say, training
programmes and the building of new infrastructure, then joint action may
become more attractive from the management’s point of view. Similarly, if
the state buffers the labour force against the negative implications of restruc-
turing by providing alternative job opportunities and help in retraining which
would lead to upward job mobility, joint action will become attractive from
labour’s point of view.
Agreements about restructuring can be struck at various levels. A decen-
tralized bargain may include new investment and wage moderation commit-
ments at the plant level. Similarly, there could be industry-wide agreements
that need not have a broader dimension nor involve the state. However, the
deeper the restructuring that is needed, the more likely it seems that there
will be externalities involved in the costs of reallocating resources and thus
the more likely it is that the state can help to bring about an efficient solution.
By definition, structural change means that some skills become redundant
and some capital equipment obsolete. If this affects a whole industry, any
social bargain has to transcend the limits of that one industry. In such a
bargain, the state can assume some of the costs of reallocating capital, of
creating new human capital, and moving labour. However, some of the
failures of the structural policies pursued in social corporatist economies in
the 1970s and 1980s can also be seen in this light. In a number of cases,
corporatist policies were restricted to individual branches such as steel or
shipbuilding, when in fact industry-transcending policies were needed (see
Strath 1988).
The comprehensiveness of corporatist arrangements is certainly relevant
for the capacity of an economy to be restructured (see chapter 1 of this
volume). The more encompassing a corporatist arrangement, the more likely
it is that individual agents are ready to accept short-term losses in exchange
for better long-term prospects: a worker (or a farmer) can accept restruc-
turing and redundancy if he can be confident that the system of concerted
action is democratic and representative enough to guarantee new jobs as
well. A less comprehensive corporatism, on the other hand, is more likely to
degenerate into a stagnant system where the interests of limited sections of
the economy are taken care of, at the expense of deeper renovation. This
criticism can possibly be raised against some developments which charac-
terized the Austrian social partnership system in the 1970s.
While the above points to a functionalist rationale for corporatist arrange-
ments in small open economies, such an approach has to be supplemented
by an analysis of the specific organizational structure and policies of em-
ployers and employees in those economies. Studies of the historical origins
of corporatism (cf. Katzenstein 1985; Esping-Andersen 1985) do show that
times of crisis and restructuring provided the impetus for organizing social
corporatist arrangements. In Sweden, the economic crisis of the 1930s
brought the lesson home, whereas in Finland the turbulence of the 1970s
showed to both parties that a more consensual approach to economic policy
might do a better job. In Austria, the civil war experience between the world
wars and the need for social and economic reconstruction after World War
II were the immediate catalyst.
7.2.2 Labour shortage and labour organization

Another factor, neglected by political scientists who have studied corpor-
atism, that may move economic policy towards corporatist arrangements is
the labour supply situation. The relative abundance of labour is probably
one key factor which determines its relative bargaining strength. In an
economy in which there is a large reservoir of labour in traditional occupa-
tions or which can rely on immigration, the bargaining power of labour is
more likely to remain relatively weak and firms will be less interested in
entering social corporatist arrangements. On the other hand, when the sup-
ply of labour becomes scarcer, firms might perceive advantages in participat-
ing in more comprehensive social regulation of income distribution. In this
essay we shall not attempt to carry out a careful statistical analysis of the
emergence of concerted action in relation to labour supply developments.
However, we would like to point out a few historical chains of events which
show that labour supply may be an important determinant of social cor-
poratist arrangements.
Finland is an example of a small country in which the end of the 1960s
effectively meant a possibly lasting evolution towards concerted action.
Finnish economic policy during the 1950s and 1960s had not been charac-
terized by partnership and joint action in economic policy formation. Eco-
nomic policy was interventionist and geared to fostering growth and
investment, but it was a case of corporatism without the social component
and was plagued by confrontations. Fiscal policy aimed at balanced budgets
on a yearly basis while monetary policy was directed towards accommod-
ating the financing needs of the open sector. This resulted in a pro-cyclical
economic policy and a very volatile economy (see e.g. Pekkarinen 1989).
The central bank was autonomous and mostly paid attention to the external
balance of the economy. Real wages followed a political business cycle: in
times of high export demand, wages increased fast, but this increase was then
checked by devaluations, coupled with austerity policies which decreased the
bargaining power of labour. However, from the end of the 1960s onwards,
a system of consensual incomes and tax policies was gradually developed,
and it is of interest to note that this coincided with the completion of an
important phase of structural change in the economy. During the 1960s, the
flow of labour from agriculture was still an important source of labour
supply, whereas by the beginning of the 1970s this process was more or less
complete.
In addition, one can consider the overall evolution of European labour
markets during the post-war era and especially the turbulent events at the
end of the 1960s. Many accounts of post-war economic growth in fact
attribute a key role to labour supply (Kindleberger 1967; Kaldor 1966;
Boltho 1982). Thus labour was quite abundant in Europe during the 1940s
and the 1950s, and indeed the historical process of urbanization and in-
dustrialization was in the 1950s still far from complete (Hyman 1978).
Hence rapid and steady growth in the 1950s was possible thanks to the
internal migration of labour from agriculture and traditional industries to
modern manufacturing and services. Towards the mid-1960s this reservoir
gradually exhausted itself, although the scarcity of labour was for a while
mitigated by migration from southern Europe.
In the 1950s and 1960s, the strong demand for labour generated first by
reconstruction and then by steady economic growth could be satisfied with-
out excessive wage inflation. During the 1960s, however, the prolonged con-
ditions of full employment, together with the slowing down of population
growth and the exhaustion of the agricultural labour reservoirs, strengthened
the bargaining power of workers and allowed an increase in their relative
income share. This enhanced bargaining power was also reflected in the
increased regulation—both by legislative means and by collective agree-
ments—of employment conditions (safety requirements, minimum wage
legislation, lay-off compensation, and other job security). These trends
peaked towards the end of the 1960s, and it is probably not coincidental that
it was at that time that a new interest in encompassing social contracts arose
in almost all the major European economies.
Changes in respective bargaining strength affected the interest that the
various parties of the labour market had in corporatist arrangements. Most
of the big European economies at that time flirted briefly with the idea of
concerted action, but in none of them did this yield any lasting results. In
West Germany, comprehensive wage settlements were agreed upon in 1967
and 1968, but in 1969 the social consensus disintegrated: the trade unions
could not contain labour unrest and the central bank reverted to an
autonomous policy of monetary orthodoxy (see Scharpf 1984). In Britain, a

social contract was attempted in 1976–9, but it broke down in the ‘winter of
discontent’, leading to a dramatic shift towards confrontational policies
under Mrs Thatcher. The ideological roots of the British labour movement
did not favour co-operation and the social contract in the form of incomes
policy did not deliver much to the workers in exchange for wage moderation.
After die electoral success of the Conservatives in 1979, confrontation was
the rational assumption for all actors in the economy. In Italy, labour markets
were turbulent from the end of the 1960s onwards, and in 1977–9 this
generated an experiment of ‘solidarity nazionale’, with the unions supporting
wage moderation in exchange for legislation on retraining, youth employ-
ment, and restructuring. As in the German case, this proved short-lived: both
sides of the labour market gradually grew more suspicious of concerted
action (Regini 1984). France fits perhaps less well into the general pattern.
Labour supply in France was scarce throughout the post-war period, and
employment in manufacturing did not increase by as much as in the other
countries. Yet a massive restructuring of labour within industry into high
growth activities seems to have taken place (Sautter 1982). An immediate
reason for the absence of concerted action before the Mitterrand government
was simply the absence of labour organizations strong and unified enough
to claim the workers’ mandate. French economic policy has been ‘statist’,
and the government largely relied on regulative measures which affected
wages and prices. It was not until Mitterrand’s presidency that the first
attempts towards a social contract with the main unions took place. How-
ever, the organizational basis for this remains weak: the socialist-controlled
CFDT is much weaker than the CGT which shows scant interest in
co-operating with a socialist-dominated government. Thus while labour
scarcity seemed to bring social corporatism on to the agenda, tendencies
towards social corporatism were in general defeated in the large European
economies.
In the 1980s, mass unemployment persisted in most OECD countries
and, after labour’s period of strength in the late 1960s, slow growth and rapid
technological change thereafter moved labour markets again towards an
excess labour supply regime. This has decreased labour’s bargaining power
and political muscle, leading in many countries to a shift towards more
orthodox economic policies. The uneven pattern of technological advances
and the associated processes of industrial restructuring also contributed to
an increasing inequality in labour’s bargaining power and thus to greater
income differentials. The SC economies have in this respect behaved quite
differently from other economies (as is discussed in more detail below), but
social corporatist structures certainly have had to and still do continue to
adjust to changing circumstances.
7.3 DOES CORPORATISM FAVOUR INVESTMENT, GROWTH AND

STRUCTURAL CHANGE? THEORETICAL ARGUMENTS
7.3.1 Wage restraint and the dynamic inefficiencies of capitalism:

the role of centralization
We now turn to the more common way of addressing the issue of corporatism
and ask whether corporatism, when interpreted as centralization of wage bar-
gaining, leads to more capital accumulation and growth. For the moment,
we adopt the narrow view, criticized above, according to which the positive
implications of corporatism work through wage moderation.
The relationship of wage setting to investment has generated a volumin-
ous literature in theoretical economics. However, the relationship of central-
ization of bargaining and investment and accumulation has been to a large
extent neglected, Sen’s isolation paradox (Sen 1961, 1967) and its extensions
(see Pohjola 1984 and Chapter 3 of this volume) being the few exceptions.
The economists’ theoretical discussion about the dynamic inefficiency
of capitalism has focused almost exclusively on settings where the game is
between two actors, the firm and the union, possibly interpreted as the
employers of the economy and the workers. Thus the centralization of the
bargaining structure has not been integrated into the analysis. This strand
of research has generated important insights about capitalism (the original
argument was in Lancaster 1973; see also Grout 1984 and Ploeg 1987 for
representative contributions). The main insights of this discussion are intuit-
ive: in a dynamic game in which the capitalists invest and the workers set the
wage, there is a dynamic externality leading to a suboptimal capital stock:
the capitalists will not invest as much as an efficient solution would stipulate,
because then the workers would have an incentive to renege on the original
wage agreements and extract the differential rent of the installed capital. It
would be advantageous to both if workers could commit themselves to lower
wages, but it is difficult to see how such commitments could be enforced in
a normal market economy based on free contracting between the agents.
On the other hand, there is the more recent literature, largely provoked
by Calmfors and Driffill (1988), which focuses on the effects of the bargain-
ing structure on employment. This literature is essentially static and does
not consider investment and capital accumulation. Of course, one can always
argue that if, in a world of centralized or completely decentralized bargain-
ing, the workers set lower wages and consequently there is higher employ-
ment, a trivial dynamic generalization would be that in such a world
employers invest more and the ensuing capital stock is higher. This general-
ization is potentially important but theoretically trivial, since it essentially
treats the long run as a sequence of short-run equilibria which are the result
of the myopic one-shot behaviour of economic actors.
Thus one would like to say more. Without working out elaborate analytical
models, we present additional arguments which potentially link together the
bargaining structure and investment and accumulation:
(a) Dynamic inefficiencies and the age profile of workers

The Calmfors–Driffill results build on the fact that a centralized trade union
aggregates the preferences of workers in various industries. However, it is
clear that a centralized trade union federation also aggregates over prefer-
ences of workers of various ages. This may have profound implications for
investment and employment if there is some labour mobility in the economy.
The intuition of this argument goes as follows: consider the labour market
of an economy which consists of overlapping generations of workers, young
and old. Thus a worker arrives at the labour market when young, works for
one period as young, then works for another period as old, and finally leaves
the labour market. Assume also that there is labour mobility in the labour
market, so that the young workers have to change to another firm once the
first period (youth) is over. Assume also that the number of workers is large
relative to that of firms, so that on average half the workers in a firm will be
young, half old. Firms invest between the periods in new capital stock and
want to maximize their profits in the usual neo-classical vein. In this eco-
nomy, the bargaining structure will affect the equilibrium capital stock.
Consider first a decentralized bargain in which at the beginning of each
period the wage agreement is struck individually in each firm. Now the old
workers in each firm have no incentive not to use their bargaining power
fully, since their eventual sacrifice in the form of wage moderation (a higher
capital stock in the next period) will not accrue to fhem since they will be
out of the market. The young could in principle enjoy the fruits of modera-
tion, but they see no incentive to disagree with their elders’ maximalist
demands, since anyway they will have a new job in the next period and their
sacrifice will not pay off for them. A centralized bargain will give a very
different result. In the centralized federation, the old will again vote for max-
imalist increases, but now the young will have an incentive for moderation:
if a generally low wage agreement is passed then, with rational capitalists,
the capital stock in the next period will be higher everywhere, and the young
will enjoy higher employment and wages, regardless of where they are em-
ployed. If the centralized federation aggregates preferences in a democratic
way so that the young have at least some say, then the wage demand will
be lower than in the decentralized case. This insight is quite robust and it
can be made precise in a variety of models, be it with wage setting by the
union or by using the Nash bargaining solution over the partners’ respective
pay-offs. Thus the centralized federation can be seen as a guarantor of moder-
ately priced labour, regardless of age and place of employment. This argu-
ment may shed light on the interpretation of the Swedish model, in which
retraining and labour mobility have always been emphasized. It fits well with
the very centralized bargaining structure; otherwise, labour mobility could
mean that there would never be any incentive for workers to make sacrifices
for the firm in which they are working at present. The argument can also be
used to assess the consequences of demographic structure on the labour mar-
ket: with an ageing population, one would expect shorter horizons for the
trade unions and consequently higher wage demands. The assumption that
all the workers change jobs is not essential—the argument carries through
even if only some of them do. This discussion can be seen as an extension of
Sen’s (1967) isolation paradox, potentially more important than the com-
mon model of infinitely lived workers employed with a constant capital stock.
(b) Information and the building of confidence and reputation

The original result of Lancaster (1973), according to which the capitalist
economy suffers from dynamic inefficiency, has been largely modified in the
more recent discussion based on modern game theory (see Kaitala and
Pohjola 1990, for example). According to this criticism, dynamically effi-
cient equilibria can in fact be sustained as non-co-operative Nash equilibria
when the players use some kind of memory strategies, which take into
account the past history of the game. Thus an efficient equilibrium could be
sustained by trigger strategies whereby the workers do not use their full
wage-bargaining power and the capitalists invest more than the static profit-
ability condition would stipulate; if any of the parties deviates from this solu-
tion, the other party punishes him either for some time or for the rest of the
game by reverting to the open loop equilibrium solution. A crucial condition
for this solution to be sustainable is that the parties can effectively monitor
each other’s behaviour and thus build reputations for ‘co-operative’ behavi-
our. It is a typical feature of economies with centralized bargaining structures
and social partnership arrangements that the labour market partners make
up many joint bodies, often with the participation of or organized by the state
which monitors the key economic variables and keeps all the parties informed
about them. This can be speculatively interpreted as an institutional attempt
to sustain long-term efficient solutions. Broadly speaking, institutions of
social partnership may induce an atmosphere of mutual confidence in
forward-looking behaviour.
(c) The process of creative destruction

Centralized incomes policies have often though not always meant a solidar-
istic wage policy whereby wage increases have been more uniform through-
out the economy. Hence firms whose profitability is above the average will
benefit from centralization whereas inefficient firms will suffer. This may
contribute to rapid technical progress and structural change: firms which
use outdated production methods will be wiped out instead of being able to
get by thanks to low wages and access to a segmented labour market. Thus
a solidaristic wage policy may ‘lend a hand to the invisible hand’; a solidar-
istic and centralized wage policy may in this way bring the economy nearer
to the Walrasian ideal of perfect labour market arbitrage. This is also an instance
in which social corporatism and the state can correct dynamic inefficiency:
the socially optimal shift of resources from declining industries is facilitated
when the state assumes some of the costs. Thus a solidaristic wage policy
may require other state intervention to work smoothly: as inefficient indus-
tries are being reorganized, it helps if the state can bear part of the social cost
of transferring resources and supporting the evolution of new industries.
7.3.2 Investment in infrastructure and technical progress

Another interpretation of the isolation paradox is in terms of technical
progress and infrastructure investment. By definition, technical progress
means the shifting of the technology frontier. It is well known in the eco-
nomic literature (see e.g. Arrow 1962) that the private costs of and rewards
for pushing the technology frontier outwards do not coincide with the
social costs and benefits, and that it is probably hard to get the frontier pushed
outwards by individual firms or groups of workers alone. This may explain
the corporatist countries’ large investments in education and infrastructure.
There are, of course, other reasons for such expenditure. In countries
which have to close a ‘technological gap’ (as was the case with Finland and
Austria for most of the post-World War II period), be they social corporatist
or not, the necessary initial outlay on training and infrastructure is simply
too large to be undertaken by the private sector. The same could still be true
for economies which are small relative to the large businesses operating inter-
nationally in certain industries and with which domestic firms have to com-
pete. In both situations one can predict that a large part of spending on research
and development and on the general upgrading of skills will be undertaken
by the state. However, in more recent times, when large businesses in
countries like Sweden and Finland have become heavily internationalized
and become successful players in a competitive international environment,
we can perceive an increasing share of spending on R. & D. and training
being internally financed by the successful firms. Such firms also have clear
advantages in retaining their labour force, because of superior employment
conditions, and thus do not have to face the problem of high labour turnover.
7.3.3 The increased power of the unions; participation and productivity

Much of the discussion so far identifies corporatism with centralization of
the representation of the workers’ interests. However, another relevant feature
of social corporatism is that workers’ participation and industrial democracy
is enhanced. This has been associated historically with centralized trade union
movements, since only in those countries in which the labour movement has
been efficiently organized has far-reaching industrial democracy legislation
been enacted. Industrial democracy may increase workers’ identification with
their firms and thus increase work motivation and profitability. On the other
hand, employers have often perceived industrial democracy as a threat.
There may be another conceptual ‘hump-shaped’ relation here: of all pos-
sible worlds, the employers would probably choose one in which the labour
and union movements were so weak that the employers could dictate wages
and the organization of work. According to this argument, increased power
of the unions lessens the employers’ ability to squeeze out as much as pos-
sible from the worker and hence a shift of power towards labour is detrimental
to productivity. The counter-arguments are obvious: labour’s bargaining
power enhances industrial democracy and motivation and hence productivity.
Thus industrial democracy could be seen as a form of efficiency wage, albeit
perhaps not very costly to the employer. Furthermore, it may be precisely
the strength of the workers which makes it worth while for the individual
worker to invest in job-specific skills: he knows that his work conditions will
remain satisfactory and his employer will not extract the fruit of his invest-
ment for the benefit of the firm alone.
To conclude, let us mention saving as a potential object for corporatist
policies. Sen’s isolation paradox was originally cast in terms of suboptimal
saving, and it is certainly noteworthy that all the Nordic countries have re-
lied on a high degree of public saving. (For a thorough development of the
notion of forced saving through taxation, see Kosonen’s chapter in this volume.)
7.4 THE CONFLICT OF INTERESTS AND THE EVOLUTION OF

SOCIAL CORPORATIST STRUCTURES
Much of the argument so far offers reasons as to why concerted action and
social corporatism may be rational for all parties. Yet corporatist arrange-
ments have emerged in only a few countries and nowhere have they emerged
by common consensus. The above arguments should thus not hide the fact
that besides the eventual mutual interest in co-operation there is also a con-
tinuous conflict of interests in capitalism, which poses a threat to and an
obstacle for corporatist institutions.
Of all possible worlds, the employers would probably choose one in which
labour was weak and they would never need collaboration. This would result
in low wages, high profit margins, and, possibly, high growth. Some of the
NICs have used this route, and this confrontational strategy is the very
essence of Thatcherism. However, once labour is strong and organized, it may
become more attractive for firms to enter corporatist arrangements, using
the state to assume some of the social costs of business activities and relying
on the labour unions’ ability to legitimate wage moderation to the rank and
file. These arrangements take some time to evolve in order to build the
trust needed. Thus both of these two social patterns may be self-enforcing:
both trust and distrust build slowly and reputations are hard to change.
Modern game theory shows that it is possible to sustain dynamically effi-
cient paths when the parties’ strategies depend on history. The parties can
use various schemes for punishing a party which deviates from the efficient
solution. This makes the analysis much more relevant than the previous
theories which seemed to predict inefficiency or require binding long-term
commitments. Thus game theory provides a useful interpretation of phe-
nomena in which periods of trust and confrontation seem to alternate.
Thus while trust builds up over longer time periods there is also learning
from periods of mistrust. Examples for dramatic shifts from non-co-operative
to co-operative behaviour are the changes in Austria before and after World
War II and also the changes which took place in the socio-political climate
in Finland in the 1970s (see discussion above). Co-operative ways of behav-
ing probably fare better in those economic environments that are not subject
to very abrupt changes. In such environments, all the agents have accumu-
lated the necessary information about the respective pay-offs and valuations
of the other agents. On the other hand, when the environment changes fast,
the players may appreciate the situation in different ways and it is probably
easy to interpret a deviation from the co-operative strategy as a betrayal or
shift of mood, even when it may be based on a different evaluation of the
changing environment. In a changing environment, it is more likely that the
parties become suspicious of each other, expecting the other party to use
the new environment to its own advantage. We argued above that one essence
of social corporatism is its capability of handling technical change; but there
is probably an inherent contradiction in it in the sense that corporatism be-
comes more difficult to sustain in periods of rapid change.
Uncertainty about the environment is related to how well the members of
the social partnership are able to allocate to themselves both power over and
responsibility for those variables which they have an interest in. For example,
the difficulties of corporatist collaboration in Finland and in Germany have
both been related to the fact that an important range of decisions, namely
those of the central bank, are effectively out of the partners’ control. This
creates another source of uncertainty. Thus a central bank which was keen
to keep its independence and maintain an orthodox monetary policy was an
additional factor which contributed to the disintegration of the German
social corporatist experiment in the 1960s. Similarly, the uncertainty about
the future exchange rate in Finland has made it more difficult for the labour
market partners to agree on wage increases. In the 1970s the central bank
sometimes threatened to use its policy tool—usually a devaluation—only
after the wage agreements had been struck. In Austria, on the contrary, the
parties of the social partnership have effectively shared amongst themselves
the mandate to control the central bank, so that both the responsibility for
and the power to make exchange rate policy and monetary policy decisions
have been thrust upon them.
Finally, if the building of trust is a continuous and contingent process,
there probably is a role for the state in fostering an atmosphere of trust and
co-operation. In Finland, the state in the 1980s clearly assumed the role of
‘mediator’ in the sense that it asserted its readiness to use its other policy
tools to make agreements easier. Thus changes in tax rules have been used
to make consensual arrangements more attractive.
7.5 INSTITUTIONAL AND STRUCTURAL CHARACTERISTICS OF

CORPORATIST ECONOMIES
There is considerable diversity in the actual growth strategies adopted by the

corporatist countries. Yet it is easy to find a common denominator in all of
them: the distrust of liberal laissez-faire. In these small economies, most of
which were late industrializers, growth and structural change were simply
perceived as too important to be left to market forces. Instead, these issues
were identified as central for national economic development. We shall now
describe in more detail those institutional and structural characteristics of
the SC economies which are related to economic growth.
7.5.1 Labour market institutions and economic growth

The discussion concerning bargaining structures and economic growth in
sections 7.3 and 7.4 was of a general nature and could not therefore present
a picture of the differentiated facets by which labour market institutions
affect economic growth in particular economies and in particular historical
circumstances. We will now discuss some of these particular features.
Of all the countries included in our comparison in this chapter, Austria
has probably the highest degree of participation of trade unions in framing
economic policy. However, while there is a high degree of centralized parti-
cipation of the trade union organization in macro-economic policy formula-
tion (the various institutions of the social partnership arrangement, such as
the ‘Paritatische Kommission’), there are also powerful branch unions par-
ticularly in the ‘older’ industries, such as steel, heavy mechanical engineer-
ing, etc. It so happens that these industries are also in public ownership. The
bargaining situation in these sectors is thus very different from the situation
faced by the employees either in the smaller-scale Austrian-owned business
sector or in the foreign-owned enterprises which operate in most of the
‘modern’ sectors of the economy (electrical engineering, instrument engin-

eering, etc.).
Industrial relations are thus rather segmented and this might explain why
the foundations for a ‘solidaristic wages policy’ (in the sense of an egalitarian
wages policy) may not have existed in Austria as compared to, say, Sweden
and Norway. It might also explain why in Austria, unions in the ‘older’
manufacturing industries were able to resist restructuring in the 1970s until
the degree of indebtedness of the nationalized industry sector could no longer
be sustained and a U-turn in the management of the nationalized industries
was initiated in the early 1980s. On a macro-economic level, on the other
hand, the clear leadership position of a few powerful branch unions may have
facilitated the implementation of national wage agreements when macro-
economic conditions required these.
In Sweden, concerted social action has been understood as a contract
between the parties of the labour market, independent of the state. The term
‘incomes policy’ has therefore also been consistently avoided by the archi-
tects of the Swedish model. Wage bargaining has been quite centralized and
the central trade union federation LO has effectively bargained on behalf of
the individual unions. Wage settlements have also been more solidaristic
than in other SC countries. Sweden could thus be singled out as an example
of an economy in which the growth of real wages in certain industries has
been held back by centralized wage agreements. It has been argued, for
example, that the low growth in real wages contributed to the growth of the
Swedish automobile industry in the 1960s (Mjoset et al. 1986).
The early commitment to a solidaristic wages policy was partly inspired
by the normative idea of similar work effort earning similar remuneration.
However, the notion of creative destruction caused by the relative uniform-
ity of wages (see section 7.3.1) also figured in the early Social Democratic
and trade union writings. The wage earners’ funds project was a logical implica-
tion of this strategy: while the micro-economic control of investment was
mainly left to private capitalists, the ‘extra’ surplus generated in efficient sec-
tors because of centralized wage settlements was to be controlled politically.
However, the real impact of the funds has apparently been quite limited.
Sweden’s trade unions are probably the most powerful in the world, with
the additional characteristic of white-collar employees being also to a large
extent unionized. The political participation of trade unions, however, is not
as extensively institutionalized as in Austria. Rather, economic policy and
labour market strategies are twin facets of the Social Democratic strategy.
Hence the integration of labour organizations into policy-making is more
political and substantial than institutional and formal. However, the strength
of the Social Democrats and the trade unions has led to the introduction of
extensive schemes of industrial democracy both at the work-place and within
the firm.
In Norway, wage bargaining has by and large been less centralized than in
Sweden. The state’s involvement in incomes policy grew in the 1960s, but
still it is probably not too far off the mark to say that the essence of Norwegian
corporatism has been in industrial policies and macro-economic planning
rather than in centralized bargaining and incomes policy.
In Finland, centralized collective bargaining and concerted action are
more recent phenomena. For two decades after World War II, the trade union
movement remained weak by Nordic standards and politically divided. It
was only at the end of the 1960s that a unitary trade union organization was
formed and the first centralized wage bargains were struck. However, the Fin-
nish trade union federation SAK has never had a mandate as extensive as that
of the Swedish LO to bargain on behalf of the individual unions. Yet the role
of the state has been prominent from the outset of social corporatism and
fiscal decisions have often been used to enhance co-operative labour market
outcomes. This is a new feature of social corporatism in Finland. Industrial
democracy and labour participation in macro-economic policy-making were
almost non-existent in the 1950s and 1960s. It was only after the harsh lessons
of the 1970s—first rapid inflation and erosion of profitability, then austerity
policies from 1975 on—that practices and institutions for concerted social
action emerged. At present, the joint monitoring body of the policy makers,
employers, and labour plays an important role in the formulation of macro-
economic policy. A factor which has probably made it more difficult to sustain
co-operation has been the independent role of the central bank which, from
the point of view of policy makers and labour market organizations, makes the
bargaining environment more uncertain. Legislation to promote industrial
democracy was introduced in the 1980s. In contrast to Sweden, profit-
sharing schemes have been completely confined to the enterprise level.
7.5.2 Finance and economic growth

Throughout the 1950s and 1960s, the role of stock markets in allocating
finance to business was unimportant in the SC countries. This did not
necessarily mean a ‘socialization of the investment function’. Austria came
nearest to this left-Keynesian idea since the mediation of finance basically
happened through banks which were nationalized. In Sweden, the banks
remained mostly in private hands although the policy makers had some
quantitative and qualitative control of flows of capital. In Finland the bank-
ing sector has been more independent and less subject to political interven-
tion, which reflects the more conservative character of the Finnish policy
model. But in all countries the banks have formally controlled the allocation
of finance, and in all countries the banking sector has been oligopolistic.
Thus important financial decisions have acquired a more strategic character
than what would have happened in a stock market; policy makers and
powerful bankers have exerted a very strong influence on key investment

decisions. In addition to the oligopolistic character of banking and finance,
monetary authorities have used various instruments of credit control. After
World War II, a policy of low interest rates was adopted in all the corporatist
countries to suit the needs of reconstruction and rapid investment. This
necessitated rationing of the available saving supply away from consumption
and credit rationing has been in force in Austria, Sweden, Norway, and
Finland. In Finland, however, the policy makers’ grip on the credit markets
was probably less powerful than in the other countries, owing to the lack of
appropriate policy instruments other than the control of banks’ behaviour at
the discount window (Vartiainen 1989).
Norway is the prime example of an active credit policy geared to enhance
structural change. In addition to public saving, Norwegian ‘Credit Socialism’
has consisted of extensive regulation of both the volume and the micro-
economic allocation of credit. State banks have played an important role in
the mediation of finance to key sectors of heavy industry, and private banks’
credit allocation decisions have also been subject to far-reaching governmental
steering.
In the 1980s, however, monetary markets were extensively liberalized in
all SC countries. In a much more internationalized financial system, the
traditional array of monetary controls has become difficult to maintain.
Furthermore, many of the large firms have internationalized their own opera-
tions and advocate access to a liberalized financial system. Finally, many of
the SC economies which started after World War II at a lower stage of indus-
trial development have now developed mature and competitive industrial
structures. One could argue that at this stage the previous regime of financial
control is less functional.
7.5.3 The role of the public sector in economic growth

(a) The role of nationalized industries: The nationalized sector played an
important role in post-war economic development in Finland, Norway, and
Austria. In Finland, it was instrumental in the emergence of important indus-
trial branches such as mining, chemicals, and petroleum refining. In Austria,
where privately owned Austrian businesses were largely small and medium
sized, the nationalized industries were the only non-foreign-controlled large-
scale firms. They were (and are) to be found largely in basic processing (steel,
mining) but also in chemicals and heavy mechanical engineering. In Sweden,
nationalized industries never played such an important role.
Nationalized industries in Austria also played an important role in the
full-employment strategy of the Austrian government after the first oil-price
shock (see Landesmann 1990). By the late 1970s, the publicly owned
companies came under much pressure—as was the case in many other
European economies—to operate in the same fashion as private businesses.

Following a period of mounting deficits experienced by these corporations,
this led to massive restructuring in the 1980s and even to privatization. In
Finland, the management of publicly owned firms was always rather inde-
pendent of political interference and their operation did not differ much from
the way private businesses operated. In Norway, publicly owned firms played
an important role in the country’s distinct regional policy. At present, though,
one can perceive changes in the orientation of Norwegian industrial policy
in this respect. In Sweden, public ownership of industries has played a less
important role, with die exception of the late 1970s when the state assumed
a substantial share of the cost of maintaining a domestic shipbuilding indus-
try (note, however, that 1976–82 was precisely the time when the Swedish
model was ‘run’ by centre-right governments, and not by its Social Demo-
cratic architects).
(b) The public sector as employer of last resort: The role of the nationalized
industries in this respect has already been discussed in the previous para-
graph. In addition, empirical evidence to be presented later on shows the im-
portant role which public sector (non-manufacturing) employment played in
all SC countries in the crucial period from 1973 to the late 1970s in maintain-
ing full employment. Such a policy was indeed difficult to pursue in circum-
stances of rapid labour shake-out in the manufacturing sector and growth in
the labour supply due to population growth and/or increased participation.
The state sector acted as a ‘sponge’ to absorb the otherwise redundant labour
force and thus helped to lift employment levels during crucial phases of
industrial restructuring in the 1970s.
(c) High spending on social services: High spending on social services has
traditionally been a hallmark of social corporatist economies. In the period
of economic crisis after 1973, such spending played an important role in main-
taining demand when investment collapsed in some of the economies. It also
contributed to the general perception that the economic crisis did not trans-
late itself into a social crisis of equal proportions. This allowed much of the
institutional framework which characterizes SC economies to stay intact.
Apart from the above, the safety-net which social services provide has
probably contributed to labour mobility across regions and occupations. It
has also brought into the labour force sections of the population (especially
women) which in many other economies have been driven out of the active
labour supply.
(d) Public investment in infrastructure: The social corporatist experiments,
particularly those of Scandinavia, have always paid special attention to
training and retraining of the labour force, which satisfies the requirements
of a continuously restructuring economy. Emphasis was also put on match-
ing skills with an upward movement of the industrial structure towards high
value-added segments of international production.
Research and development spending started from a relatively low level,

particularly in Finland and Austria. Its level (measured as a proportion of
GDP) is still low in these two economies, but the growth rates of such spend-
ing are comparatively high. Sweden on the other hand spends a higher pro-
portion of its GDP on R. & D., and is among the technologically most
sophisticated of the industrial world.
7.5.4 Saving and investment

(a) Structure of saving: The ‘Nordic model of accumulation’ is extensively
analysed in Kosonen’s chapter in this volume. Kosonen demonstrates that
private saving rates are very low by international comparison. On the other
hand, public saving financed through a relatively high burden of taxation
compensates in such a way that investment rates are in general not low by
international comparison. Recently, however, tax reforms were introduced
in most of these economies, attempting to put a halt on further increases in
taxation. In the future, therefore, it is questionable whether public savings
will continue to play such an important role.
(b) Support for investment: Various schemes to support private investment
activity such as early depreciation allowances (particularly generous in
Austria) and interest rate allowances have been used very prominently in the
SC countries. For a detailed discussion of these policies see Landesmann’s
chapter in this volume.
7.6 EMPIRICAL OVERVIEW OF FEATURES OF ECONOMIC

GROWTH IN SOCIAL CORPORATIST ECONOMIES
(a) GDP and manufacturing output growth: Tables 7.1 and 7.2 present growth
in GDP and GDP per capita for the four economies discussed in this paper
and for the OECD and OECD Europe as a whole. We can see that the
growth in GDP and GDP per capita was higher than in OECD Europe in
three of the four economies (Swedish growth rates being somewhat lower).
Norway’s high GDP growth in the 1970s due to North Sea oil is particularly
striking.
Output growth in manufacturing (Table 7.3), however, is particularly low
in Norway over the 1970s and the 1980s reflecting the fast process of
deindustrialization. It is very high in Finland over the entire period and in
Austria until the early 1980s. Sweden experiences a recovery in the 1980s
after a depressed manufacturing growth performance in the 1970s.
All in all, a comparison of the GDP and manufacturing output growth per-
formance with the OECD European average reveals nothing very distinctive
about this group of economies. Two of the countries (Austria and Finland)
TABLE 7.1 GDP (growth per annum)
(%)
1960–8 1968–73 1973–9 1978–87 1960–87
Large OECD
countries 5.0 4.4 2.7 2.6 3.7
Austria 4.2 5.9 2.9 1.6 3.5
Finland 3.9 6.7 2.3 3.3 3.9
Norway 4.4 4.1 4.9 3.3 4.1
Sweden 4.4 3.7 1.8 1.8 2.9
OECD Europe 4.6 4.8 2.5 1.8 3.3
TABLE 7.2 Real GDP j per capita (growth)
(%
1960––8 1968–73) 1973–9 1978–87 1960–87
Large OECD
countries 3.9 3.3 2.0 1.9 2.4
Austria 3.6 5.4 3.0 1.6 3.2
Finland 3.3 6.5 2.0 2.9 3.5
Norway 3.6 3.3 4.4 2.9 3.5
Sweden 3.6 3.1 1.5 1.6 2.5
OECD Europe 3.6 3.9 1.9 1.3 2.6
TABLE 7.3 Real value-added in manufacturing (growth)

(%)
1960–8 1968–73 1973–9 1978–87 1960–87
Large OECD
countries 6.3 5.4 2.3 2.6 4.1
Austria 4.5 7.4 2.7 1.5 3.7
Finland 5.3 8.6 2.9 3.7 4.8
Norway 4.3 5.0 0.2 0.9 2.5
Sweden 5.3 9.0 0.5 1.8 2.9
OECD Europe 5.6 6.1 2.0 0.8 3.5
experience in general higher manufacturing growth performance than the

OECD European average and the other two economies (Norway and Sweden)
lower growth. The same could be said about productivity performance in
manufacturing (see Fig. 7.1 and Table 7.4).
(b) Productivity and real earnings growth and the recovery of investment: Fig. 7.2
compares the real earnings and productivity paths in manufacturing for each
of the four SC economies and also for OECD Europe. Here a distinct picture
does emerge: all the four economies experience over the 1970s and the 1980s
FIG. 7.1 Productivity in manufacturing (1973 = 100)
TABLE 7.4 Real value-added in manufacturing per person employed

(%)
1960–8 1968–73 1973–9 1978–87 1960–87
Large OECD
countries 4.5 4.6 2.6 3.7 3.9
Austria 5.2 5.5 3.5 3.0 4.2
Finland 4.2 3.6 2.5 4.8 3.9
Norway 3.3 — 0.5 (3.3)a —
Sweden 5.2 5.4 1.2 (2.8)a —
OECD Europe 5.0 5.4 2.9 2.6 3.9
a
One year missing.
distinct periods in which (labour) productivity growth outstrips real earnings

growth. (Norway, however, also experiences a period in the mid-1970s in
which earnings growth outstrips productivity growth but these trends get
reversed in the 1980s.) This pattern (which, incidentally, also occurs in the
USA and Japan) is a feature which does distinguish this group of social
corporatist economies from the behaviour of other European economies. As
Fig. 7.3 shows there are distinct periods in which real earnings growth has
been substantially reduced and lags behind productivity growth. This pat-
tern, which has been accompanied by a substantial recovery of profitability
in the 1980s (in Austria this happened only in the late 1980s), explains also
the distinct recovery of manufacturing investment in the 1980s in three of
FlG. 7.2 Real earnings and productivity in manufacturing (1973 = 100)
FlG. 7.3 Real earnings in manufacturing (1973 = 100)

the four economies (with the exception of Austria; see Table 7.5). Both
Finland and Sweden, which experienced a dramatic slump in investment in
the mid-1970s, show a remarkable recovery in the 1980s (much higher than
OECD Europe as a whole).
(c) Investment shares and profitability: Tables 7.6 and 7.7 on investment shares
(gross fixed capital formation as a percentage of GDP) and rates of return,
as well as Fig. 7.4 which plots profitability indicators and investment shares
TABLE 7.5 Real gross fixed capital formation (growth)

(%)
1960– –8 –73 1973–9 1978–87 1960–87
1968–
Large OECD
countries 6.3 5.8 1.5 2.3 3.9
Austria 5.6 8.1 1.2 0.7 3.6
Finland 2.5 8.7 –1.3 3.2 3.0
Norway 5.5 6.3 2.1 3.0 4.1
Sweden 5.0 6.5 –2.9 4.9 3.9
OECD Europe 5.8 5.2 0.2 1.1 3.0
TABLE 7.6 Gross fixed capital formation as a percentage of GDP
1960– –8 –73 1973–9 1978–87 1960–87

1968–
Large OECD
countries 20.5 21.9 22.1 20.7 21.2
Austria 26.4 27.2 26.4 23.2 25.7
Finland 26.6 26.2 27.4 24.5 26.1
Norway 28.0 27.4 32.9 26.1 28.7
Sweden 24.1 22.6 20.6 18.0 21.6
OECD Europe 23.0 23.6 22.4 20.2 22.2
TABLE 7.7 Rates of return in the business sector
(%)
Pre–1979 1975–9 1980–6
Large OECD
economies 24.2 19.8 18.8
Austria 23.6 17.4 15.4
Finland 17.0 12.8 13.3
Norway 14.9 12.4 11.5
Sweden 19.0 15.3 14.6
OECD Europe 19.8 16.9 16.3
Source: OECD, Economic Outlook (June 1989).
FIG. 7.4 Investment shares and profit ratios in OECD economies, 1960–1985
Notes: Investment shares = net capital formation/national income; profit ratios = operating
surplus/compensation to employees.
Source: OECD, National Accounts.
for all OECD economies, reveal another distinct feature of the social corporat-
ist economies: while rates of return in the business sector are rather low (with
the expansion of Austria) in comparison with OECD Europe, investment shares
are high (except for Sweden after 1973). This points to a particular charac-
teristic of the SC economies: these economies seem to be able to maintain
comparable or even higher investment activity compared to other OECD eco-
nomies while showing significantly lower rates of return or profit shares in
national income. One reason for this situation could be the particular struc-
ture of saving typical for these economies (see discussion in section 7.5.4).
Another reason could be compensating factors such as good industrial rela-
tions, expectations of stable growth, and continuity of macro-economic policy
direction. This conclusion is also borne out by a cross-country regression
analysis of investment in which investment shares over various subperiods are

regressed on profitability and growth indicators. For most of the time periods,
all the corporatist countries lie above the regression line, as is apparent from
Fig. 7.4.
In section 7.3, the abundance of labour was emphasized as one determin-
ant of growth, suggesting that social corporatism may be a co-operative solu-
tion for dealing with labour shortage. While the paucity of data does not
allow a very far-reaching analysis of the role of labour supply, it is certainly
of interest to note that a comparison of investment ratios reveals an even
stronger regularity if labour supply is used as a conditioning factor. Using
Kindleberger’s (1967) classification of countries into high and low labour
supply ones, we have investigated investment ratios within the low labour
supply growth group, to which the social corporatist economies of Sweden,
Austria, and Norway belong. Within the low labour supply growth group,
the investment ratios of the SC economies are clearly higher than those of
the non-corporatist low labour supply economies of Belgium, The Nether-
lands, and France; the former lie consistently above the OECD average
whereas the latter remain below it (see Fig. 7.5).
(d) Sectoral patterns of employment and productivity growth and implications for
long-run competitiveness of the tradable sector: Table 7.8 shows employment
growth in the manufacturing and service sectors as well as in public employ-
ment. We can see that the negative employment growth in the manufacturing
FIG. 7.5 Investment ratios in three groups of countries

Note: Investment ratios - investment (gross)/GDP.
TABLE 7.8 Employment in manufacturing, services, and government (%)

1960–8 1968–73 1973–9 1978–87 1960–87
Manufacturing
Large OECD
countries 1.7 0.8 –0.3 –1.0 0.3
Austria –0.7 1.8 –0.7 –1.5 –0.5
Finland 1.0 4.8 –0.1 –1.0 0.8
Norway 1.0 — –0.2 –1.1 —
Sweden 0.1 –1.3 –0.6 (–0.7)3 —
OECD Europe 0.6 0.7 –0.09 –1.6 –0.9
Services
Large OECD
countries 2.3 2.7 2.5 2.1 2.4
Austria 1.3 1.8 2.4 1.3 1.6
Finland 2.9 3.7 2.2 2.5 2.8
Norway 2.0 — 4.0 2.5 —
Sweden 2.0 3.2 2.8 (1.6)a —
OECD Europe 1.9 2.1 1.8 1.8 1.9
Government
Large OECD
countries 3.2 1.9 1.7 1.0 2.0
Austria 2.2 3.6 3.9 (1.9)a 2.9
Finland 4.5 4.7 4.7 2.8 4.1
Norway — 4.6 3.6 2.6 3.5
Sweden 5.2 6.3 4.9 (2.0)a 4.3
OECD Europe 2.3 3.5 2.4 1.3 2.2
a
Years 1978–89.
sector in these economies over the 1970s and 1980s was very much in line
with average OECD experience. The manufacturing sector did not react very
differently in contributing potentially to severe unemployment. However, fig-
ures on employment generation in services and particularly in public employ-
ment show much higher rates of employment growth than in the OECD in
general. Referring back to relative productivity growth in manufacturing (not
very different from the OECD European average, see Table 7.4) and services
(in general much lower than the OECD European average in the 1970s and
1980s, see Table 7.9), this confirms the picture obtained in our analysis: that
these economies attempted to restructure their tradable sectors (in most cases
successfully) so as to maintain (or even improve) their competitiveness while
using the sheltered sectors of the economy to avoid unemployment.
Table 7.10 presents figures for the trade balance as a percentage of GDP.
We can see that after a period of negative performance on the trade balance
in 1970s, there is in most cases a remarkable turn-round in the 1980s (with
TABLE 7.9 Real value-added in services (growth), and per person employed
(%)
1960––8 1968–73 1973–9 1978–87 1960–87
Growth
Large OECD
countries 5.2 9.6 3.4 2.9 4.0
Austria 4.0 5.5 3.3 2.0 3.5
Finland 4.6 7.4 2.8 3.8 4.5
Norway 4.6 4.6 3.3 2.5 3.7
Sweden 4.1 3.8 2.5 1.7 3.0
OECD Europe 4.5 4.9 3.2 2.3 3.6
Per person employed
Large OECD
countries 3.0 2.0 12.0 0.8 1.7
Austria 2.7 3.7 0.9 0.7 1.8
Finland 1.7 3.7 0.6 1.4 1.7
Norway 2.7 — –0.7 (–0.09)3 —
Sweden 2.0 0.7 –0.3 (0.06)a —
OECD Europe 2.8 2.9 1.4 0.6 1.8
a
Years 1979–89.
TABLE 7.10 Trade balance of a percentage of GDP (average)
1960–8 1968–73 1973–9 1978–87 1960–87

Large OECD
countries 0.6 0.6 0.1 –0.2 0.3
Austria –0.1 0.6 –0.7 0.2 0.0
Finland –1.1 –0.1 –0.7 0.9 –0.2
Norway –1.8 0.0 –4.7 4.9 –0.1
Sweden –0.1 0.7 –0.7 1.4 0.4
OECD Europe 0.0 0.6 –0.1 0.7 0.3
Austria somewhat lagging behind). While this indicator has to be used with
caution and interpreted in the light of terms of trade movements and relative
GDP growth in these economies relative to their trading partners, it does
point to a relatively successful performance in which a twin strategy of, at
the same time, maintaining or improving the competitiveness of the tradable
sector while keeping low levels of unemployment and maintaining public
services (see Table 7.11) far above the OECD European average was
pursued.
2
For an extensive discussion of structural adjustment in the tradable sectors in these
economies and for the presentation of various indicators of competitiveness, see Landesmann
in this volume.
238
MICHAEL LANDESMANN AND JUHANA VARTIAINEN
TABLE 7.11 General government claim on resources
Total outlays as%ofGNP Government employment share (%)
1960 1975 1980 1986 1960 1975 1980 1986

Large OECD
countries 28.8 37.7 38.4 40.0 11.5 14.7 14.6 14.7
Austria 35.7 46.1 48.9 51.9 10.5 16.4 18.2 20.5
Finland 26.6 36.1 36.5 42.3 7.7 14.6 17.8 20.8
Norway 29.9 48.4 50.7 52.0 — 19.3 21.9 23.3
Sweden 31.0 48.9 61.6 63.5 12.8 25.5 30.7 33.0
OECD Europe 31.5 44.1 45.8 48.9 10.4 15.4 16.7 18.1
Source: OECD (1989: table 5.2, p. 158).

7.7 CONCLUSION
This chapter has presented an overview of certain features of social corpor-

atist economies and longer-term economic development. We should like in
conclusion to discuss some of the challenges which lie ahead, particularly
with respect to the adjustment of social corporatist structures to the process
of closer European economic integration in the 1990s.
There were two aspects which we emphasized when discussing the emer-
gence of social corporatist institutions in a number of European economies
in the course of this century. One was the observation that in many countries
the emergence of social corporatism coincided with some crucial ‘catching-
up’ phase of the economy. The other was that they were (and are) small,
open economies.
In the present context of closer European integration we can perceive that
most northern and central European economies have converged in terms of
the degree of maturity of their industrial structures and, while there are still
some ‘technological gaps’ to be closed (e.g. in the case of Austria), most coun-
tries now compete more or less successfully on equal terms with other indus-
trialized economies.
They will continue to be relatively ‘small’ which will make them predes-
tined to be ‘niche’ producers in an international economy where intra-industry
specialization and trade are the main expression of the international division
of labour amongst advanced economies. However, closer political integra-
tion in Europe will limit the extent to which countries will be able to have
recourse to national economic policy instruments as a means of achieving
long-term competitive advantages.
7.7.1 Where does that leave social corporatism?

In our opinion the debate concerning social corporatism will shift over the next
few years to a discussion of political-economic structures at a transnational
level. The recent debate concerning the ‘Social Charter’ of the European
Communities has already shown this. While the main pressure for complete
integration of the European market by 1992 might come from business, that
debate has shown that, given the political and social history of the different
European economies, the consideration of the ‘social’ aspects of such an in-
tegration process cannot be neglected. An institutionalization (e.g. in the form
of a harmonization of European legislation) of social policy questions at a pan-
European level amounts to a Europeanization of social corporatist structures.
However, another aspect of such internationalization seems to have been
lagging behind such developments. It concerns the internationalization of
3
Austria has already submitted its application for full EC membership, and it is expected
that Sweden and Finland will follow suit in the course of the 1990s.
wage-bargaining arrangements or union policy in the wider sense at a

pan-European level. While in certain branches unions have developed cross-
country strategies to face multinational employers, such strategies are so far
few and not always very successful. Union organizations still seem to be very
nationally orientated and find it difficult to become active in the ‘trans-
national’ arena. There is, however, also a more structural aspect to this prob-
lem: Collective organizations gain their effectiveness from the close monitoring
of members’ interests and the flexibility they show to new demands which
emerge from changing circumstances. Such monitoring of demands is- by
definition, easier in smaller-scale organizations, and, equally, the flexibility
they can show towards the particularities of the interest groups they represent
is also in most cases inversely related to size.
The danger of organizational centralization at a high level (e.g. at a Euro-
pean level) thus carries risks for trade union organizations as far as the
representation of and adaptability of their actions to members’ interests is
concerned. On the other hand, the traditional rationale of trade union organ-
izations in capitalist market economies is that they enforce certain common
standards across competitors and these standards counteract the unbalanced
growth and segmentation processes which would otherwise emerge. This
reasoning behind the potential for organization of union and labour repre-
sentation at a pan-European level will continue to exist and the formulation
of the ‘Social Charter for Europe’ is an expression of it. Initiatives in this
direction are evidence of an international integration process of social
corporatist structures.
REFERENCES
ARROW, K. J. (1962), ‘The economic implications of learning by doing’, Review of

Economic Studies, 29: 155–73.
BOLTHO, A. (1982), ‘Growth’, in A. Boltho (ed.), The European Economy: Growth and
Crisis, London: Oxford University Press.
CALMFORS, LARS, and DRIFFILL, JOHN (1988), ‘Bargaining structure, corporation
and macroeconomic performance’, Economic Policy, 6: 14–61.
DUBOIS, P. (1978), ‘New forms of industrial conflict in 1960–74’, in C. Crouch and
A. Pizzorno (eds.), The Resurgence of Class Conflict in Western Europe since 1968, ii,
London: Macmillan.
ESPING-ANDERSEN, G. (1985), Politics against Markets: The Social-Democratic Road
to Power, Princeton, NJ: Princeton University Press.
GROUT, P. A. (1984), ‘Investment and wages in the absence of binding contracts: a
Nash bargaining approach’, Econometrica, 52: 449–60.
HYMAN, RICHARD (1978), ‘Occupational structure, collective organisation and in-
dustrial militancy’, in C. Crouch and A. Pizzorno (eds.), The Resurgence of Class
Conflict in Western Europe since 1968, ii, London: Macmillan.
redistribution in capitalism: efficient game-equilibria in a two-class neoclassical
KALDOR, NICHOLAS (1966), Causes for the Slow Rate of Economic Growth in the United
Kingdom, Cambridge: Cambridge University Press.
KlNDLEBERGER, C. (1967), Europe’s Postwar Growth, Cambridge, Mass.: Harvard
University Press.
LANCASTER, KELVIN (1973), ‘The dynamic inefficiency of capitalism’, Journal of
Political Economy, 81: 1092–109.
M.J0SET, LARS, (1986), Norden dagen derpa, Oslo: Universitaets förlaget.
OECD (1989), Economies in Transition: Structural Adjustment in OECD Countries,
Paris.
PEKKARINEN, JUKKA (1989), ‘Keynesianism and the Scandinavian models of
economic policy’, in Peter Hall (ed.), The Political Power of Economics Ideas:
Keynesianism across Nations, Princeton, NJ: Princeton University Press.
PLOEG, FREDERICK VAN DER (1987), ‘Trade unions, investment and employment: a
non-cooperative approach’, European Economic Review, 31:1465–92.
POHJOLA, MATTI (1984), ‘Union rivalry and economic growth: a differential game
PONTUSSON, J. (1987), ‘Radicalization and retreat in Swedish democracy’, New Left
Review, 165: 5–33.
REGINI, M. (1984), ‘The conditions for political exchange: how concertation
emerged and collapsed in Italy and Great Britain’, in J. H. Goldthorpe (ed.), Order
and Conflict in Contemporary Capitalism, Oxford: Clarendon Press.
SAUTTER, C. (1982), ‘France’, in A. Boltho (ed.), The European Economy: Growth
and Crisis, Oxford: Oxford University Press.
SCHARPF, F. W. (1984), ‘Economic and institutional constraints of full-employment
strategies: Sweden, Austria and West Germany 1973–1982’, in J. H. Goldthorpe
(ed.), Order and Conflict in Contemporary Capitalism, Oxford: Clarendon Press.
SEN, A. K. (1961), ‘On optimising the rate of saving’, Economic Journal, 71: 479–96.
------(1967), ‘Isolation, assurance and the social rate of discount’, Quarterly Journal
STRATH, B. (1988), The Politics of Deindustrialization, London: Croom Helm.
VARTIAINEN, JUHANA (1989), ‘Understanding Finnish monetary policy’, University
of Helsinki, Dept. of Economics (mimeo).
8
Industrial Policies and Social Corporatism
Michael Landesmann
8.1 INTRODUCTION
This chapter discusses the relationship between industrial policies and ‘so-
cial corporatism’. The notion of ‘social corporatism’ has been defined (see
e.g. Goldthorpe 1984; Schmitter and Lehmbruch 1979) as the tripartite
(business, labour, and the state) institutionalized involvement in the for-
mulation and execution of economic and social policies. The group of Scan-
dinavian countries (Sweden, Norway, Finland) and Austria have been
singled out as representing the most developed forms of corporatist arrange-
ments of this type at the present time and this chapter sets out to examine
the experience of these economies vis-d-vis the success or failure of the in-
dustrial (or structural) policies pursued in them, particularly over the 1970s
and the 1980s.
As we will see, the experiences even within this group of economies are
rather varied. They had in common with other advanced economies the
strong pressure to undergo industrial structural adjustment in the wake of
the two oil-price shocks, the lower growth environment of the world eco-
nomy, and the increased degree of international competition both from
NICs and from the increasing degree of international integration amongst
advanced economies. They also faced the additional pressure—which is
particularly strong in social corporatist economies—to pursue a policy to-
wards industrial structural adjustment which puts heavy emphasis upon
minimizing the social costs (and the even distribution of the gains) of such
adjustments. This led to an interesting mix of both defensive and construct-
ive policies (the former refers to policies designed to smooth the process of
structural adaptation, the latter to policies orientated towards the support
of new activities, skills, etc.) which were used with varying and changing
emphasis in the four economies and over the period examined in this
chapter.
The structure of the chapter is as follows: section 8.2 addresses the gen-
eral question of the relationship between industrial policies and social cor-
poratism, section 8.3 describes the evolution of industrial policies and of
structural change in the four economies in some detail, and section 8.4
summarizes some of the interesting differences which the country studies
reveal about the experiences of the four economies.
INDUSTRIAL POLICIES AND CORPORATISM 243
8.2 GENERAL QUESTIONS CONCERNING THE RELATIONSHIP

BETWEEN CORPORATIST STRUCTURES AND
INDUSTRIAL POLICY
8.2.1 Why is there such a strong emphasis in the economic literature

examining corporatist experiences on incomes policy and
on wage-setting institutions?
At first sight, the almost exclusive attention on the relationship between
corporatist institutions and one particular macro-economic variable, the
general level of wage settlements, seems odd for the following reasons.
Corporatist arrangements seem to be designed to operate in the longer
run, i.e. they are supposed to provide a framework in which the mediation
of specific group (or class) interests is achieved and in which medium- to
long-run social and economic targets can be met. In the recent economic
debate, however, the role of corporatist institutions has mostly been dis-
cussed within the context of short run macro-economic analysis.1 As is well
known, short-run macro-economic analysis today, as it was the case in the
inter-war period, focuses upon the labour market as an explanation for the
level of employment or unemployment and, in particular, upon the role of
wages in regulating the level of employment. It seemed natural for macro-
economists, who were interested in finding an explanation for the varying
experiences of different economies with respect to unemployment and the
growth of employment, to make differences in the wage-setting mechanisms
in these different economies responsible for the different unemployment
(employment) experiences. The relatively narrow focus in the economic liter-
ature on wages policies in different economies and the role of corporatist
institutions in affecting wage levels thus satisfied the needs of current (and
pre-war) macro-economic analysis.
Another reason why the issue of wage settlements has been in the
forefront of the analysis of the economic impact of corporatist institutions
is that wages could be seen as the most explicitly targeted variable in many
corporatist arrangements. Even if that can be agreed upon in the case of a
number of European economies, it must be recognized that the general level
(or the rate of growth) of wages can only represent an intermediate target
variable. The ultimate targets which are seen to be affected by this variable
are rather the overall level of employment (or unemployment), the rate of
inflation, the rate of growth, etc.
Apart from the fact that the relationships between the wage level and
these other, ultimate target variables are hotly debated in current macro-
economic analysis, it is quite clear that these variables are also affected by
1
See Bruno and Sachs (1985), Calmfors and Driffill (1988), Newell and Symons (1987),
Freeman (1988).
244 MICHAEL LANDESMANN
a host of other variables which are also subject to the influence of corporatist
arrangements. For example, it is difficult to predict, even in an orthodox
economic model, how wage settlements will affect the rate of inflation
without at the same time knowing how productivity growth is affected.
The point which is at issue here is the following: corporatist arrangements
are by their very nature slowly evolving longer-term historical features of
capitalist economies,2 which in many cases affect comprehensively the ways
in which economic processes evolve in these economies. A singular concern
with the impact of corporatist structures upon one particular variable, the
general level of wage settlements, and then indirectly, upon employment,
inflation, and growth, is thus seriously misleading.
In this essay, I will discuss the influence of corporatist arrangements upon
industrial policies and thus analyse their impact upon a host of other
variables and relationships, such as productivity growth, the sectoral ori-
entation of industrial growth, training, and research and development.
While these are the variables which affect the industrial growth process in
the long run, it seems none the less to be indispensable to consider these
variables if one is to understand and compare the employment, growth, and
inflation experiences of a number of European ‘social corporatist experi-
ments’ over the 1970s and the 1980s.
8.2.2 Why do industrial policies often occupy a secondary place in social

corporatist settings?
The first answer to this question could simply be that industrial policies are
in fact central to social corporatist experiments but they appear not to be
so and are overshadowed by other policies such as incomes policies.
There is—in my opinion—something to this argument: industrial policies
by their very nature attempt to influence economic performance in the long
run. They are also composed of a wide variety of heterogeneous instruments
which operate in very specific—often micro-economic—ways in different
sectors of the economic system. There is no reason to presume that they
do not affect macro-economic performances in a substantial way just as the
use of straightforwardly macro-economic instruments3 (such as exchange
rates, fiscal and monetary policy, and, indeed, incomes policies) might do.
But their effects are very difficult to assess quantitatively precisely because
of the longer-term nature of their impact and because of the heterogeneous
use and effects of these instruments. Macro-economic policy instruments,
2
See e.g. Maier (1984) for a historical account of the evolution of corporatist structures in
a number of advanced economies.
3
Nobody would deny, for example, that training and educational policies in the German
economy have had a marked effect upon its industrial performance and similarly one would
agree with respect to the role of Japan’s industrial policy in the post-war era.
including wages and incomes policies, on the other hand, are often used in
a discrete manner, frequently responding to situations of economic crisis,
and thus their effects are more easily discernible in the short term.
There is however also a deeper reason why industrial policies are not
perceived as belonging to the core of social-corporatist experiments.4 At
first sight it seems as if a consensus on industrial policy targets (such as
high industrial growth, increased competitiveness, etc.) would be easier to
obtain than, say, an incomes policy where the underlying conflictual situ-
ation (over income shares) is more apparent. Industrial policies are targeted
towards increasing national wealth as a whole and they thus open up positive
sum options from which everybody could gain.
In actual practice, however, industrial policies are designed to be specific,
i.e. directed towards particular industries, firms, regions, groups in the
labour market, etc., rather than general. Even in those cases in which they
are general (such as general tax allowances), they have a differential impact
upon different parts of and actors in an economy. Implicit in industrial
policy formulation and execution are therefore always trade-offs between
different groups, regions, industries, etc.
As such, industrial policies are less conducive to centralized agreements
(see also the discussion below on the effective use of industrial policy
instruments) or to an easy summary in the form of macro-economic target
variables (such as a general target of annual wage increases in the case of
incomes policies) around which a consensus on the direction of macro-
economic policies could emerge.
Furthermore, industrial policies have the same problem as all disag-
gregated policies and especially discretionary ones. They are prone to the
impact of particularistic group interests and, in the case of industrial
policies, of groups which want to preserve their current position in an
economy if they are threatened by processes of structural change. Industrial
policies which are meant to assist structural change could thus easily be
turned into structure-preserving policies as a result of the specificity and
possibility for discretion in the instruments employed and the political
processes involved. This possibility is reinforced by the fact that the estab-
lishment of new industries, the support of R. & D. and training programmes
in new directions, etc. only has ‘potential’ advocates in industry, while
existing activities, skills, research lines, etc. are often represented by very
vocal, well-entrenched interest groups. There is also an element of lack of
transparency in the use of a large set of heterogeneous policy instruments
which makes them more accessible to the lobbying tactics of particular
4
In other forms of corporatism, for example the technocratic form of corporatism which
characterized Japan’s post-war economic policy performance, industrial policies are seen as very
much at the core of the relationship between the corporate sector and the state (see e.g. Johnson
1982).
interest groups, often uncontrolled by the wider public scrutiny which more
‘visible’ policies are exposed to.
The difficulty which social corporatist experiments face with respect to
industrial structural change in general, and industrial policies in particular,
can also be expressed in another way. Social corporatism has the problem
not only of finding a consensus (or a mode of conflict resolution) between
classes, but also of building on a consensus within classes. At the least it
has to show that it can successfully provide the framework to mediate
between different segments of the same class and thus avoid open conflict.
Policies which speed up structural change, and which assist the uneven
process of industrial development, face difficulties in being recognized as a
basis for a consensus between the different segments of the same class. This
is not necessarily and not in all circumstances the case: if structural change
is perceived as indispensable for achieving industrial growth (as seems to
have been recognized by the Swedish trade union and labour movement
early on),5 and adjustment to structural change is perceived as leading to
an upward movement in real income positions, job content, etc. which will
be widely diffused, then such consensus support may be found.
There is, however, a problem with the time horizon of such policies:
industrial policies are not, in most cases, expected to have an immediate
effect upon growth prospects and the modernization of the industrial appar-
atus. To carry along the support or at least understanding of those groups
which stand to lose in the short run from structural adjustment processes
but could gain in the longer run (which, in turn, will depend on the
distribution of such gains) might be a difficult undertaking. Adjustment
assistance, such as temporary employment schemes, mobility support,
training for potential new jobs, is one way of assuring that the negative
aspects of structural change processes get weakened or, in a forward-looking
manner, provisions are made for a ‘new structure’ to emerge. The experi-
ence in different economies (such as Austria, Norway, Sweden) shows that
such measures are able either to support or to inhibit and even prevent
structural change from taking place.
We also know that—depending upon the time horizon and the scope of
such measures—these (adjustment-assisting) policies might absorb a large
(and often growing) fraction of the total funds available for industrial policy
purposes. In the case where the modernization process takes longer than
expected or is unsuccessful (sometimes because of international circum-
stances), the budgetary problems posed by policies which assist structural
adjustment may become very severe and, in some cases, unsustainable. (See
the case histories in section 8.3 of this essay.)
A further problem which arises in connection with industrial policy-making
5
See the early trade union documents LO (1951) and LO (1961, 1966).
in a corporatist setting is that—unlike other components of social corporat-

ism—industrial policies are often not formulated or executed in a tripartite
setting, i.e. with representatives of the state and both labour and management
being involved.
Many industrial policy measures are formulated and executed as a result
of the direct interaction between ministerial bureaucracies and the manage-
ment or owners of firms or their representatives, with little participation of
labour or their representatives. This also explains the much more central
role which industrial policies play in forms of corporatism other than social,
such as the technocratic form of corporatism pursued by Japan or France
in the post-war period.6 Often there is also a division of responsibilities for
different types of industrial policies (regional policies, manpower planning
and training, R. & D. support and high-technology schemes, the provision
of public utilities, and the running of nationalized industries, etc.) within
the overall ministerial bureaucracy. These different branches of public
policy-making bodies search for the agreement and support of different
interest groups with regard to different types of policies. Thus manpower
planning and training schemes might in fact include trade union repres-
entatives in the consultative process, while R. & D. support might not;
regional grants might involve local authorities, while various types of invest-
ment support might be dealt with exclusively in interaction with the
management of firms, etc. From the point of view of resolving conflicts,
such a decomposition of industrial policies into different parts might be a
good strategy. However, the same decomposition might not be conducive
to achieving broad support for a powerful and coherent industrial policy
programme (an ‘industrial strategy’) which would receive weight and ur-
gency similar to short-run macro-economic policies.
Industrial policies are thus caught in a dilemma: from the point of view
of obtaining maximum and broad-based support, industrial policies should
be consistently formulated, be as transparent as possible, and the rationale
of a particular package should be widely disseminated and made easily
understandable. From this point of view use should be made mostly of
non-discriminatory schemes (such as general tax allowances, access to
training for as wide a section of the potential labour force as possible,
general R. & D. support, etc.). However, from the point of view of max-
imum effectiveness, industrial policies should be ‘tailored’ to the specific
needs and circumstances of particular regions, industries, segments of the
labour force, etc. This would allow, for example, easier monitoring of the
fulfilment of the so-called ‘additionality’ requirement often imposed in
government schemes and would allow a concentrated use of funds where
they are most needed. It would also allow a more effective targeting of
6
See e.g. Pempel and Tsunekawa (1979).
instruments to specific requirements. This, though, would reduce the over-

all transparency needed for broad-based involvement and also increase the
feeling that discriminatory practices were being pursued and increase the
conflict potential between different interest groups.
There is a final point which has to be made with respect to the problem-
atic relationship between social corporatism and industrial policy. As men-
tioned before, industrial policies are aimed at affecting industrial structures
and industrial practices of an economy in the long run. They are thus
designed to have an impact upon the structural characteristics of an eco-
nomic system, while these are taken as given for the purposes of short-term
macro-economic policy-making which are not themselves aimed at changing
such characteristics. Industrial policies thus have the potential to affect the
way in which economic processes are carried out and can—potentially—
qualitatively transform an economic system. It is this potential of industrial
policies to transform the qualitative characteristics of an economic system
which has caused—in a number of historical circumstances—sharp political
conflicts over the formulation and implementation of such policies. Ex-
amples for such conflicts are the debates concerning nationalization after
both World Wars I and II in a number of European countries,7 the conflicts
which erupted with Mr Benn’s tenure of office as Minister for Industry in
the Labour cabinet of 1974/5 in the UK and his proposals concerning
corporate planning agreements, the conflicts over the Swedish wages funds
in the late 1970s and early 1980s, etc.8
While, in the longer run, capitalist structures often do exhibit remarkable
abilities to adjust to such reforms, at the point of actual implementation
such policies often lead to acute political and social conflicts which can
threaten existing social corporatist agreements even though these might get
re-established in a different form alongside or after the implementation of
such reforms.
8.3 SOCIAL CORPORATISM AND INDUSTRIAL POLICIES IN AUSTRIA,

FINLAND, SWEDEN, AND NORWAY
In this section we shall deal with the particular experiences of a group of

social corporatist societies vis-d-vis structural change and industrial policies
over the period of the 1970s and the 1980s. We shall first present some
statistical background information concerning macro-economic developments
in the different economies and then discuss in some detail their specific
experiences with respect to structural change and industrial policies.
7
See e.g. Weissel’s (1976) excellent discussion of the socialization debates in Austria after
World War I.
8
See e.g. Pontusson (1987) concerning the debates on the establishment of the Swedish-
wages funds.
Table 8.1 shows the percentage shares of the labour force employed in
agriculture, industry, and services and Table 8.2 presents some figures
concerning economic developments in the four economies over the 1973–9
and the 1979–86 periods. The figures in Table 8.1 show that the three
economies found themselves by the early 1970s still at substantially different
stages of industrial development. Austria and Finland still had a substantial
share of the labour force employed in the primary sector, while Sweden and
Norway had already embarked, by the early 1970s, upon a process of
‘deindustrialization’ (defined by a declining share of the labour force em-
ployed in manufacturing). That process of deindustrialization was most
rapid in the case of Norway (which became an oil economy over that
period), followed by Sweden. The share of the labour force employed in
manufacturing in Austria is high by international standards.
As regards developments over the subperiods 1973–9 and 1979–87
(Table 8.2), we can see that Sweden achieved relatively low rates of growth
of GDP over both periods and Austria’s growth decelerated substantially
in the second period, while Finland’s increased.
The growth rates of fixed capital formation show that both Sweden and
Finland experienced a dramatic slump in investment activity after the first
oil-price shock but recovered thereafter. Figures on unemployment rates show
that all countries achieved substantially lower rates than the OECD average
(although Finland’s unemployment rate jumped to about 6 per cent by 1979).
How these unemployment rates were achieved given the different demo-
graphic conditions, developments in participation rates, migration, etc. in
the different economies is discussed in other chapters of this volume. Here
we simply produce the figures on employment growth over the two sub-
periods (which show rather different experiences for the different eco-
nomies) and figures on labour productivity growth. The latter show the
rather strong decline in productivity growth over the 1970s and 1980s for
Austria, Finland, and Sweden relative to the pre-1973 period, a well-known
feature experienced throughout the OECD from which this group of social
corporatist economies did not escape. Productivity growth in Sweden was
particularly low.
We shall now turn our attention to the particular features of industrial
structural change and of industrial policies in the four economies with
emphasis on the 1970s and the 1980s.
8.3.1 Austria
Austria was engaged in reconstruction after World War II and was closely
linked (through trade and investment) with the West German economy. It
had few problems in the 1950s and 1960s in moving ‘upstream’ in the de-
velopment of its industrial structure: Austria was—relative to Germany—at
250
MICHAEL LANDESMANN
TABLE 8.1 Percentage of the labour force employed in agriculture, industry, and services in Austria, Finland,
Norway, and Sweden
Austria Finland Norway Sweden
1970 1975 1980 1986 1970 1975 1980 1986 1970 1975 1980 1986 1970 1975 1980 1986
Agriculture 18.9 12.5 10.5 8.6 22.6 14.9 13.6 11.0 13.9 9.9 8.3 7.2 8.2 6.4 5.6 4.2
Industry 40.3 40.9 40.3 37.8 34.6 36.1 34.6 32.0 37.2 34.4 29.1 26.7 38.4 36.5 32.2 30.1
Services 40.8 46.5 49.2 53.6 42.8 48.8 51.8 57.0 48.8 56.3 62.5 66.1 53.5 57.1 62.2 65.6
Source: Calculated from OECD (1988d).

251
INDUSTRIAL POLICIES AND CORPORATISM
TABLE 8.2 Developments in Austria, Finland, Norway, and Sweden, 1973–1987 (average annual percentage growth rates)
Growth of GDP Growth of employ– Unemployment Growth of labour productivity
Growth of gross
capital formation ment rates in the business sector
1973–9 1979–87
1973–9 1979–87 1973–9 1979–87 1973 1979 1987 Pre–1973 1973–9 1979–85
Austria 3.23 1.97 1.19 1.07 0.29 0.31 0.9 1.7 3.8 5.5 3.9 2.2
Finland 2.96 3.76 0.24 3.28 0.29 1.10 2.3 5.9 5.0 5.0 3.2 2.0
Norway 4.79 3.42 4.01 2.13 1.84 1.54 1.5 2.0 2.1 3.0 4.3 2.4
Sweden 2.13 2.00 –0.70 2.21 0.97 0.64 2.5 2.1 1.9 3.1 1.9 1.4
Source: All the figures have been calculated from OECD (1988e), with the exception of the labour productivity figures which were taken from
OECD (1987).
an earlier stage of industrial development, and the further evolution towards

a more mature industrial economy was relatively straightforward. It involved,
for example, the development of its chemicals and electrical engineering
industries and the lowering of its dependence upon natural-resource based
industries such as wood and wood-processing.
The social corporatist institutions which developed after World War II
did not reflect the need for an explicit industrial policy but rather provided
a framework in which the ‘social partners’ (labour and business) could
co-operate on incomes policy targets and on broad issues of macro-
economic policy.9 The strong presence of foreign businesses (multina-
tionals) in most advanced areas of industrial activity, as well as the small
and medium-sized character of the domestic privately owned business
sector, also made it difficult to develop an industrial policy framework a la
‘Japan Inc.’ or ‘Sweden Inc.’. The reasons why the large sector of publicly
owned enterprises, particularly in the basic industries such as steel, coal
mining, chemicals, and machinery production, was not more actively used
to further the modernization process of Austria’s economy (as was, for
example, the case in Finland) need further study.
A straightforward argument could be that things were going well in the
1950s and in the 1960s and little needed to be done in terms of selective
intervention by the government. Investment rates were high (also due to a
policy of generous early depreciation allowances) and the consensus on
macro-economic policy management also turned out to be robust with
respect to the impact of the first oil-price shock in 1973. Austria responded
with a strong continued commitment to maintaining full employment. It
used fiscal policy to that purpose but also preserved employment levels in
the nationalized industry sector in order to achieve this goal.10
Industrial structural change from the late 1960s onwards proceeded in
line with other industrial economies, but at a pace which did not allow
much further closing of the gap with more advanced economies such as
West Germany and Sweden. (See e.g. the comparisons made in the OECD
Economic Survey on Austria: OECD 1985a: 39–40.)
The OECD report mentions a WIFO study (WIFO 1982) in which
productivity differentials between Austrian and German industries are ex-
amined. Between 1964 and 1980, these differentials shrank for some indus-
tries (such as fabricated metal products, machinery and equipment, and
chemicals) but increased (for basic metals) or remained constant for others.
Overall, the productivity differential to Germany still remains important
(see also Aiginger 1982: 222). Compared with other OECD economies,
9
For a general discussion of the development of social partnership arrangements in Austria
after World War II, see e.g. Nowotny (1982).
10
See Table 8.A1 for information about employment and productivity development in
Austria’s nationalized industries.
Austria experienced a sharp deceleration of relative productivity growth in

the basic metals and chemicals industries after 1973. In other industries,
Austria compared favourably with productivity developments abroad.
Austria’s trading position towards other advanced economies still shows
a high export dependence upon basic and semi-finished goods. The share
of high-technology products in its exports is still low, and not increasing
rapidly (see Table 8.3). Austria’s spending on R. & D. is very low compared
to the OECD average and, although the ratio of R. & D. to GDP is
increasing, it is not increasing at the rate it does for some of the other
countries (particularly Finland) discussed in this chapter (see Table 8.4).n
Patent applications, while being relatively high in Austria for historical
reasons, have actually fallen over the 1970s and the 1980s (see Table 8.5).
As mentioned above, a distinct feature of Austria’s industrial structure is
the strong presence of foreign-owned companies particularly in advanced
industries, such as electrical engineering, and also the large share of publicly
owned companies.12 These two together account for a large share of total
R. & D. spending in Austrian industry. The small to medium scale of
Austrian-owned private businesses is a factor explaining the comparatively
low overall spending on R. & D.
TABLE 8.3 Share of high-technology products in exports of

manufacturing in Austria, Germany, and Sweden
1970 1975 1980 1982

Austria 7.6 7.6 8.9 8.7
Germany 13.6 13.4 14.8 16.0
Sweden 10.6 11.6 12.0 12.9
Source: OECD (1985a: table 13, p. 41).
TABLE 8.4 R. & D. expenditure as % of GDP in Australia, Finland and

Norway
1970 1975 1980 1985

Austria 0.61 0.92 1.10 1.27
Finland 0.78 0.91 1.07 1.53
Norway 1.10 1.34 1.27 1.53
OECD average 2.12 1.99 2.07 2.45
Source: OECD (1987: table 24, p. 46).
11
See also the detailed figure presented by E. Volk in s. 12 of the WIFO Report (WIFO
1987).
12
In addition to the publicly owned companies under the control of the OEIAG, the state
holding company, which accounted for about 13% of industrial employment in 1983, there are
many companies controlled by state-owned banks which account for an additional 10% of
industrial employment (see OECD 1985a: n. 24; see also the discussion in Aiginger 1988).
TABLE 8.5 Total patent applications in Austria, Finland, Norway, and Sweden
1965 1970 1975 1980 1983
Austria 2,714 2,267 2,525 2,345 2,388
Finland 819 861 1,164 1,356 1,719
Norway 870 938 752 716 825
Sweden 4,814 4,343 4,042 4,126 4,331
Source: Englander et al. (1988: table 3, p. 16).
The nationalized sector is concentrated in the basic industries such as

mining, basic metals, chemicals, and machinery production, while Austrian-
owned private firms are to be found particularly in textiles and food produc-
tion. As we mentioned earlier (see also n. 11 above), these two parts of die
economy behaved very differently over the different subperiods since 1973.
Between 1973 and 1978, in the wake of the first oil-price shock, employ-
ment in the nationalized sector increased by about 5 per cent, and fell in
the private incorporated business sector by 12 per cent (see Aiginger 1982:
223). After 1982, on the other hand, the employment losses in the national-
ized sector were much more severe. We will return to these more recent
developments, after describing certain features of Austrian industrial policy.
The institutional basis for an active industrial policy is poorly developed
in Austria. The ‘Paritatische Kommission’ occupies a central position in
Austrian economic policy formation, but it concentrates its policy recom-
mendations on questions of wages and prices and general macro-economic
stability. Until the early 1980s the awareness of a need for an industrial
policy framework was very low, although a number of studies had been
commissioned by the Beirat fur Wirtschafts- und Sozialfragen (another
powerful social partnership institution; see Guger in this volume) in the late
1970s which reflected a concern about the lack of a properly formulated
and executed industrial policy, particularly in the areas of technology policy,
direct export, and investment support. Existing investment support was
criticized as being too broadly diffused, and as favouring capital-intensive
and large enterprises.13 Direct industrial support has often proved (given
the political mechanism of its allocation) to be structure preserving in terms
of both industrial and regional employment patterns. Otherwise, Austrian
industrial policy is heavily biased in favour of indirect investment promo-
tion in the form of accelerated depreciation and investment allowances, low
interest credits from the (largely publicly owned) banking system,14 and the
formation of tax-free allowances.
13
See e.g. Haas and Wehsely (1977) and Kager and Kepplinger (1981).
14
OECD (1985a) estimates that about 30% of bank credits to industry are subsidized in
Austria.
There is some evidence to suggest that the massive financial support given
to the nationalized industries during the 1970s, which has undoubtedly
helped to preserve employment levels in these industries over that period,
has not however been used efficiently to develop effective training schemes
(as happened, for example, in Sweden over the same period). These could
have helped to equip the work-force with additional skills and supported a
drive to restructure the enterprises and make them viable in the longer run.
By the late 1970s and early 1980s the budgetary situation and the in-
debtedness of the nationalized industry sector had deteriorated to such an
extent that the overall policy direction started to change quite dramatically
from the early 1980s onwards.15
The government commissioned a comprehensive review of the entire
system of subsidies and transfers, a dramatic reorganization of the Austrian
state holding company (OEIAG) was initiated, and the privatization of some
of its subsidiaries was envisaged.16 There is a plan to step up R. & D.
spending to a level of 1.5 per cent of GDP and stricter criteria are being
set up to select projects for industrial policy support. Finally, a revision of
the existing law regulating business activity (the Gewerbeordnung) is under
way which is designed to strengthen the small and medium-sized enterprise
sector by removing barriers to entry and exit.
TABLE 8.6 Unemployment in Austria
1974 1980 1982 1986

a
Unemployment rates 1.5 1.9 3.7 5.2
Males 0.8 1.6 3.9 5.1
Females 2.2 2.3 3.5 5.2
Unemployment durationb
6–12 months 10.2 14.1 14.1
12 months or over 4.2 5.7 12.6
Long–term unemploymentc
Youths 3.3 10.3d 16.4
Adults 7.9 10.3d 16.4
Older workers 22.1 21.6d 30.4
a
Registered unemployment as a percentage of the dependent labour force.
b
Percentage share in total unemployment.
c
Percentage in total unemployment within age and gender groups.
d
1983.
Source: OECD (1988a: tables 18 and 19, p. 44).
15
Government aid in the form of subsidies and capital transfers had reached a peak of 5%
of GDP in 1981.
16
See OECD Economic Survey 1987/88 on Austria (OECD 1988a: 18–19).
Unemployment rates have risen dramatically, with youth unemployment

deteriorating steadily since 1980, and the incidence of long-term unemploy-
ment is also increasing fast. The fall in employment levels since 1980 was
particularly dramatic in the nationalized industry sector as Fig. 8.1 shows.
The latest OECD report on Austria (OECD 1989a), however, testifies to
a positive outcome of the reorganization process initiated in the early 1980s
in the nationalized industry sector and of various measures in the area of
technology policy. The OECD report records that the restructuring meas-
ures adopted by the nationalized industry sector have led to a dramatic
turn-round of the revenue situation of the OEIAG which was forecast to
close 1988 with a gross operating surplus on current transactions of AS 1
bn., as compared with a loss of more than AS 10 bn. in 1987: ‘About half
of the improvement can be attributed to the favourable demand and price
situation while the remainder reflects rationalisation and consolidation
efforts’ (OECD 1989a: 25).
By the mid-1980s, therefore, the Austrian government17 had embarked
upon a set of policies different from those pursued in the 1970s. These include
fiscal consolidation, tax reform, reorganization of the publicly owned com-
pany sector (including the privatization of some segments of it), and initiat-
ives in the areas of technology policy and the small and medium-sized
enterprise sectors which seem to have won the support of the majority of
FIG. 8.1 Employment in Austria

Note: Indices 1970 = 100.
Source: OECD (1988a: diagram 12, p. 53).
17
A coalition government is presently in power which includes the two major parties, the
Socialist Party and the People’s Party.
the political constituents of Austria’s social partnership arrangement. There

is evidence to suggest that the government has acted effectively on the basis
of that new consensus. Opposition to these developments within the ‘Grand
Coalition’ has manifested itself in stronger support for parties outside
that coalition such as the Greens and the right-wing Austrian Freedom
Party (FPO).
8.3.2 Finland
Finland has, in issues concerning industrial structural change and industrial
policy, turned out to be a success story. Figs. 8.2–8.4 reproduce some of
the distinguishing features which characterize Finnish manufacturing per-
formance over the 1970s and the 1980s: high overall growth in manufac-
turing output (Fig. 8.2), a very high investment rate (very similar to that in
Japan, see Fig. 8.3), and a dramatic increase in R. & D. spending (Fig.
8.4). In the mid-1970s, following the first oil crisis, Finland experienced,
however, a dramatic slump in investment, low GDP growth, and very low
employment growth. In the late 1970s and throughout the 1980s, GDP
growth was high, investment boomed, and there were declining unemploy-
ment rates (see Tables 8.1, 8.2 above for statistical information).
We also showed in Table 8.1 that Finland had, in 1970, still about 23
FIG. 8.2 Growth of manufacturing output (index 1960 = 100)

Source: OECD (1988ft: diagram 9, p. 31).
FIG. 8.3 Investment rate in manufacturing, 1970–1986

Note: Fixed investment as a share of manufacturing value-added.
Source: OECD (1988*: diagram 11, p. 36).
per cent of its labour force employed in agriculture; this proportion had
come down, in 1986, to 11 per cent. Over the period under analysis, Finland
could thus be regarded as having been ‘catching up’, particularly in relation
to countries such as Sweden and West Germany. Commentators (see e.g.
Pekkarinen 1988, or OECD 19886, particularly diagram A3, p. 85) have
remarked that Finland had, until recently, a very specialized structure of
industrial exports, more so than any of the other countries surveyed in this
chapter. It has in the past, therefore, been particularly vulnerable to chan-
ging world market conditions. The notion of a ‘devaluation cycle’ has been
coined (see Korkman 1978, mentioned in Pekkarinen 1988) to describe the
impact of an ever-recurring balance-of-payments constraint upon macro-
economic policy formulation in the 1960s and 1970s.
While Finland’s exports immediately after World War II consisted almost
entirely of forestry products, over the last two decades its export structure
has become more diversified. Finland is a very open economy and Finnish
industrial performance is heavily influenced by external factors. A study
undertaken by the economics division of the Bank of Finland (reported in
OECD Report 19886: 43) suggests that the geographical and product com-
position of Finnish exports, which in the early 1970s still had a negative
impact on Finnish export performance in Western markets, became much
FIG. 8.4 R. & D. expenditure and fixed investment in Finnish manufacturing

constant prices
Source: OECD (19886: diagram 12, p. 39).
more favourable in the 1980s. A constant market share analysis confirms

that both the product composition and die geographical distribution of
Finnish exports no longer contribute negatively to export growth (see
Table 8.7).
The relatively strong emphasis in Finland upon supply side policies, and
relative neglect of Keynesian demand management policies, can thus, in
our view, be explained, firstly, by the relatively recent industrialization
process of the Finnish economy and, secondly, by the strong incentive to
TABLE 8.7 Constant market share analysis of Finnish exports

(yearly average change, current prices)
1972–6 1977–81 1982–6
Regional structure –1.1 –2.6 0.1
Product patterns –0.6 –1.0 0.9
Residual factor –2.3 5.4 0.8
Change in market share –4.2 1.8 1.9
Source: OECD (19886: 43, table 12).
diversify its industrial and export structure in order to become less vul-
nerable in its external economic relationships.
Different from Sweden, but similar to Austria, Finland has a strong presence
of state-owned companies employing about 12 per cent of the industrial
labour force and accounting for about 30 per cent of industrial investment.
These are run on a commercial basis and are largely independent of direct
government supervision.18 State-owned companies are to be found particu-
larly in the highly capital-intensive branches such as mining, energy, and
heavy industry. State initiatives were also instrumental after World War II
in establishing the metal industry, which has developed into an important
export industry.19
Investment rates in manufacturing have traditionally been very high in
Finland (see Fig. 8.3) and these have been seen for a number of reasons20
as being essential to maintain and improve competitiveness. Industrial
policies in the post-war period were particularly directed towards achieving
high rates of investment (through depreciation and investment allowances).
Since the 1970s, however, there has been a strong emphasis upon produc-
tivity-enhancing policies (training and a very wide access to educational
facilities, as well as support for research and development). Finland’s spend-
ing on research amounted to 1.6 per cent of GDP in 1986. During the
period 1980–7, finance for research in industry ‘has grown at a real rate of
5 per cent per annum, while funds set aside for target oriented research
have soared by about 20 percent in real terms each year’ (Ministry of Trade
and Industry 1987: 27). There have been a large number of government
schemes supporting research and development activity in Finland. The
Ministry of Trade and Industry Report (1987: 37) mentions ‘tax reforms
to promote research activities, the removal of restrictions on the deprecia-
tion of research equipment, the provision of opportunities to use investment
reserves, and the introduction of an extra deduction for research’.
Finland also undertook in the 1980s a strong expansion of spending on
education and training. Spending directed particularly towards supporting
research activities in higher education expanded at a rapid rate.
The rate of growth of total factor productivity (TFP) decelerated in die
1970s but it accelerated again from the late 1970s. For example, manufac-
turing TFP accelerated from 1 per cent in 1973–9 to 3.3 per cent in 1979–86
(see OECD 19886: table 7, p. 32). Apart from the high R. & D. spending
and high output growth, the OECD report (19886: 35) also mentions
a number of other factors which explain the favourable productivity
18
See OECD (19886:52).
19
See Pekkarinen (1988: 17).
20
‘Geographic and climatic conditions as well as the sparsity of the population also provide
an explanation for the relatively high investment ratio’ and ‘The capital-intensive structure of
Finnish industry together with environmental conditions require a higher level of investment
than is necessary in competitor countries’ (in: Ministry of Trade and Industry 1987: 33, 34).
performance: ‘Wages in Finland have risen considerably faster than in

competitor countries, and hence the need for rationalisation and the pres-
sure for scrapping of inefficient capital equipment have increased; from
1979 to 1986, industrial wages rose on average about 2]A percentage points
faster per year than in the OECD area. Due to the higher investment ratio
and the investment boom of 1979–80, when in most other OECD eco-
nomies investment decelerated, the age structure of the Finnish capital stock
is now rather favourable.’ An important and distinct feature of Finland’s
technology policy is its emphasis on decentralized and regional supply of
technological information (see Ministry of Trade and Industry 1987: 28
and 42–9). Technology centres (TEKES) have been established in a num-
ber of regional centres and, in addition to traditional measures of regional
assistance, efforts have been made to develop forms of aid, in the areas of
marketing, product development, and know-how. A number of measures have
been introduced to support the medium- and small-sized enterprise sector
(such as the establishment of a new Business Development Department in
the Ministry of Trade and Industry) which provide a network of company
services and use the Regional Development Fund (RKA) to that effect.
As regards agricultural policy, Finland has an agricultural sector which
is one of the most subsidized in the OECD area (see OECD 1989e and
OECD 19896: 56). Subsidies amounted to around 70 per cent of receipts
in 1987. ‘As a result of this policy, Finland has a degree of self-sufficiency
which is significantly above 100 percent for dairy products and meat and
above 100 percent for grain’ (OECD 19886: 56).
While the Finnish government continued to take a protective stance
vis-d-vis the agricultural sector, Finnish manufacturing had—as mentioned
above—a very strong productivity performance in the 1980s, its R. & D.
soared, it can rely on a good educational and training infrastructure to
develop the skills of its labour force, and has consequently pursued a
successful strategy of upgrading its production activities and the product
composition and regional orientation of exports. It has become a successful
‘niche producer’ in particular specialized areas of industrial production and
has become a world leader in technology in branches such as forest
products, paper machines, and specialized shipbuilding. Recent OECD
estimates of the income elasticity of demand for Finnish exports was 1.86
compared to an income elasticity of demand for imports of 0.99.21 Finnish
patent applications have risen dramatically (see Table 8.5 above) and its
business sector has recently embarked upon a strategy of internationaliza-
tion22 (as have Swedish producers over a much longer period).
21
See OECD (19886: table A4, p. 95). The estimation period for these countries was 1975
to 1986.
22
OECD (19886: 37): ‘In the period 1981–86, the volume of investment abroad increased
8.3.3 Sweden
The need for industrial structural change was accepted early on in Sweden.
The economic policy orientation of both the trade union movement (the
‘LO’ or ‘Landsorganisationen’) and the employers’ federation (‘SAF’ or
‘Svenska Arbetgivarforeningen’) included a commitment to full employ-
ment but also to the necessity of industrial structural change.
The reasons for this are quite clear. Relative to other European powers
(England, France, and even Germany) Sweden was a late industrial
developer and thus had to conceive early on of a strategy to promote exports
and substitute imports (see e.g. Heckscher 1954; Lindgren 1979; and
Lundberg 1985). It was also a feature of the Social Democratic policies
after World War II that rising living standards and an extension of social
welfare were dependent upon industrial structural change and technological
advance (see LO 1951; 1961, 1966).
A number of famous schemes were devised to combine social policy tar-
gets with the requirements of structural change and modernization. The
most famous of these is the one associated with the name of the LO economist
Gösta Rehn (see Rehn 1948, 1952; and the discussion in Lundberg 1985).
The principal idea behind this scheme was to put pressure upon the
business sector to modernize and keep wage inflation down while at the
same time both supporting the modernization process and ensuring full
employment through an active labour market policy. The logic of the Rehn
scheme was the following: a strong adherence of the (centralized) trade
union movement to a solidaristic wages policy—which ensures that the same
pay is given for the same type of work regardless of profitability and
productivity in the various branches and firms—would ensure that strong
pressure was put upon the low-productivity but also the labour-intensive
sectors and firms of the economy. This would either lead to the shedding
of labour in these sectors or firms and/or lead to an improvement in their
efficiency. A necessary complement of such a policy to ensure full employ-
ment was therefore an active labour market policy which allowed the
efficient transfer of labour to more productive jobs, which supported labour
mobility through retraining schemes, through subsidies for geographical
transfers, through efficient labour exchanges, etc. Finance for these labour
market policies (but also for the rapidly expanding social services, etc.)
would come from successively increased taxes including payroll taxes. If the
squeezed profits led to lower private savings, this could be compensated for
by increased public savings (see also the contribution by Kosonen on the
‘Nordic system of accumulation’ in this volume).
There is evidence that the effect of the implementation of this scheme
by 30 per cent annually and reached 19 per cent of the value of industrial investment in 1986 .. .
during the three years 1983–86 Finnish companies bought 179 foreign companies.’
(under conditions of an open trade policy) was an industrial structural
transformation away from labour-intensive activities such as textiles, cloth-
ing, and footwear industries, and towards expanding export industries such
as engineering and capital-intensive branches such as iron and steel, ship-
yards, and petrochemicals.23 Public expenditure on labour market policies
became a hallmark of the ‘Swedish model’ and they rose from about 1 per
cent of GDP in the early 1960s to about 4 per cent in the 1980s. There is
also evidence to suggest that the Rehn model contributed to the strong
overall shift of employment out of industrial employment and into public
sector employment.24
The period between the mid-1970s and the early 1980s was a difficult
one for the Swedish economy. After the investment boom of 1973–4, the
volume of investment dropped by about 40 per cent and investment stagna-
tion lasted until about 1983.25 There was also a dramatic increase of state
support for ailing industries. In Fig. 8.5 we can see the increase in direct
industrial subsidies over the period 1975 to 1979. Industrial subsidies as a
whole increased from about 1.3 per cent of GDP in 1970 (or about 4.9 per
cent of value-added in mining and manufacturing) to about 3.6 per cent of
GDP in 1978 (or 16 per cent of value-added in mining and manufacturing).
Bo Carlsson estimates that over the period 1970–9 75 bn. Skr. were paid
out in industrial subsidies. This compares with spending on labour market
measures of 50 bn. Skr. over the period 1970–9.
As we can see from Fig. 8.5, the measures which increased the most over
the period are firm-specific measures, i.e. direct measures largely designed
to save ailing firms. In an attempt to compare the industrial subsidy
programme across a number of European economies, Carlsson also finds
that the level of firm-specific subsidies was particularly high in Sweden and
a very high proportion of these were on rescue operations. Carlsson’s
statistics are reproduced in Table 8.8.26
23
See Lundberg (1985: 19).
24
F. Scharpf in Goldthorpe (1984: 266–7) presents some comparative statistics on public
sector employment. According to OECD statistics (see OECD 1982), the share of public sector
employment in total employment rose from 20.6 to 29.8% between 1970 and 1980. In Austria
and West Germany the corresponding figures were increases from 14.1 to 18.5% in Austria and
increases from 11.2 to 14.7% in West Germany. According to ILO ( Yearbook of Labour Statistics)
figures, the share of employment in ‘community, social and personal services’ increases in
Sweden from 29 to 35% between 1972 and 1980, in Austria from 23 to 24%, and in West
Germany from 19 to 23%. Scharpf concludes (p. 267): ‘By either measure, Sweden now has
the highest share of publicly financed employment of any country in Western Europe’.
25
Lundberg describes the dramatic increase in public spending and public sector employ-
ment over that period: ‘The volume of total public-sector expenditures (including transfer
payments and gross investment expenditures by general government) was expanding by 6 per
cent per annum (1970–1982), attaining close to 70 per cent of GNP while GNP in real terms
was stagnating (1 per cent growth rate). The share of total employment in the public sector
increased from 20 per cent to 33 per cent between 1970 and 1983’ (Lundberg 1985: 20 n. 11).
26
For details on the compilation of these figures and the caution needed in making
international comparisons in this area, the reader is referred to Carlsson (1983).
Source: Carlsson (1983: 4).
TABLE 8.8 Firm-specific subsidies by receiving sector in Sweden, 1977–1979

Million No. of
Skr. employees
In relation to
Value- Total (thousands)
added wage bill
(%) (%)
(of skr.)
Shipyards 9,094 72.3 120.2 282
Steel 4,880 35.6 33.4 92
Forest-based 2,012 11.2 12.4 32
Mines 1,666 32.9 41.2 100
Textile and clothing 1,125 9.5 11.6 21
All manufacturing 20,238 6.9 8.5 21
Note: Sum of grants and loans; paid-out amounts; credit guarantees not included.
Source: Carlsson (1983: table 2, p. 11).
In Sweden, three-quarters of the firm-specific subsidies in 1977–9 went

to the shipbuilding and the steel industries. Over the 1977–9 period, the
shipbuilding subsidies amounted to about 120 per cent of the total wage
bill, and in 1978 and 1979 the subsidies exceeded the value-added
produced in the industry.
While over the period 1975 to about 1982 the emphasis of industrial
policy measures was thus clearly to help ailing industries to restructure their
operations, since 1982 the character of industrial policy support has
changed quite dramatically. There was a move away from rescue and
restructuring support and towards the promotion of high-technology areas,
export support, and support to the small firm sector (see Fig. 8.6). At this
stage it might be appropriate to relate some of the specific features of
Sweden’s industrial structure to the restructuring processes which took
place in Swedish industry over the 1970s and 1980s. Table 8.9 shows the
breakdown of Swedish exports by industrial branch. It shows clearly that
the traditional comparative advantage areas of Sweden’s industrial structure
(forest products, shipbuilding, crude ore) have declined both in terms of
shares in Swedish exports and in terms of world market shares (where the
FIG. 8.6 Structure of industrial support in Sweden

Note: Adjustment/contraction = temporary support + sectoral support; growth/renewal
supports for R. & D., regional development, exports, and small enterprises.
Source: OECD (1985c: diagram 16, p. 46).
TABLE 8.9 The structure of Swedish exports

Percentage of total exports World export share %
1960 1970 1977 1961 1976
Declining
Pulp 13.9 8.5 5.8 12.7
Wood and lumber 8.8 5.9 4.7 14.7 7.0
Ores 8.4 4.0 2.2 7.5 3.0
Stable
Paper and
paper products 10.2 8.8 9.0
Metal products 3.0 3.4 4.0
Shipbuilding 7.0 5.2 6.0 18.0 15.0
Steel 8.0 9.0 7.0 4.5 4.2
Textile and clothing 1.5 1.9 2.6
Growing
Chemicals and
pharmaceutical 2.9 4.1 4.8 1.3 1.5
Mechanical engineering 13.7 17.5 18.2 3.7 3.4
Electrical engineering 4.4 7.4 8.1 2.7 3.6
Transport equipment
(excl. shipbuilding) 5.4 9.7 11.3 1.4 2.5
Other 15.8 18.4 21.5
Source: Boston Consultancy Group (1979), Main Report, Exhibits 1 and 2.
declines are particularly strong). Electrical engineering, transport equip-

ment excluding shipbuilding, chemicals and pharmaceutical, and, to some
extent, mechanical engineering exhibit, on the other hand, both increasing
world market shares and increases in their share in the value of Swedish
exports.
These last four branches show an extremely high degree of industrial
concentration. The top twenty-eight companies in the Swedish engineering
and chemical-related sectors accounted for 77 and 55 per cent of total
exports and employment respectively in these industries in 1977. (See e.g.
Boston Consulting Group (1979), Main Report, Exhibit 19.)
Public ownership in these sectors, on the other hand, is very low as Table
8.10 demonstrates. State ownership was concentrated (apart from tobacco
and beverages which are rather special cases) in shipbuilding, mining, and
a particular segment of the steel industry.
Given this background concerning the particular features of Sweden’s
industrial structures, the restructuring process over the 1970s and 1980s
had the following characteristics. On the one hand, there were the tradition-
ally dominant industries such as wood and wood processing, iron ore and
TABLE 8.10 State ownership in industry in Sweden in 1976a

Tobacco 100 Wood products and paper 10
Shipbuilding 73 Pharmaceutical 10
Mining 60 Chemicals, plastics 9
Beverages 58 Engineering 3
Commodity steel 53 Transport (excluding
Textiles, clothing, leather 14 shipbuilding) 3
a
% of total sector sales generated by state-owned companies.
Source: Boston Consultancy Report (1979).
steel, and shipbuilding which faced very serious problems over the period
and which saw a significant reduction in their activities, levels of employ-
ment, and importance in Swedish exports.27 They also absorbed a colossal
amount of public direct subsidies over the period. State ownership in this
sector (apart from forestry products and paper) is high. On the other hand,
the other important group of manufacturing industries (electrical engineer-
ing, chemicals and pharmaceutical, transport equipment other than ship-
building, and mechanical engineering) pursued over the same period a
relatively successful path of modernization, attempting to withstand the
much fiercer climate of international competition. This group of industries
was traditionally dominated by very large firms (given Sweden’s size overall)
and these firms’ activities were highly internationalized.
The OECD report on Sweden (OECD 1985c) testifies that R. & D.
spending in Sweden is highly concentrated in large firms: 95 per cent of R.
& D. takes place within firms with more than 500 employees and of the
firms which report R. & D. activities the ten largest account for two-thirds
of all R. & D. spending. We also know that R. & D. expenditure has risen
dramatically over the 1970s and 1980s,28 while the level of total investment
has stagnated or contracted, and government-funded R. & D. as a percent-
age of total R. & D. spending has fallen over the 1970s.29 The increase in
R. & D. spending must therefore have been largely ‘in house’ within the
corporate sector (exceptions to this might have been defence-related firms).
The strategy towards modernization of the Swedish corporate sector
seems to have been remarkably successful over the 1970s and 1980s and
has proceeded in the following directions:
27
See the book by Strath (1988) describing the restructuring process in the Swedish and
other European shipbuilding industries.
28
See OECD (1985c: diagram 14, p. 42).
29
Government-funded R. & D. as a percentage of total industry R. & D.: 1973: 35% (total),
15% (excluding defence); 1975: 27% (total), 15% (excluding defence); 1977:29% (total), 13%
(excluding defence) (Boston Consultancy Group 1979: Appendix B, p. 52). The same source
also presents some figures concerning public R. & D. support to the large Swedish firms which
are rather low by international comparison.
• We have already remarked on the very rapid growth of R. & D.

spending; R. & D. spending in Sweden in 1983 was about 2.3 per cent
of GDP (in 1987/8 about 3 per cent of GDP), for Germany 2.2 per
cent, for the USA 2.0 per cent, and for Japan 1.6 per cent. The OECD
report (1985c) mentions that R. & D. now represents almost 10 per
cent of value-added in industry and industry finances 90 per cent of its
research.
• There has been much progress in introducing modern techniques of
production. The OECD report mentions that the use of robots and
computer-aided design and computer-aided manufacturing methods
(CAD/CAM) is among the highest in the world.
• Swedish firms have successfully attempted to obtain market
dominance in well-defined sectors and niches of the world market (see
Boston Consultancy Group Report 1979: appendix 13).
• Finally, large Swedish firms have continued their process of inter-
nationalization at a very rapid rate: the OECD report (1985c) estimates
that the twenty most multinationalized Swedish firms, which account
for about 80 per cent of employment in Swedish foreign subsidiaries,
have invested some 70 to 80 per cent of their total investment since
1982/3 abroad.
8.3.4 Norway
Norway is an example of a social corporatist economy which, partly owing
to macro-economic circumstances and partly owing to the particular ori-
entation of its structural policies, faced especially severe structural adjust-
ment problems which it has—so far—not overcome.
Norway experienced in the late 1970s/early 1980s a dramatic process of
deindustrialization, as the figures in Table 8.11 testify. We can see that over
the late 1970s/early 1980s there was a marked shift of employment out of
manufacturing and most of the employment absorption took place in the
general government (public services) sector. The most important explana-
tion for this process of deindustrialization is that Norway became an oil
TABLE 8.11 Sectoral employment shares in Norway

(%)
1965 1975 1987
Primary sector 17.5 9.3 6.5
Secondary sector 36.4 34.3 26.5
Tertiary sector 46.1 56.4 66.9
of which:
General government 13.8 19.3 27.3
Source: OECD (1988a: table 16, p. 48).
economy. The key variable by which North Sea oil affected the other sectors
of the economy was, as in the case of the United Kingdom, the exchange
rate. While other social corporatist economies such as Sweden and Finland
took advantage of exchange rate devaluations as a key mechanism by which
greatly to improve the prospects of the export sector, this instrument was
denied to Norway as it became an oil economy. By the mid-1980s, the oil
and gas sector produced about 18 per cent of GNP, 20 per cent of fixed
capital formation took place in this sector, its exports constituted about
one-third of all Norwegian exports, and the traditional (non-oil) exports
were only able to finance less than two-thirds of non-energy imports.
Relative labour unit costs of the manufacturing sector deteriorated strik-
ingly in the mid-/late 1970s (see Fig. 8.7) owing particularly to the ex-
change rate appreciation but also owing to unfavourable relative labour
productivity growth; relative wage rates did not, until the mid-1980s, move
out of line with development in competitor countries; after this, however,
they rocketed.
There was a dramatic deterioration in the terms of trade of the exposed
as against the sheltered sector of the economy. ‘From 1975 to 1982 the
FIG. 8.7 International competitiveness in manufacturing in Norway

Source: OECD (1989c: diagram 10, p. 51).
output price of the exposed sector relative to that of the sheltered sector
fell by 15 per cent. With money wages rising uniformly across sectors, this
has led, given productivity developments, to a profit squeeze of the exposed
sector’ (OECD 1985c: 1). These developments led to a sharp drop in the
early 1980s of manufacturing investment, and manufacturing production
virtually stagnated after 1973.
The revenues from North Sea oil permitted, however, high GDP growth
(about 4 per cent per year between 1972 and 1983 compared to an average
of 2.5 per cent per year in the OECD as a whole). The gains from oil were
‘largely redistributed to firms and households through subsidies and trans-
fers, higher public spending and—more secretly—tax reductions’ (OECD
1985^: 85).
On the industrial policy side, regional support has always played an
important role in Norwegian structural policy. Manufacturing production
is geographically dispersed and some 100 ‘one company towns’ account for
about 10 per cent of manufacturing production. Plant closures thus would
have had a marked regional impact.
Public subsidization—already high by international standards—rose
dramatically in the late 1970s when it amounted to close to 8 per cent of
GDP (see Fig. 8.8). Just as in some of the other Scandinavian countries—
particularly Sweden—industrial support was particularly targeted towards
shipbuilding (see Table 8.12). Government assistance was provided in a
number of forms: direct subsidies, subsidized loans, reduced taxes and
FIG. 8.8 The degree of subsidization: an international comparison (Norway,

Sweden, the FRG, and the United States)
Source: OECD (19856: diagram 8, p. 25).
TABLE 8.12 Industrial support to shipbuilding and state companies in Norway

Shipbuilding _______________ State companiesa_____________
Percentage Percentage Percentage Percentage
of gross of operating of gross of operating
value-added surplus value-added surplus
1973 5.2 22.1 4.2 15.9
1977 10.5 65.3 4.1 37.7
1980 20.2 149.6 5.1 16.9
1981 33.1 157.7 9.7 61.2
1982 28.1 131.0 26.2 271.9
1983 24.7 141.8 12.7 38.6
1984 22.3 205.7 19.2 42.5
1985 29.2 190.9 11.8 40.0
a
Value-added are operating surpluses in sectors where state companies are
particularly concentrated: mining, manufacture of ferrous and non-ferrous metals,
and manufacture of metal products (except machinery and equipment). The
figures thus understate the importance of industrial support to the state enterprise
sector.
Source: OECD (1988c : table 19, p. 67).
duties, and the covering of losses of public enterprises. The textile industry
was furthermore supported through import restrictions.
The OECD report (OECD 1988c: 68) assesses the public support to
industry in the following way:
Government support to industry has typically aimed at keeping employment in
industries which otherwise would no longer be viable. These industries are among
the most capital-intensive in the economy. Consequently, most of the support can be
said to have subsidised the use of capital rather than labour. As an employment measure,
the support policy must be judged as inefficient, in particular as the labour market
has generally been unduly tight. However, the location of the targeted industries and
the regional problems which would follow abrupt closure of individual industries may
explain why a policy of easing the strain of adjustment had been pursued.
By the mid-1980s the employment losses in the declining industries were
none the less severe, similar to Sweden although less than in the UK. Apart
from government support to industry, income support to farmers had been
very high. Until the mid-1970s, agricultural policy had been directed
towards increasing production and this was achieved particularly through
import restrictions. Since then, income support has become the prime target
which includes an attempt both to reduce income differentials between
farmers (for geographical or efficiency reasons) and to narrow the gap
between the real incomes of farmers and those of industrial workers. The
first of the two objectives required direct subsidies while the second could
be achieved through price support policies. The budgetary cost of this

policy was very high. Public expenditure related to the implementation of
agricultural policy amounted to about 5 per cent of total government outlay
and more than 6 per cent of central government spending (in the US 4 per
cent, in Japan 3 per cent, and 2.3 per cent in the European Community;
see OECD 1988a: table 16, p. 63). Production subsidies as a percentage
of gross value-added in agriculture amounted to more than 40 per cent in
the early 1980s.
TABLE 8.13 Employment trends in three ailing sectors

(Norway, Sweden, Germany, France, United Kingdom)
Iron and steel Textiles Shipbuilding
1985 1984 1985
(1974 = 100) (1974 = 100) (1975 = 100)
Norway 53 65a 29
Sweden 62 62b 11
Germany 66 60 45
France 52 64a 57
United Kingdom 31 42 25c
a
1983
b
1982
c
1984
Source: OECD (1988 c : table 17, p. 66).
As regards the present state of competitiveness of Norwegian industry,

there is evidence to suggest that—similarly to the UK economy—the shift
of resources out of the traditional tradable sector in the late 1970s and early
1980s due to North Sea oil has taken a heavy toll. With the fall in oil prices
a rather dramatic deficit on the current account of the balance of payments
has opened up (the deficit on the current account amounted to 6.5 per cent
of GDP in 1986; see OECD 19896: 39). There are also signs of a deterior-
ating deficit on the balance of service trade. This mainly reflects the fall in
net shipping revenues due to a decline in the shipping fleet under the
Norwegian flag.
The decline in the exchange rate from 1989 onwards improved competit-
iveness somewhat (in fact the volume of Norwegian exports grew at relat-
ively high rates) but it will take time to reverse the effect of the uneven
growth of the oil and the non-oil sector of the Norwegian economy which
took place in the late 1970s/early 1980s. The long-run viability of the
tradable sector has also not benefited to the same degree as it did in Sweden
and Finland, from a targeting of structural policies towards modernization
and the development of new activities. Rather, these policies were defensive
in character and most government assistance concentrated on shipbuilding,

mining, and basic metals industries. Structural adjustment and the reori-
entation of industrial policies are hence delayed relative to countries such
as Sweden and Finland.
8.4 LESSONS FROM THE FOUR COUNTRIES’ EXPERIENCES WITH

RESPECT TO INDUSTRIAL STRUCTURAL CHANGE AND INDUSTRIAL
POLICIES IN THE 1970S AND THE 1980S
As with most other OECD economies, the group of social corporatist

economies surveyed in this chapter had to face severe problems of structural
adjustment in the 1970s and 1980s. However, the developmental state of
their economies (including certain features of their respective industrial
structures) as well as the historical traditions of their corporatist arrange-
ments were different. In the following, we will argue that the above two
types of factors were both important in explaining the attitudes of social
corporatist institutions in the four economies towards the use of industrial
policy instruments in the 1970s and the 1980s.
As argued above, Finland was, as far as the general level of industrial
development was concerned, still a relatively ‘backward’ economy in the
1950s and 1960s and its industrial structure (and particularly its export
structure) was highly specialized. Just after World War II, the state took it
upon itself to attempt to broaden the industrial base of the Finnish
economy, and public ownership in a variety of industries (particularly
mechanical engineering) was used to that effect. But Social Democratic
traditions were much weaker in Finland than in the other Nordic economies
(see Pekkarinen 1988) and, hence, the emphasis upon the role of the state
in guaranteeing full employment and extending welfare services was weaker.
Public ownership in certain areas of industrial activity was seen as a prag-
matic step to promote industrial development rather than to ‘socialize’
sections of a capitalist economy. The weakness of the external balance and
the strong dependence upon natural-resource-based export commodities
was continuously felt in the post-World War II period and gave rise to sharp
economic fluctuations (in the form of a sequence of ‘devaluation cycles’;
see discussion above). There was thus a strong incentive to improve the
export capacity of the Finnish economy.
Finland’s response to the world-wide developments of the 1970s and
1980s can be seen as a continuation of its attempts to promote industrial
development, to move ‘upstream’ and broaden its export base, and to
continue to develop an educational and training infrastructure to enable it
to do so. The relative weakness of its Keynesian tradition prevented as
strong a concern with maintaining full employment levels as was the case
in the other countries of our sample (although compared to average OECD

performance in this regard, Finland did quite well). The publicly owned
enterprise sector continued to operate along commercial lines and did not
act as the employment absorber of last resort. Being closely linked to
economies with a more mature and broadly based industrial structure
(Sweden, Germany), Finland had to move into specialized product areas if
it wanted to export, and this tendency helped it to continue to embark on
a successful ‘market niche’ strategy in high value-added areas of interna-
tional export markets in the 1970s and 1980s. Even the ‘problem sectors’
such as shipbuilding and steel were already more specialized in Finland
than in, say, Sweden, so that the burden of restructuring of these sectors
was much lighter.
Austria had, over the post-World War II period, built up a strong, institu-
tionally reinforced basis for ‘social partnership’ co-ordination of economic
policy formation and execution. Its economic policy orientation centred
around what came to be known as ‘Austro-Keynesianism’. This involved a
strong commitment to maintain full employment relying on the ‘social
partners’ (labour and business) and on a hard-currency policy (the Austrian
Schilling remained closely linked to the D-Mark) to control inflationary
pressures. As argued in section 8.3 above, the awareness in Austrian
economic policy-making circles of the need for an explicit industrial policy
was very weak until about the early 1980s. Economic management was
almost entirely centred around the use of aggregate macro-economic
policies and a strong emphasis on wages and incomes policies. The experi-
ence of fast economic growth and generally rising living standards after
World War II did not generate any incentive to change that economic policy
orientation or to modernize the institutional basis of the social partnership
arrangements existent in the Republic.
Austria responded to the post-1973 situation with a continued strict
adherence to its previous policy targets and continued to maintain low
unemployment rates and low inflation. The high degree of unionization and
the particularly strongly organized unions in the nationalized industries
(steel, heavy mechanical engineering, chemicals) assured that these sectors
did not react in the same manner to the new international circumstances
(international excess capacity, strong competition from NICs and other
advanced economies facing low demand) as they did in other advanced
industrial economies. Employment levels in the nationalized industry sector
were held up; however, the ‘breathing space’ enjoyed by these industries
through heavy public subsidization was not efficiently used to reorganize
production lines, retrain the labour force, and change managerial practices.
The experience of the Austrian nationalized industrial sector over the period
1973 to 1982 is an example of an important failure of social corporatist
arrangements.
Although it had all the back-up from the state and operated in a relatively
non-hostile political environment, it did not manage to combine the twin
goals of modernizing its operations while adhering to the socially desirable
target of contributing to employment maintenance over this period. The
reasons for this failure deserve further careful study, especially since it did
seem to have contributed substantially to the change in the political and
ideological climate in the early 1980s and to the shift in economic policy
orientation. The sharp contraction of employment levels in the nationalized
industry sector in the mid-1980s has been described in section 8.3. It was
also argued that while the strong presence of a nationalized industry sector
could have contributed to the formulation of a consistent and effective
industrial policy programme, and so could the available institutional appar-
atus of the social partnership organizations (see also Haase 1982), the
ownership structure in industry as a whole is rather fragmented (with a
strong presence of foreign-owned enterprises in many modern industrial
activities and Austrian-owned firms mostly in the small to medium-sized
sector). This might have contributed to difficulties in developing and co-
ordinating an industrial policy effort. However, in the most recent period,
the awareness of industrial policy issues and of the importance of further
modernization of Austria’s industrial structure has increased appreciably
and one has to wait to see in which way the social corporatist arrangements
are going to embrace this new orientation and with what success.
Sweden had a very difficult transition in the 1970s and 1980s towards
what seems now a relatively successful survival of the ‘Swedish model’. In
this chapter we have argued that what made this transition relatively suc-
cessful was, to an important extent, the ‘dual structure’ of the Swedish
economy. For historical reasons, Sweden has developed a highly concen-
trated and strongly internationally orientated corporate sector which was
able to modernize its operations in the face of fierce international competi-
tion in the 1970s and 1980s. The state involvement in the restructuring
processes of the ‘ailing’ industrial sector (particularly shipbuilding and steel)
was very great, and commentators (see IUI 1981) have argued that the
heavy subsidization programme over the 1970s contributed substantially to
the slow growth performance of the Swedish economy over that period.
However, in spite of slow overall economic growth, the social policy features
of the ‘Swedish model’ seem to have been maintained and even
strengthened over that period (low unemployment with rising participation
rates particularly of women, good training and retraining programmes, good
public services, etc.), while the corporate sector of the Swedish economy
has built up a strong competitive position. From the early-mid-1980s, the
profitability of the Swedish business sector improved, the devaluation in
1981/2 and the growth of demand in export markets supported export
performance; investment growth was rapid and in particular growth in
intangible investment (in R. & D., marketing, etc.) was very high. Improved
output and employment growth allowed the dismantling of various selective
support schemes and subsidies (see OECD 1989d) and also the scaling
down of various labour market measures. The government’s budgetary
situation improved dramatically over the period 1986–8.
There is evidence of a strong process of internationalization by the large
Swedish companies and there is a question whether this might weaken the
domestic base in the longer run. Net direct investment abroad has been
boosted additionally by the prospects of an integrated European market.
The pressures towards internationalization are now felt throughout the
group of countries considered in this chapter and will have a significant
impact upon changes in the industrial organization of these economies over
the coming years.
The macro-economic circumstances faced by Norway in the 1970s and
the 1980s were quite different from the other three economies. Owing to
the arrival of North Sea oil, Norway achieved high real income growth in
the 1970s, and employment generation and investment performance were
strong. However, Norway also experienced in that period a strong process
of ‘deindustrialization’ (as indicated by a large shift of employment out of
manufacturing) and devaluation was not an instrument available to support
the non-oil export sector of the economy (as it was in Finland and Sweden);
rather, manufacturing investment and profitability suffered from a strong
appreciation of the currency which was reversed only in the mid-1980s.
Since overall income growth was high and the budgetary situation very
favourable (owing to high tax income from North Sea oil), structural
policies were used mainly to support declining industries (particularly ship-
building) and the traditional regional policy orientation of subsidizing
geographically dispersed manufacturing activity and the very high support
for agriculture were maintained. Unlike Sweden and Finland, Norway did
not pursue with rigour a policy of infrastructural support (R. & D., train-
ing, etc.) and of modernization of its manufacturing sector until the early-
mid-1980s when the oil price fell and a dramatic weakening of the trade
balance emerged. At present Norway is still struggling to secure competitive
foundations in its non-oil tradable sector in order to regain external equi-
librium with high economic growth. Rather as in Austria, the move from a
defensive orientation of industrial policies towards offensive modernization
measures is imposed upon Norway at a time when circumstances (particu-
larly the budgetary situation and foreign debt position) have become more
difficult.
TABLE 8.A1 Economic performance of state-owned industries in Austria compared to industry as a whole (Index: 1973 = 100)
INDUSTRIAL POLICIES AND CORPORATISM

1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983
Employment
State-owned industry, total 100 104 103 102 101 99 100 99 96 951 911
Excluding oil and mining 100 104 102 102 101 101 102 101 97 93 89
Total industry 100 100 94 93 94 92 92 93 91 87 83
Production
Total industry 100 105 97 104 108 111 119 122 120 119 120
Productivity
Total industry 100 105 103 112 115 120 130 132 132 137 145
Estimate.
Source: OECD (1985a : table 5, annex).
277
REFERENCES
ABELE, H., NOWOTNY, E., SCHLEICHER, S., and WINCKLER, G. (1982), Handbuch
der osterreichischen Wirtschaftspolitik, Vienna: Manzsche Verlags- und Univers-
itatsbuchhandlung.
AlGlNGER, K. (1982), ‘Industrie’, in Abele et al. (1982).
------ (1988), ‘Industriell-gewerblicher Sektor’, WIFO (mimeo).
Boston Consultancy Group (1979), A Framework for Swedish Industrial Policy, Stock-
holm: Liber Forlag.
Basil Blackwell.
CARLSSON, B. (1983), ‘Industrial subsidies in Sweden: macroeconomic effects and
an international comparison’, Journal of Industrial Economics, 32: 1–23.
ENGLANDER, A., EVENSON, R., and HANAZAKI, M. (1988), ‘R&D, innovation and
the total factor productivity slowdown’, OECD Economic Studies, 1: 7–42.
Clarendon Press.
HAAS, E., and WEHSELY, H. (1977), ‘Die direkte Investitionsforderung 1948 bis
1976’, Wirtschaft und Gesellschaft, 3: 229–47.
HAASE, K. (1982), ‘19 Thesen fur eine osterreichische Strukturpolitik nach japani-
schem Vorbild’, Wirtschaftspolitische Blatter, 1.
HECKSCHER, E. (1954), An Economic History of Sweden, trans. G. Ohlin, Cambridge,
Mass.: Harvard University Press.
IUI (1981), The Industrial Subsidy Program and its Macroeconomic Impact, Stockholm.
JACKSON, P., and SiSSON, K. (1976), ‘Employers confederations in Sweden and the
significance of industrial infrastructure’, British Journal of Industrial Relations, 14:
306–23.
JOHNSON, C. (1982), MITI and the Japanese Miracle: The Growth of Industrial Policy,
1925–1975, Stanford, Calif.: Stanford University Press.
KAGER, M., and KEPPLINGER, H. (1981), ‘Investitionsforderung in Osterreich’,
Schriftenreihe des Ludwig Boltzmann Institutes fur Wachstumsforschung No. 6,
Vienna.
KORKMAN, S. (1978), ‘Exchange rate policy, employment and external balance’,
Bank of Finland B: 33, Helsinki.
LiNDGREN, H. (1979), Corporate Growth: The Swedish Match Industry in the Global
Setting, Stockholm: Liber Forlag.
LO (1951), ‘Trade unions and full employment’, presented to the 1951 LO Congress.
----- (1961, 1966), Economic Expansion and Structural Change: Research Report Sub-
mitted to the 1961 Congress of Landorganisationen Sverige (The Swedish Federation
of Trade Unions), trans. T. L. Johnsten, London: Allen & Unwin.
LUNDBERG, E. (1985), ‘The rise and fall of the Swedish model’, Journal of Economic
Literature, 23: 1–36.
MAIER, C. (1984), ‘Preconditions for corporatism’, in Goldthorpe (1984).
Ministry of Trade and Industry (1987), Industrial Policy Review: Finland, Helsinki.
NEWELL, A., and SYMONS, J. (1987), ‘Corporatism, laissez-faire and the rise in
unemployment’, European Economic Review, 31: 567–614.
NOWOTNY, E. (1979), ‘Verstaatlichte Unternehmen als Instrument der
Beschaftigungspolitik in Osterreich’, Zeitschrift fur offentliche und gemein-
wirtschaftliche Unternehmen, 2: 252–72.
----- (1982), ‘Institutionen und Entscheidungsstrukturen in der osterreichischen
Wirtschaftspolitik’, in Abele et al. (1982).
OECD (1982), Employment in the Public Sector, Paris.
----- (1985a), Economic Survey 1984/85 on Austria, Feb., Paris.
----- (19856), Economic Survey 1984/85 on Norway, Paris.
----- (1985c), Economic Survey 1984/85 on Sweden, May, Paris.
----- (1987), ‘Total factor productivity’, in Economic Outlook, Dec: 39–48.
----- (1988a), Economic Survey 1987/88 on Austria, Paris.
----- (19886), Economic Survey 1987/88 on Finland, Paris.
----- (1988c), Economic Survey 1987/88 on Norway, Paris.
----- (1988d), Labour Force Statistics, 1966–1986, Paris.
----- (1988e), OECD Economic Outlook, Paris.
----- (1989a), Economic Survey 1988/89 on Austria, Paris.
----- (19896), Economic Survey 1988/89 on Finland, Paris.
----- (1989c), Economic Survey 1988/89 on Norway, Paris.
----- (1989d), Economic Survey 1988/89 on Sweden, Paris.
----- (1989c), National Policies and Agricultural Trade: Country Study: Finland, Paris.
PEKKARINEN, J. (1988), ‘Keynesianism and the Scandinavian models of economic
policy’, WIDER Working Paper No. 35, Helsinki.
PEMPEL, T. J., and TSUNEKAWA, K. (1979), ‘Corporatism without labour? The
Japanese anomaly’, in Schmitter and Lehmbruch (1979).
PONTUSSON, J. (1987), ‘Radicalization and retreat in Swedish democracy’, New Left
Review, 165: 5–33.
REHN, G. (1948, 1952), ‘The problem of stability: an analysis of some policy
proposals’, in Wages Policy under Full Employment, ed. R. Turvey, London: W.
Hodge.
SCHARPF, F. (1984), ‘Economic and institutional constraints of full-employment
strategies: Sweden, Austria and West Germany 1973–1982’, in Goldthorpe
(1984).
SCHMITTER, P., and LEHMBRUCH, G. (1979) (eds.), Trends towards Corporatist
Intermediation, Beverly Hills, Calif.: Sage.
STRATH, B. (1988), The Politics of Deindustrialization, London: Croom Helm.
WEISSEL, E. (1976), Die Ohnmacht des Sieges: Arbeiterschaft und Sozialisierung nach
dem Ersten Weltkrieg in Osterreich, Vienna: Europa Verlag.
WIFO (1982), Perspektiven der osterreichischen Industrie, Vienna.
------(1987), ‘Die internationale Wettbewerbsfahigkeit Osterreichs’, Osterreichische
StrukturberichterstattungKembericht 1986, ii, Vienna.
9
Exchange Rate Policy and Employment in Small
Open Economies
Sixten Korkman
9.1 INTRODUCTION
Employment in a large and relatively closed economy is often thought to

be mainly demand-determined in the short run, in which case the causal
significance of the real wage may be limited. For small open economies, by
contrast, there is widespread agreement that the real wage matters (at least
in the medium to long run). Traded goods may be sold on international
markets even though domestic demand is insufficient, as the price elasticity
of demand facing producers in a small economy can be taken to be high.
A reduction in the real wage should in these circumstances stimulate produc-
tion and employment through a number of routes: by changing relative
prices and switching world demand towards domestic output, by raising
profitability and encouraging increased utilization of existing capacity, by
inducing investment, etc. There is a presumption that large-scale and per-
sistent unemployment in a small open economy is classical in the sense of
being associated with a real wage and a real value of the currency which
exceed the levels compatible with full employment at external balance.
The main proximate determinants of the real wage are the money wage,
world market prices, and the exchange rate. Policy analysis of the real wage-
employment nexus should thus be concerned with two sides of the same
coin, i.e. with wage (and price) formation as well as with exchange rate changes.
It is the interaction between wage developments and exchange rate policy
which is crucial.
This chapter approaches the aforementioned issue from the perspective
of central bank policy, mainly as pursued in the Nordic countries (Denmark,
Norway, Sweden, and Finland). I will start by enumerating in section 9.2
some important changes that have taken place in the environment of cen-
tral bank policy in the past decade. I will also characterize what seems to
be the dominating doctrine of the role of monetary policy in a small eco-
nomy. As will be pointed out, this doctrine is not a fully convincing one.
Section 9.3 will then outline alternative monetary regimes or exchange rate
policy systems and their consequences for inflation and employment. Par-
ticular emphasis is paid to the macro-economic significance of alternative
rules or commitments on the part of the central bank. Section 9.4 discusses
EXCHANGE RATE POLICY AND EMPLOYMENT 281
the relation between exchange rate policy and corporatism. Section 9.5
makes some observations on the institutional choices open to small
European countries and the EMS option, in particular. Section 9.6 concludes.
9.2 MONETARY POLICY: CHANGING ENVIRONMENT

AND SHIFTING PERCEPTIONS
The role assigned to monetary policy has changed in all countries since the
early 1970s. Underlying these changes are shifts in policy perceptions and
priorities as well as a major transformation of the economic and financial
landscape. Some of the main changes from the point of view of small eco-
nomies like the Nordic ones may be outlined as follows:
1. International economic developments have been characterized by
severe real shocks and financial turbulence. Oil crises and stagflation
in the 1970s were followed by high real interest rates and disinflation
in the 1980s. The key currencies of the world economy have not only
been flexible or floating, but also volatile and misaligned. For long
periods they seem to have evolved quite independently from economic
fundamentals. The unstable international environment has been asso-
ciated with increased macro-economic imbalances and adjustment
problems as well as increased micro-economic risks.
2. Fiscal policy in many European countries, in particular, has been
constrained by large financial imbalances in the form of deficits in the
government budget and/or in the current account. In Denmark and
Sweden, for instance, there was widespread agreement in the late 1970s
and the first half of the 1980s that the room for fiscal policy action
was insignificant because of the dangers of the debt trap. Declining
fiscal policy ambitions tended to increase the importance attached to
monetary policy as a tool for demand management. For the same rea-
sons, monetary policy has been pursued under conflicting pressures
related to external and internal balance, i.e. to the foreign exchange
reserves and the current account, on one hand, and the inflation-un-
employment constellation, on the other hand.
3. Money and capital markets have expanded enormously and admin-
istrative controls of financial markets have been dismantled.
Deregulation in Scandinavian countries and elsewhere has concerned both
domestic interest rate determination and bond issues as well as foreign ex-
change transactions and capital flows. The interest rate mechanism has re-
placed credit rationing as the vehicle for equilibrating demand and supply
in loan markets. A particularly noteworthy feature of financial developments
is integration across national borders, which has reduced the autonomy of
monetary policy.
282 SIXTEN KORKMAN
The changes in the financial environment have been associated with

corresponding shifts in the perceived role of monetary policy. There is quite
a difference between the views that were held on this issue in Scandinavian
countries during the decades after the war, and what seems to be the con-
sensus view or the ‘official doctrine’ in the late 1980s.
The traditional view was that the central bank should maintain low and
stable interest rates and fixed exchange rates. Capital flows were regulated
and interest rates were set administratively at below market-clearing levels.
Credit rationing was widespread and monetary policy was predominantly
passive.
During the 1980s, by contrast, market forces have come to the forefront.
As mentioned above, rationing and regulation have been replaced by flexible
interest rates and central bank operations in open markets. Furthermore,
the size and interest sensitivity of (deregulated) capital flows have increased.
The emerging or established view on the role of monetary policy in the
new circumstances is very simple: it states that there is no room for an
independent policy to be pursued in a small economy. Interest rates should
be allowed to be determined by markets so that the balance of payments
remains in equilibrium at the given exchange rate. This view takes for
granted that capital flows should not and, in any case, cannot be regulated,
and that domestic and foreign financial assets and debts will then be close
substitutes in domestic and foreign portfolios.
Another cornerstone of the officially held doctrine is that the authorities
should make no exemptions to the fixed exchange rate policy. (Earlier it
was considered permissible and even recommendable to change the ex-
change rate in conditions of a so-called fundamental disequilibrium.) The
resulting view that monetary policy is impotent is then easily under-
standable, since free capital flows and fixed exchange rates are indeed a
combination which minimizes the scope for monetary policy and makes the
level of interest rates and financial conditions in the small economy exter-
nally determined.
There are two main reasons for being somewhat sceptical as regards the
sustainability of the (permanently) fixed exchange rate regime in small
economies. First, with fiscal policy being as inflexible as it is, there is not
necessarily an instrument to contain the current account deficit and thus
to maintain external equilibrium. The interest rate will preserve equilibrium
in the overall balance of payments, but may allow net capital inflows to
finance even sizeable current account deficits, which may then be reversed
only at exceedingly high interest rates associated with speculative attacks
on the currency and financial distress. Monetary policy under fixed ex-
change rates cannot be used to regulate the level of demand or the current
account.
A second observation is that many countries have in fact repeatedly
departed from the exchange rate regime that they officially adhere to. In
particular, the Nordic countries undertook more than twenty devaluations
in the late 1970s and the early 1980s. One could reply to this that neither
Sweden nor Finland (or Denmark) has devalued since 1982, and that this
marks the start of a new era (while Norway is slipping behind). For Sweden,
however, it should be noted that the currency has been effectively
depreciated since 1985 because of the large share of the dollar in the
Swedish currency basket. Another conceivable reason why there have not
been any more devaluations may be the fact, interesting in itself, that the
1982 devaluation seems to have been so effective that no new devaluation
has so far been needed. A complementary interpretation would point to the
favourable international environment which has prevailed in recent years.
The status of exchange rate policy is still an unresolved issue in these
countries. The official doctrine is that fixed exchange rates are the basic
norm to which the economy and policies have to adapt themselves. How-
ever, the question is whether this doctrine is credible and, even more
important, whether it is sensible and desirable: should the monetary
authorities stick to the fixed exchange rate with uncompromising vigour, or
should exchange rate changes be used actively to promote favourable
macro-economic developments?
9.3 THE RELATION OF EXCHANGE RATE POLICY TO INFLATION

AND EMPLOYMENT
An assessment of exchange rate policies should be based on a number of

criteria such as the prevalence of various shocks, political priorities, the size
of foreign trade elasticities, the strength of wage responses to price changes,
and the character of the mechanism of wage formation in general. In the
following I shall first outline briefly the (familiar) effects of an isolated
exchange rate change, before turning to the significance of alternative
exchange rate policy systems or regimes.
9.3.1 Effects of a change in the exchange rate
Assume that the currency is devalued in a situation where there is a deficit

in the current account and unemployment. What are the most important
links in the chain of effects thus initiated?
1. Price competitiveness will improve and the real wage will fall. Price
competitiveness may refer to the price ratio of domestic to foreign
goods and/or of traded to non-traded goods. The real wage will fall
because of the increase in the domestic currency price of imported
284 SIXTEN KORKMAN
or traded goods, given at least short-run rigidity of the nominal wage

rate.
2. Improved competitiveness should increase export and decrease import
volumes and thus strengthen the real trade balance. The market share
responses may, however, take time to materialize, and the initial effect
on the value of the trade balance may therefore be negative (the
J-curve). The short-run effect on output may also be negative as the
fall in the real wage as such reduces consumption. In the somewhat
longer run, however, trade elasticities are generally thought to be large
enough for a devaluation to improve both output and the trade ac-
count (given that the level of unemployment and slack in the economy
are sufficient for output to be responsive to demand).
3. Lower real wages are associated with an increase in profitability—
through higher profit margins as well as increased capacity utiliza-
tion—which should encourage investment in new capacity. A higher
investment ratio supports growth of output and productivity, thus
raising the sustainable level of employment and real wages. The
hoped-for effects on profitability and growth (rather than on the trade
balance) have indeed been the main motivation for most of the
devaluations undertaken by the Nordic countries.
4. Devaluation will induce inflation both through direct cost effects and
through demands for wage compensation (and possibly via an im-
proved labour market situation). The wage response will be large and
fast, particularly if wage indexation is widespread. The increase in the
price level will reduce the real value of domestic currency assets and
debts and may also lead to higher interest rates.
A widely held view, consistent with the sequence of effects outlined above,
is that devaluation will have real effects only in the short to medium term,
while the main long-run effect will be an increase in the general price level.
Furthermore, the real effects may be quite modest if investors and exporters
discount expected future wage increases and therefore do not respond to
the (temporary) improvement in competitiveness and profitability.1
The aforementioned view presumes that there is a unique equilibrium
for real variables; the only conceivable role for a change in the exchange
rate is then one of speeding up the process of adjustment from an initial
disequilibrium. While limited, this role may nevertheless be quite construct-
ive if exogenous disturbances are large and frequent and if nominal rigidities
are important. An even more essential role for central bank policies will
emerge if there are ‘hysteresis effects’—implying that present imbalances
have repercussions on developments in an indefinite future—or if policy
1
For an analysis which recognizes the effects on investment of expected future wage
reactions to a devaluation, see Risager (1988).
regimes have effects on strategically important parameters characterizing

private sector behaviour. We now turn to these issues.
9.3.2 Effects of exchange rate policies

Most economists feel strongly that the monetary system needs a nominal
anchor in the form of a controlled money supply (alternatively a target for
nominal income) or a fixed exchange rate. The motivation for this concern
is that fully accommodating policies might render the price level indeterm-
inate. For a small economy, in particular, it is generally thought that the
fixed exchange rate is the most natural mechanism for establishing an
anchor or a benchmark for inflation expectations and wage formation.
Correspondingly, the argument against using discretionary exchange rate
changes as a means of facilitating the adjustment process is basically that
such changes will remove the ‘fixed point’ and the associated ‘discipline of
the market’, thus having detrimental consequences for wage behaviour and
price stability.
In this view it is taken for granted that devaluations or an accommodating
exchange rate policy will not generate any additional employment in the
long run. Recurrent devaluations only cause permanently high inflation as
trade unions demand compensation for expected and realized inflation. Non-
accommodation in the form of a permanently and rigidly fixed exchange
rate is therefore often strongly recommended even though it is admitted
that credibility of the fixed exchange rate regime can perhaps be achieved
only by accepting a temporary phase of above-normal unemployment.2
One may (and should) ask why the Nordic countries have in fact chosen
to pursue devaluation policies if these are indeed futile or only harmful in
the long run. The simplest explanation is to draw attention to the import-
ance of the short run and to the magnitude and persistence of the adjust-
ment problems that may be encountered. The conflict between internal and
external balance may result in strongly deflationary pressure and high
unemployment. The alternative to a devaluation—restrictive demand manage-
ment policies that dampen wage growth by increasing unemployment—
works slowly and its results are politically painful.
Furthermore, contractionary demand management policies risk leading
to a lasting increase in unemployment because of asymmetrical reactions
which imply that higher actual unemployment also raises ‘equilibrium’
unemployment (hysteresis). Laid-off workers may not be re-employed even
2
This claim has been forcefully stated in a sequence of reports by the Swedish group of
economists called ‘Konjukturradet’. Results of formal analyses dealing with the consequences
of exchange rate policies in game-theoretic terms are actually quite sensitive to model specifica-
tion; cf. the articles in the Scandinavian Journal of Economics (1985). For instance, wage claims
may in fact be moderated if unions acknowledge that large nominal wage increases are useless
since they only trigger offsetting devaluations anyway.
286 SIXTEN KORKMAN
though output revives, partly because the perceived quality of labour may
be negatively affected by the fact of being unemployed. Deflation causes
irreversible scrapping of capacity, and with low utilization rates capital for-
mation becomes too small and unsupportive of future growth revival. The
structure of investment tends to be ‘defensive’, i.e. biased towards cost-
saving rather than output-increasing equipment etc. In sum, unemployment
is more easily increased than decreased; cf. Johansen (1982). Fear of a
deflationary spiral with permanent stagnation as the outcome is certainly
one of the reasons why Nordic governments (except Denmark) have so far
hesitated to endorse the ‘restoration of credibility just takes some temporary
suffering’ doctrine.
Another explanation for accommodating policies might be that these
policies are simply an illustration of the famous time-inconsistency problem.
The interests of society might indeed be better served by sticking to the
fixed exchange rate, but in the short run the government will find it optimal
to devalue (given an adjustment problem as well as the time lags of the
inflation effects of a devaluation and the time and pain needed to make
non-accommodating policies credible). Understanding this, the private sec-
tor expects policies to remain accommodating and behaves accordingly,
thus magnifying the problem and making expectations of accommodation
self-fulfilling.3 Precommitment may be the way out of this dilemma but more
so in a theoretical than in a practical context.
A third possibility worth considering is that there is something missing
in the ‘devaluation does not matter in the long run’ characterization of
exchange rate policy given above. One particular line of reasoning, which
does not seem far-fetched in a Nordic context, goes as follows: a basic
precondition for growth of investment and output is that entrepreneurs have
confidence in future demand and profitability conditions. The required
confidence was bolstered in the first decades after the Second World War
by successful government demand management policies and by the strong
commitment to full employment. More recently, this confidence has
evaporated and the investment propensity has fallen because of the risks
and uncertainty associated with real shocks and the unresolved conflict
between full employment and price stability.
For obvious reasons, domestic demand management is of less importance
in small economies. This is so particularly for the manufacturing sector and
exports, which in the longer run are of paramount importance for overall
macro-economic developments through their effects on external equilib-
rium. Actual and expected profitability in small open economies is more
directly dependent on wages, foreign prices, and exchange rates (as noted
in section 9.1).
3
The problem is magnified if there is a two-party system, since the incoming government
will then often have a particularly strong temptation to start its term with a devaluation.
Accommodating exchange rate policies may in these circumstances en-

hance investment by reducing the uncertainty of profit expectations.4 In-
vestment is discouraged by the fact that the yield on long-term investment
projects can easily be wiped out by unforeseeable price or wage develop-
ments or changes in world market conditions. The essential role of accom-
modating exchange rate policies is to reduce these risks by smoothing out
profit developments, or by redistributing risks so that they are shared more
equally as between sectors and agents. Given nominal wage rigidity and
imperfect capital markets, accommodating exchange rate policies may help
diversify risks and may to some extent act as an insurance system for private
sector profitability and for manufacturing profitability, in particular.
The argument that accommodation enhances investment does not re-
quire that devaluations have permanent effects on competitiveness and the
real wage (only that they reduce profit uncertainty). However, there are
actually two ways for devaluations to safeguard the level of manufacturing
profitability even though the competitiveness effects are temporary. First,
even temporary reductions in the real wage may enhance profitability if they
are recurrent. Second, devaluation policies may ease the lot of debtors—
such as industrial corporations—by maintaining inflation at a high level.
Needless to say, high inflation reduces the real value of debt (and assets),
particularly if interest rates are set administratively and are not adjusted in
response to changes in the rate of inflation.
Another perspective on the argument that accommodation supports in-
vestment is provided by the literature initiated by Lancaster (1973) on the
dynamic inefficiency of capitalism. Lancaster views workers and capitalists
as being involved in a strategic game where workers decide on real wages
and distribution, while capitalists decide on investment (and thereby deter-
mine future employment and consumption opportunities). Lancaster’s
point is that there is dynamic externality arising from the fact that workers
are unable to affect the allocation of profits as between capitalists’ consump-
tion and investment, while the capitalists cannot ensure that they will receive
the desired share of future output as returns to present investments. The
consequence of the dynamic externality is that equilibrium will be inefficient
as real wages will be too high and investment activity too low.
Accommodating policies may reduce this dynamic externality by setting
limits on the real effects that may be achieved through aggressive wage
policies. Devaluation can be used to wipe out any real wage gains and to
restore competitiveness, at least in the short run. In order to avoid an
accelerating inflation-devaluation spiral it may, however, at times be neces-
sary to reduce the influence of trade unions by pursuing deflationary policies
in the form of a credit squeeze or tight fiscal policy.
4
Cf. Korkman (1985).
288 SIXTEN KORKMAN
It is not necessarily the case that trade unions oppose devaluations, how-
ever, even though these might seem to mitigate the effects of wage increases.
One reason is that trade unions may accept (if only implicitly) the commit-
ment to certain profitability standards. More importantly, devaluation may
be acceptable to all participants in a conflict where rivalry between com-
peting unions is an essential element. As Pohjola (1984) has demonstrated,
union rivalry may cause an ‘excessive’ real wage, slow growth, and high
unemployment, when each union sets its wages without due regard to their
overall impact. Devaluation is a way of reducing the average level of real
wages without interfering with the wage relativities that union rivalry is
about. Again, the role of exchange rate policy is to modify and to reduce
the consequences of a dynamic externality arising from strategic interaction.
(Obviously, this is a reformulation of the old argument that labour accepts
reduced real wages through an increase in the price level more easily than
through lower money wages.)
In view of the interactions discussed above, it should be clear that the
exchange rate may have a pivotal role to play in corporatist policies in a
small economy. Before dealing further with corporatist policies, however,
it may be noted that the scope for using the exchange rate as an instrument
of such policies is importantly affected by the working of the financial
markets. When nominal interest rates were set administratively at low levels,
devaluation policies would, as noted above, ease the lot of debtors through
inflation. More recently, interest rates and capital flows have been deregu-
lated and have become more flexible. Now devaluation and inflation ex-
pectations will feed into nominal interest rates. In addition, exchange rate
uncertainty may create a risk premium, raising real interest rates as well.
The constraints on exchange rate policy imposed by the financial markets
are severe because of the large size of potential ‘hot money movements’.
Conceivably, uncertainty and speculative capital flows may in the present
circumstances be kept within limits—and interest rates at reasonable
levels—only by establishing confidence in the fixed exchange rate. Freely
floating rates may be the only viable alternative. Barring this alternative, the
regime of fixed exchange rates may thus be imposed upon the economy by
the financial community irrespective of the role that the exchange rate could
perhaps otherwise play as part of the macro-economic policies pursued by
the government or by the corporatist system.
9.4 EXCHANGE RATE POLICY AND CORPORATISM
A key goal of corporatist policies is to regulate income distribution in a way

that is acceptable to all parties and at the same time conducive to macro-
economic developments. The exchange rate is unavoidably essential to such
policies, but its role will vary according to the circumstances. The variation
in national experience is considerable.
In Finland it is generally recognized that recurrent devaluations have been
part of a growth-orientated policy. Sizeable devaluations were undertaken
in 1949, 1957, 1967, 1977–8, and 1982. A common interpretation is that
there used to be an ‘implicit understanding’ between industry, the forest
industry in particular, and the central bank, that investment activity should
be strong, and that, as part of such a policy, the exchange rate would be
devalued if this was necessary to ensure some minimum return on invest-
ment. Arguably, this is one explanation for the high investment ratio which
has been characteristic of the Finnish economy. The development of the
manufacturing sector, in particular, remained favourable in the 1970s and
1980s (Fig. 9.1). It seems obvious that the devaluations in 1977–8 were
instrumental in restoring competitiveness (Fig. 9.2) to a level consistent
with continued industrial expansion.
The attitude of trade unions to devaluations has been mixed. On some
occasions, they have resisted devaluations and demanded compensation for
the inflation effects. To avoid being caught in an accelerating devaluation-
inflation spiral it has, at times, been found necessary to limit the influence
of trade unions by pursuing quite deflationary policies in the form of a
monetary policy-induced credit squeeze and/or tight fiscal policy. As a
FIG. 9.1 Industrial production in Finland, Austria, Denmark, Norway, and

Sweden (s.a. 1970 = 100)
290 SIXTEN KORKMAN
FIG. 9.2 Finland: international competitiveness
result, Finland has pursued an accommodating policy with recurrent

devaluations but with a stop-go profile associated with brief deflationary
periods preceding the devaluations (whence the name ‘devaluation cycle’
to characterize Finnish exchange rate policies in the post-war period).
At other times, devaluations seem not to have been opposed by the
unions; in some instances they have even actively advocated devaluation (if
only behind the scenes). Union rivalry is probably one explanation for this
(cf. above). Another explanation is that devaluations have been decided by
coalition governments with a strong representation of left-wing parties. Typ-
ically, the exchange rate has been devalued by predominantly left-wing
governments once wage policies have failed to maintain the conditions
necessary for external balance.5 An additional reason for the rather modest
resistance to devaluations might be that the inflation-induced reallocation
of wealth (reductions in the real value of financial debts and assets) is perceived
as being favourable by both corporations (large debtors) and households6
5
There is a growing strand of literature containing game-theoretic analysis of the strategic
interaction between a wage-setting monopoly union on one hand, and an exchange rate-setting
government cum central bank on the other. Clearly, however, this approach is not fully
appropriate if governments and trade unions are not really independent decision makers, but
are rather to be viewed as different parts of the same power structure (with the Socialist Party
being the common denominator).
6
Cf. Oksanen(1988).
(typically also debtors)—while losses have been incurred mainly by the
pension funds (which nobody seems to include on their assets side).
In Sweden, the fixed exchange rate is an essential part of the well-
established corporatist doctrine called the EFO (or the Scandinavian)
model. This model, which was worked out by a group of experts from the
trade unions and from the employers’ association, states that nominal wages
should be kept within the limits set by increases in productivity and world
market prices in combination with the requirement that the functional
distribution of income be unchanged. It is a cornerstone of this model that
exchange rates be fixed. Nevertheless, in the difficult period since 1973,
Sweden has had to abandon this policy, at least temporarily.
Expansionary fiscal policy ran into difficulties, and a series of devaluations
was undertaken in order to restore conditions conductive to industrial growth.
The commitment to full employment was stronger than the commitment
to the fixed exchange rate. Again, devaluations have been undertaken by
left-wing governments as well (or in particular). It would seem that the
devaluations have been successful7 in improving competitiveness (Fig. 9.3)
and in recent years also in stimulating industrial growth (Fig. 9.1).
FIG. 9.3 Sweden: international competitiveness

7
Sweden’s 1982 devaluation seems, by and large, to have been a success (even taking into
account the favourable external environment). Nevertheless, practically all Swedish economists
are united in their claim that no new devaluation is conceivable. A more favourable attitude to
a soft exchange rate policy was taken by the group of economists associated with the Brookings
Institution.
292 SIXTEN KORKMAN
In Austria, a hard-currency policy has been pursued in the form of a fixed

peg, mainly to the German Mark. As a consequence, the Schilling has
appreciated 30 per cent in effective terms since 1970 and relative unit labour
costs have risen by some 20 per cent (Fig. 9.4). Yet industrial output has
developed rather favourably (Fig. 9.1). It would seem that wage formation
in Austria is so effectively controlled by corporatist policy that changes in
the exchange rate have no useful role to play (as is supposedly the case in
the Scandinavian model too). On the other hand, the industrial perform-
ance of Austria has deteriorated in recent years, and the authorities are
increasingly facing a dilemma.8
FIG. 9.4 Austria: international competitiveness
Norway has mainly pursued a hard-currency policy but only because it

could afford to do so. Competitiveness has deteriorated markedly since the
early 1970s (Fig. 9.5). More recently, falling oil prices have caused large
imbalances and an apparent switch to a soft-currency policy. A major factor
inhibiting devaluation seems to be that it would simply be pointless since
at present there is excess demand as a result of lax fiscal policies. The
institutional environment of exchange rate policy is much the same as in
8
The hard-currency policy has been associated with some lack of industrial flexibility (and
considerable support to government enterprises). This is interesting, as a major criticism of
accommodating devaluations is that they are ‘defensive’ and allow outdated activities to be
maintained.
FIG. 9.5 Norway: international competitiveness
Sweden. Industrial production has developed favourably but only because

of energy-related activities.
Finally, Denmark also undertook large devaluations in the early 1980s
(Fig. 9.6), and these were followed by a revival of industrial output. More
recently, the Danish authorities have placed strong emphasis on their com-
mitment to the fixed exchange rate in the hope of achieving low interest
rates and modest inflation. So far, however, domestic inflation has remained
higher than in competitor countries. In addition, in the EMS context, the
fixed nominal exchange rate (vis-a-vis ECU) has implied a substantial
strengthening of the effective exchange rate and a devastating loss of com-
petitiveness. Imbalances remain large, the manufacturing sector performs
poorly, and no resolution to the Danish dilemma is in the offing, cf. Akerholrrr
(1988). The Danish experience supports the view that the road to a credible
commitment to the fixed exchange rate regime is long and fraught with
hardship.
9.5 THE INTERNATIONAL DIMENSION
Exchange rate changes that are unilaterally decided are often regarded as
internationally harmful beggar-thy-neighbour policies. After the war, the
Bretton Woods institutions were created in order to avoid, among other
294 SIXTEN KORKMAN
things, competitive depreciations. Exchange rates were supposed to be

basically fixed and adjusted only with due regard to international aspects.
Ex post it would seem that the Bretton Woods system served its purposes
well; the 1950s and 1960s were a period of unprecedented growth and
stability.9 In contrast, the fifteen years of experience of floating rates are
generally assessed to have been disappointing. Needless to say, floating rates
are not to be blamed for real disturbances like the oil shocks. Nevertheless,
floating rates have been part of the problem rather than of the solution
because of their association with large trade imbalances and financial tur-
bulence (related to uncertainty hampering trade and investment, waning
confidence in the financial system, and the debt crisis, etc.). The experience
of the last decade has demolished the myth that floating rates make inter-
national policy co-operation redundant (and that monetarist policies are a
road to price stability with full employment).
Greater fixity of exchange rates is currently in vogue again: intervention
in exchange markets takes place on a large scale, international co-ordination
of macro-economic policies is being called for, and target zones for the
major currencies are being discussed. One concrete expression of the move
9
At the time, however, the verdict of most economists was different. Fixity of exchange rates
was condemned and flexible rates were advocated by both monetarists and Keynesians. The
former considered free exchange rates as a natural part of an unregulated price mechanism; the
latter hoped to escape the constraints on macro-economic policies imposed by the requirements
of external equilibrium.
towards fixed rates is provided by the vitalization of the European Monetary

System. The EMS has secured relative stability of European exchange rates,
and inflation differentials between member states seem to have narrowed.
The EMS has consequently come to be regarded as an example of successful
monetary co-operation on a regional basis. An enlargement of its member-
ship seems likely, as does also the future development of EMS co-operation
towards a full monetary union with a European central bank as the final (if
distant) goal. There is no doubt that many small countries, in particular,
will seriously consider joining the EMS in the coming years. Denmark is
already a member, Austria might join both the EMS and the EEC, and
membership is being discussed in Sweden and Norway. Interest will in-
crease when the United Kingdom joins the EMS.
EMS membership is of course quite attractive to those who stress the
importance of credibility and commitment to the fixed exchange rate
regime. The advantage of the EMS from this point of view is that unilateral
exchange rate changes would not be allowed, and that the discipline im-
posed by the pegged rate would therefore be more severe. At best, EMS
may enhance both domestic price stability and international policy co-
operation.
On the other hand, EMS is no automatic route to credibility (cf. the
Danish experience where nominal and real rates of interest are persistently
high). Moreover, EMS membership largely eliminates the freedom to use
the exchange rate as part of adjustment policies implemented in a corpor-
atist context or otherwise. This consideration is of some importance in the
Nordic context because these countries have an economic structure (e.g.
the forest industry) and political priorities (e.g. die commitment to full
employment) which differ from the rest of Europe. Disturbances requiring
adjustment of real exchange rates are therefore likely to occur in the future
as well.
In retrospect, the currency-basket system operated by the Nordic coun-
tries seems to have worked reasonably well. Large and persistent misalign-
ments of exchange rates have been avoided. The devaluations in 1977–82
in Finland and Sweden contributed to developments in employment which
are internationally satisfactory. The manufacturing sector, in particular, has
been supported by the soft exchange rate. It may be that the policy of an
adjustable peg to a basket of currencies cannot be continued in the future.
Deregulation of financial markets and capital flows have changed the situa-
tion. Discretionary exchange rate changes are difficult to implement in a
world with large and mobile capital flows. A choice between EMS mem-
bership and free floating may be the options that remain.
In this respect, the EMS is a very noteworthy alternative. For an enlarged
EMS system to work satisfactorily, however, it will probably have to develop
in the direction of more symmetry (as regards the adjustment burden
296 SIXTEN KORKMAN
between surplus and deficit countries) and flexibility (as regards the size
and/or frequency of realignments). Alternatively, the Nordic countries
should retain the right to introduce temporary capital controls in times of
financial unrest, or they should be allowed larger bands around the central
rates (maybe even larger than the 6 per cent allowed for Italy).
9.6 CONCLUDING COMMENTS
Competitiveness and the real wage are bound to be important for the
macro-economic performance of small economies. This is widely recog-
nized by labour market organizations as well as by governmental bodies. It
is therefore generally thought to be imperative that institutional arrange-
ments be conducive to the necessary synchronization of wage, price, and
exchange rate developments.
One of the main advantages of the fixed exchange rate is that it clarifies
the allocation of roles and responsibilities in the policy-making context. In
particular, it puts severe constraints on central bank policies and at the same
time yields a simple norm for appropriate wage formation in normal cir-
cumstances (the Scandinavian model). With fixed exchange rates and exo-
genous world market prices (and productivity developments), the ‘room’
for domestic wage increases may be reasonably well assessed. From this
perspective, the fixed exchange rate may be viewed as an essential element
of the system of corporatist policies aimed at regulating income distribution.
In the past two decades, small economies have been faced with large
shocks and severe imbalances. Domestic wage flexibility cannot ensure full
employment in such circumstances. The imbalances may become cumulat-
ive unless adjustment is effective. Given nominal wage rigidity, effective
adjustment policies are likely to include a change in the exchange rate as
part of the policies.
In the five countries briefly considered above, exchange rate policy has
indeed been closely associated with corporatist policies. For instance, in
Austria the hard-currency policy is a cornerstone of the corporatist system
of real wage regulation. The same used to be the case in the Scandinavian
countries. More recently, Sweden has actively used devaluation as an in-
strument of adjustment policy. Again, this has taken place within the
framework of policies decided by a socialist government and implemented
by the monetary authorities and trade unions. In Finland, arguably, accom-
modating exchange rate policy was earlier based on an understanding
mainly between the forest industry and the central bank (1940–70), while
it has subsequently been based more on an understanding between the
labour market organizations and the government. There are many reasons,
some of which were dealt with above, why these policies may have had
important effects not only on nominal variables but on real variables like
growth and employment as well.10
Conceivably, the manipulation of exchange rates in order to support
industrial growth and employment is a policy strategy which is not viable
in the new financial environment characterized by large and unregulated
capital flows. If so, the fixed exchange rate system may be sustainable in
Nordic countries only if corporatist policies can be strengthened and made
more flexible, possibly in the context of an enlarged EMS. Otherwise,
floating exchange rates may be the unavoidable even if undesirable out-
come. Also, the future role and power of the corporatist institution is
conditional on its ability to cope with the new financial environment.
REFERENCES
AKERHOLM, J. (1988), ‘External adjustment in small open economies: some recent

experience’, Bank of Finland Discussion Paper 6/88, Helsinki.
EICHENGREEN, B., and SACHS, J. (1984), ‘Exchange rates and economic recovery in
the 1930Y, National Bureau of Economic Research, Working Paper 1498.
JOHANSEN, L. (1982), ‘Arbeidslosheten: Letter opp enn ned’, Sosialokonomen, 36:
29–40.
KORKMAN, S. (1985), ‘Devaluation policy and employment’, in Claes-Henric Siven
(ed.), Unemployment in Europe, Analysis and Policy Issues, Stockholm: Timbro.
LANCASTER, K. (1973), ‘The dynamic inefficiency of capitalism’, Journal of Political
Economy, 81: 1092–109.
OKSANEN, H. (1988), ‘Inflaatio-oikaisut kansantalouden sektoreiden tulotileihin’, in
S. Honkapohja and A. Suvanto (eds.), Raha, inflaatio ja talouspolitiikka, Helsinki:
Valtion painatuskeskus.
POHJOLA, M. (1984), ‘Union rivalry and economic growth: a differential game
RiSAGER, O. (1988), ‘Devaluation, profitability and investment’, Scandinavian Journal
Scandinavian Journal of Economics (1985), 87/2, ‘Proceedings of a conference of trade
unions, wage formation and macroeconomic stability’.
10
There is no obvious correlation between competitiveness and industrial performance. This
is not surprising in view of the two-way causality involved: strong price competitiveness most
likely supports industrial production, but rapid growth will also lead to higher wages and lower
price competitiveness. For the 1930s, it has been argued that the crisis was much less severe in
those countries that undertook large devaluations than in other countries: see Eichengreen and
Sachs (1984).
10
Corporatism and Economic Performance in
Sweden^ Norway, and Finland
Jukka Pekkarinen
10.1 INTRODUCTION
Three Nordic economies, Sweden, Norway, and Finland, share common

characteristics which make them an interesting subject for study from the
point of view of corporatism and economic performance.
First, as far as the institutions bearing on industrial relations and economic
policy are concerned, all three countries rank relatively highly on the various
scales that are used to indicate the degree of corporatism—union density,
centralization of wage bargaining, involvement of interest organizations in
economic and social policy design, and, except Finland, number of strikes, etc.
Second, the recent economic performance of these countries, in particular
with respect to employment, has been impressive by European standards.
They have achieved high employment levels and their unemployment rates
have remained low.1 This may to some extent reflect the readiness of the
Nordic countries to accept higher inflation rates and larger current account
deficits than elsewhere, but empirical studies frequently suggest that they
have favourable trade-offs between inflation and unemployment and display
relatively flexible real wages.2
Comments from Peter Katzenstein, Paulette Kurzer, Steven Marglin, Wolfgang Merkel,
Matti Pohjola, Bob Rowthorn, Juliet Schor, and other participants in the project are gratefully
acknowledged. I also thank Malcolm Waters for correcting my English. The research was
supported by financial assistance provided by the Finnish Ministry for Foreign Affairs to the
project being carried out at the Labour Institute for Economic Research. The chapter was partly
written during the author’s stay at Harvard University in the academic year 1988–9. This visit
was made possible by grants from the Academy of Finland, the Yrjo Jahnsson Foundation, and
the Finnish Cultural Foundation.
A qualification is needed for Finland, which has represented a kind of intermediate case
between low unemployment and high unemployment countries; Finland is nevertheless one of
the very few OECD countries which succeeded in consistently reducing unemployment during
the first half of the 1980s and has now reached an unemployment rate lower than that of Norway
or Austria. The labour market performance of the corporatist economies is examined more
closely by Bob Rowthorn in this volume.
2
See e.g. Grubb et al. (1983); Kahn (1984); McCallum (1986). Except for Norway, these
results are broadly confirmed by those obtained in the project on ‘Wage formation in the Nordic
countries’, in which Sweden scores the highest elasticity of real wages with respect to unemploy-
ment. See Calmfors (1990). For a discussion of the issue of real wage flexibility, see also Matti
Pohjola in this volume.
CORPORATISM IN THE NORDIC COUNTRIES 299
Third, as shown by Bob Rowthorn in this volume, the economic and social
record of these countries is enhanced by the relatively low degree of wage
dispersion across industries in all Nordic countries, including Denmark.
This is further strengthened by the heavy reliance on a progressive income
tax as a source of public sector revenue and by an extensive welfare sector,
as discussed by Katri Kosonen in this volume.
The Nordic countries appear to be good examples of social corporatism,
in the sense of the terms used in this volume: they show a combination of
centralized bargaining, heavy government involvement, and a high degree of
solidarity. Taking into account their good employment performance, the
question naturally arises whether the latter has been caused by the former.
Before jumping to such a conclusion two counter-arguments addressed in
this chapter must be considered:
1. The nature of corporatism and its linkages with economic performance
vary strongly from one country to another even among the subgroup
of social corporatism.
2. It has frequently been argued that the corporatist ‘star performers’,
using the term due to Rowthorn and Glyn (1990), have not done very
well from a long-run point of view as far as growth and structural
change are concerned. If true, this claim casts doubt on the sus-
tainability of their performance.
Some introductory remarks on these two issues are in order:
1. The rediscovery of corporatism by economists, many years after it had
been extensively dealt with by other social scientists already preparing to
leave the subject, exemplifies the general revival of interest in institutions in
economics. The most narrow characterization of corporatism relates to certain
structural characteristics of the wage-bargaining institutions. High degrees
of unionization and vertically centralized organization structures among
employees and employers are usually observed. They give rise to centralized
wage bargaining. A broader structural definition encompasses a state ready
to negotiate with the interest organizations and interventionist in its economic
policies. Corporatism is thus conceived as a tripartite bargaining structure
between labour, capital, and the state. The power balance of this triangle
varies and this gives rise to different variants of corporatism (cf. Katzenstein
1985). The variety is further increased by the fact that institutions are
historical and carry the legacies of specific national traditions concerning, for
example, the agenda of economic policies (cf. Pekkarinen 1989).
The historical contingency of corporatism appears especially important
from the point of view of the definition of corporatism adopted in this project.
The distinction between conflict- and consensus-based varieties of corpor-
atism refers to differences in the goals the interest organizations, the trade
union movement in particular, strive for. This difference in goal setting is at
300 JUKKA PEKKARINEN
least partially independent of the degree of centralization. Moreover, goals

may change over time. This suggests a dynamic dimension to corporatism.
2. The relevance of the dynamic dimension is further enhanced by the fact
that, apart from the rediscovery of institutions, economists’ concern with
corporatism is related to the debate on flexibility evoked by the largely
structural interpretation of the recent economic crises. In the late 1970s and
early 1980s, the call by policy makers, international economic organizations,
and many economists for greater flexibility of the economic system was
generally interpreted as implying a move toward decentralization and the
freer play of market forces. More recently, acquaintance with the corpora tist
experiments has introduced more nuances into the flexibility issue by sug-
gesting that centralized economic systems in which labour is well organized
and in which government intervention is widely used can show high flex-
ibility in adjusting to changing economic circumstances.
There are several channels through which corporatism can contribute to
flexibility. As emphasized by Bob Rowthorn in his essay in this volume,
centralized wage bargaining is a necessary condition for direct control of the
general wage level and consequently, at least in a small open economy, the
distribution of income between wages and profits as a policy target. This may
have enhanced the capability of the corporatist economies to deal with the
supply shocks of the 1970s. As far as the micro aspects of flexibility are
concerned, it is suggested in this book that the distinction between the
inclusive type of corporatism based on institutionalized conflict and the
typically more exclusive type of consensus-based corporatism is relevant.
The former type of corporatism, which more or less characterizes the Nordic
countries, tends to freeze the wage structure and thus weaken the role of
relative wages in the reallocation process.
On the other hand, active manpower policies and union involvement in
the management of the reallocation of labour may enhance micro-flexibility.
Furthermore, in contrast to the standard view, the high priority placed on
full employment and low wage differentials characteristic of the social
corporatist economies may facilitate micro-flexibility: when jobs offering
largely the same pay are available elsewhere in the economy, workers’
resistance to restructuring may be diminished. It may also be the case that
the principle of democracy or non-exclusion lengthens the horizon of the
interest organizations and enhances their readiness to accept far-reaching
restructuring measures considered necessary for preserving the economic
viability of the system while at the same time contributing to a more
solidaristic sharing of the burden of adjustment. Bearing on this issue, this
chapter discusses flexibility from the point of view of the stability or integrity
of corporatism over time. If corporatism results in a poor long-run perform-
ance with respect to growth and structural change, it endangers the pre-
conditions for its own existence in the course of time. Bargaining has to take
place under deteriorating economic conditions. The co-operative mood of

the game between the interest organizations weakens as also may their
internal solidarity.
The Nordic countries are usually regarded as prime examples of social
corporatism. Without challenging this interpretation, this chapter stresses
the point that there are important differences between them. This diversity
implies certain qualifications concerning the relation between corporatism
and economic performance. More importantly, the experience of each
Nordic economy can be used to illuminate certain specific aspects of cor-
poratism. One distinction can be drawn between Sweden and Norway on
the one hand and Finland on the other. Whereas the former countries repre-
sent mature corporatism, the relevant institutions having become established
long before the economic turmoil of the 1970s, Finland, like Australia
discussed by Robin Archer in this volume, is an example of strengthening
corporatism, where institutional realignment was combined with the chal-
lenge posed by the economic crises of the 1970s.
Against this background, the following sections take a closer look at
economic developments in the three countries since the mid-1970s. Sweden
provides an example of steady corporatism. Viewed from the perspective of
the late 1980s and overlooking the current problems of overheating, the
Swedish economy seemed to have succeeded in combining continuous full
employment and restructuring. The role played by the expanding oil sector
cannot be overlooked when accounting for the success of the Norwegian
economy in the late 1970s and early 1980s. But the economic situation
deteriorated sharply as a result of the fall in oil prices in the mid-1980s. The
question to be faced in the Norwegian case is whether postponement of the
restructuring of the non-oil economy weakened its long-term growth pro-
spects. If so, the Norwegian case may illuminate the thread of instability or
fragility of corporatism.
Finland figures as a country which, after a sharp decline in the mid-1970s,
rapidly improved her relative economic performance. As this coincided with
the rise of Finnish corporatism, the question naturally arises whether the
Finnish case exemplifies the positive economic consequences oi strengthening
corporatism. This hypothesis has to be set against the alternative view
according to which certain favourable, but not necessarily irreversible,
changes in the international economic environment were decisive for the
improvement of economic performance in Finland.
In what follows section 10.2 describes in general terms the structural and
institutional background of each country. Sections 10.3–10.5 discuss how
economic policies, the behaviour of the private sector, and changes in the
economic environment have shaped economic performance in each country
since the early 1970s and the role of corporatism in this. Conclusions are
drawn in section 10.6.
10.2 SOCIAL CORPORATISM IN THE NORDIC COUNTRIES:

BACKGROUND
As far as the economic structure is concerned, all three countries meet

Katzenstein’s (1985) criterion of external vulnerability as a precondition for
social corporatism. Their industrialization was export-led. While Sweden
developed a relatively diversified export structure in the inter-war period, the
exports of Norway and Finland have traditionally relied on a few bulk
products the markets for which are very volatile. Reflecting this difference,
Swedish exports taken as a whole tended to be more stable than those in the
other two countries. One would thus expect the external balance in Finland
and Norway to have been more volatile than in Sweden. In Finland this
indeed was the case whereas in Norway, curiously enough, swings in exports
and imports were so closely synchronized that serious current account
disturbances were rare (for possible explanations, see Pekkarinen 1989).
These differences notwithstanding, up until the first oil crisis the broad
international economic environment of these three small open economies
was quite similar. While the industrialized countries in general slid into a
deep recession, which strongly influenced the overall policy assessment in
the Nordic countries as well, Sweden, Norway, and Finland, as a con-
sequence of the central role of resource-based goods in their exports and the
expansionary economic policies pursued, experienced rapid growth in 1974.
But in 1975 exports contracted strongly. While domestic demand still
remained buoyant and inflation accelerated, external balances deteriorated
sharply. These imbalances set the economic policy agenda for many years to
come, each of the three countries reacting to them in its own way.
Among the institutional preconditions for corporatism, one may first note
the fairly high degree of concentration among firms, characteristic of the
Nordic export industries in particular. Together with a centralized trade
union movement and also a fairly centralized state structure, this has
facilitated corporatist approaches.3 In contrast to the formal vertical hier-
archy among each of the interest organizations, the horizontal interaction
between state, capital, and labour has been quite flexible. This combination
of horizontal flexibility and vertical hierarchy has been considered typical of
corporatist arrangements.4 One should, however, remember that in social
corporatism of the Nordic type the local level has played an active role in the
trade union movement.
The fairly high degree of centralization of the Nordic state structures
facilitates government involvement in the economy.5 This is also supported
by the fact that the Nordic countries lack a strong laissez-faire tradition; a
3 4
See Czada (1987). This has been emphasized e.g. by Lehmbruch (1982).
5
For the relevance of this point, see Weir and Skocpol (1985).
strong economic role for the state has in principle been accepted across a
wide political spectrum. The Nordic countries also have a tradition of extens-
ive committee work in the preparation of major policy reforms. As these
committees have relied on the representation of interest organizations, they
have performed as a catalyst for corporatism.
Broad similarities in the economic role of the state do not preclude some
crucial differences in economic policies. These are in many respects greatest
between Finland and Sweden, while Norway, although closer to Sweden,
forms an intermediate case.6 Whereas in Sweden and Norway the Keynesian
ideas of demand management were already widely accepted before the
Second World War, economic policy in Finland resisted these ideas. Fiscal
policy, in particular, followed a clearly pro-cyclical pattern in Finland until
quite recently. The main attention in economic policy was paid to the relative
costs of the export sector. Moreover, in Finland the central bank retained an
unusually high degree of autonomy whereas in Sweden and Norway the
treasury gained effective control of it.
Norway shares with Finland the characteristic that direct state interven-
tion in production has been extensive. In both countries, the state has
actively fostered industrialization by setting up state-owned enterprises and
by granting subsidies. In Sweden such direct state intervention has been
exceptional, notwithstanding the high level of public sector saving—a com-
mon feature in all three countries emphasized by Katri Kosonen in her essay
in this volume. Until the late 1970s, state ownership and subsidies to
non-farming industries did not play a big role. Moreover, in Sweden the state
has traditionally not relied on dirigiste incomes policies. Wage control has
been considered the responsibility of the labour market organizations.
Although Agrarians and Liberals have often allied with Social Democrats
to form a majority in major reforms in Sweden,7 the general pattern of
economic policies has frequently been challenged by the non-socialist par-
ties. This gave rise to deep conflicts such as the controversies over planning
in the late 1940s, pension funds in the 1960s, and wage earner funds in the
1970s. The strong presence of conflict in the Swedish case has raised the
question in the literature as to whether it should be considered a corporatist
case at all in the conventional ‘Continental’ use of the term. While referring
to the interpretation of corporatism as a system where there is consensus at
the top level of organizational hierarchies and the trade union movement
accepts the responsibility for suppressing the demands of labour from below,
Walter Korpi and Gösta Esping-Andersen, for instance, regard Sweden as a
case of democratic class struggle where labour is strong and the corporatist
6
I have argued for this point elsewhere, referring to, as national economic policy models,
the thinking patterns concerning economic policy and cutting across the political spectrum.
They seem to show remarkable persistence over time. See Pekkarinen (1989).
7
See Marklund (1988).
features of which represent a kind of tactical compromise, each stage of

which will be challenged by the increasing power of the left.8 This view
closely resembles Göran Therborn’s notion in this volume that Sweden
represents a case where relatively peaceful industrial relations are maintained
by institutionalized conflict. Indeed, in terms of the distinction drawn in
many essays in this volume between consensus- and conflict-based cor-
poratism, Sweden is a prime example of the latter.
In Norway and in Finland consensus concerning the basic thrust of policy
has traditionally been more prevalent. In Norway it was established during
the reconstruction period after the war under a Social Democratic hegemony
that remained unchallenged until the 1960s. In Finland changes in cabinet
coalitions were frequent and majority cabinets were an exception until the
late 1960s. Since then, quite stable coalitions between socialist and bour-
geois parties, unknown in other Nordic countries, have been the main rule
in Finland. The strict majority rules of the Finnish constitution have com-
pelled governments to negotiate with the opposition on many major issues
requiring new legislation. Combined with the emphasis on consensus and
heavy government involvement in bargaining between capital and labour
that has been prevalent in Finland, this raises the question, to be addressed
below, as to whether Finland fits the description of conflict-based corporatism.
The present systems of industrial relations in all three countries are charac-
terized, first, by high degrees of organization among employers and
employees.9 Labour is organized in industry-wide unions of manual workers
and lower-paid salaried employees on the one hand and in mostly craft-based
professional unions on the other. The biggest of the central organizations are
the LOs in Sweden and Norway and the SAK in Finland which represent
the industry-based unions in manufacturing, construction, and the low-paid
workers’ and salaried employees’ unions in the public and private service
industries. In Sweden, the LO alone covers about half of the labour force,
somewhat more than the LO in Norway or the SAK in Finland. The
coverage of the LO in Norway is closely confined to manufacturing, con-
struction, and communications. The craft-based professionals’ unions are
organized in two or three central organizations. Norway also has some
independent professional unions.
The overall unionization rate exceeds 85 per cent in Sweden and Finland,
with little difference between manufacturing and services or between
workers and salaried employees in either country. In Sweden the increase in
8
Korpi (1983); Esping-Andersen (1985).
9
There is an extensive comparative literature on the system of industrial relations in Norway
and Sweden, whereas that on the Finnish system is more sparse. See e.g. Flanagan et al. (1983);
Braun (1986); Elvander (19886). Most recently, wage formation in the Nordic countries,
including Denmark, has been studied extensively in the project on wage formation in the Nordic:
countries mentioned in n. 2 above. See Calmfors (1990).
the unionization rate has been fairly steady, whereas in Finland unionization
remained quite low, around 40 per cent, until the late 1960s. During the
next ten years, it increased rapidly and has been quite stable more recently.
In Norway the degree of unionization has not increased much since the late
1960s, and it is now less than 70 per cent, being significantly lower among
salaried employees and in the service sector than among workers in manufac-
turing or construction.
In Scandinavia, the LOs have been politically closely allied with the Social
Democratic parties, which in turn have normally dominated governments.
In Finland this tie has been looser, and there has been a strong presence of
Communists in the blue-collar trade union movement.
Private employers were organized roughly at the same time as employees
around the turn of the century. Nowadays, a strong presence of the central
organizations of employers in wage negotiations is a common characteristic
of all Nordic countries. Public and private sector employers are organized in
separate central organizations, however, as are, in the case of Finland, private
employers in the secondary and tertiary sectors. Political ties between
confederations of employers and bourgeois parties, although present, have
not been as systematic as between trade unions and parties on the left.
Compared internationally, wage bargaining in the Nordic countries is
characterized by a high degree of centralization; there has not, however,
been full co-ordination between settlements for blue-collar and white-collar
unions, or private sector and public sector unions, for instance. Standard
features of centralized bargaining—multi-year wage contracts, use of index
clauses and earnings guarantees, etc.—have, to a somewhat varying degree,
been present in all three countries. These common features aside, there are
some differences in the bargaining procedures among the Nordic countries.
First, centralization is higher and of older origin in the Swedish and Nor-
wegian LOs than in the Finnish SAK or among the professional unions.
While Sweden had an unbroken sequence of centralized settlements in the
period 1956–82, recourse to bargaining at industry level has been more
frequent in Finland and also in Norway. Moreover, whereas for example in
Sweden the role of industry-wide wage negotiations is normally downplayed
by the binding centralized agreement and the local level is important in
applying the central agreement, in Finland the industrial unions are more
autonomous. They specifically decide whether or not to accept the central
agreement negotiated by the central organization. On the other hand,
although conditional upon the acceptance of the industrial unions, each
central organization has normally been a partner in accepting the central
agreement in Finland whereas in Sweden for example the central agreement
is negotiated by the LO for the private sector only. Other central organiza-
tions do not negotiate for their member unions, which have formed separate
negotiation cartels for the private sector and the public sector. Finally, in
Finland the government has been an active partner in the centralized wage
negotiations that have traditionally been quite explicitly tied up, as com-
prehensive ‘package solutions’, with issues of taxation and social policy. In
Norway, compulsory arbitration enacted by parliament has been fairly
frequent and the government has also intervened sometimes in the bargain-
ing between the labour market organizations by means of tax concessions.
Most recently, wage restraint has been imposed by legislation.
A common feature of all Nordic countries, which distinguishes them from
many European countries, is the rather strong presence of trade unions at
the plant level. Shop stewards and other trade union representatives play an
important role at the work-place where unions face no competing institu-
tions of work-place representation such as codetermination bodies. Local
bargaining at firm (or plant) level is also an important element of wage
bargaining. It is reflected in ‘wage drift’, i.e. the excess over centrally
negotiated wage increases, which accounts for not less than roughly half of
total wage increases in the Nordic countries.10
Finally, the Nordic countries had to face decentralization tendencies in
wage bargaining in the course of the 1980s. These developments will be
discussed more closely in the context of the country analysis and in the
concluding section below.
10.3 SWEDEN: SOLID CORPORATISM?
In the early 1980s, the performance of the Swedish economy deteriorated

sharply (see Table 10.1). Output growth stagnated; in manufacturing the
volume of output was 4 per cent and the volume of investment no less than
40 per cent lower in 1982 than in 1976. The rate of productivity growth
decreased. Swedish exports suffered heavy losses in the market shares. A
whole bunch of crisis industries emerged in the previously flourishing Swedish
manufacturing sector. Subsidies, which by the early 1980s amounted to 2
per cent of GDP, were paid to keep them alive. The macro-economic
symptoms of external and internal imbalance were alarming. The current
account, in a comfortable surplus in 1973, swung into a deficit amounting
to nearly 4 per cent of GDP by 1982, and the fiscal balance of the
consolidated public sector (including pension funds), which had shown a
surplus of 4.6 per cent of GDP as late as 1976, had turned into a deficit of
10
There is an ongoing debate on the determinants of wage drift in the Nordic countries:
whether it is accounted for in the central negotiations, whether, in contrast, it is the ultimate
determinant of wage increases which offsets variations in centrally negotiated wages, or whether
the two components of wage increases are independent of one another. For contributions to
this debate, see the papers by Jackman, Holden, Holmlund and Skedinger, Flanagan, and
Pissarides and Moghadam in Calmfors (1990).
TABLE 10.1 Economic indicators on Sweden, 1973–1988
Year Change Change Change Unemploy- Employ- Labour Change Manufacturing Manufacturing Effect- Change RULCC Current
in real in real in real ment rate ment rate3 market in con- Real Nominal Negoti- Wage ive in real account
GDP exports domestic meas- summer hourly hourly ated wage drift exchange output per (%ofGDP)
(%) (%) demand ures prices earnings earnings rate person employed
(%) (%)
1973 4.0 13.7 2.3 2.5 73.6 2.8 6.7 1.4 8.2 4.1 4.0 143.3 5.1 120.3 2.8
1974 3.2 5.3 4.6 2.0 75.4 2.5 9.9 i.o 11.0 5.0 6.8 143.0 0.3 116.4 –1.0
1975 2.6 –9.3 4.1 1.6 77.2 2.3 9.8 4.6 14.8 10.5 7.5 148.2 –1.3 121.7 –0.5
1976 1.1 4.3 2.6 1.6 77.6 2.6 10.3 6.9 17.8 7.9 5.4 151.4 3.5 136.8 –2.1
1977 –1.6 1.5 –3.1 1.8 77.8 3.3 11.4 –4.3 6.7 3.7 3.5 145.4 –2.2 134.4 –2.6
1978 1.8 7.8 –2.1 2.2 77.8 3.6 10.0 –1.1 8.7 4.8 3.2 131.1 0.8 123.4 0.0
1979 3.8 6.1 5.5 2.1 78.8 3.6 7.2 0.5 7.8 4.4 3.8 131.0 6.1 116.0 –2.2
1980 1.7 –0.5 1.9 2.0 79.4 2.9 13.7 –4.3 8.8 6.1 3.2 130.2 0.4 115.2 –3.6
1981 –0.3 1.1 –2.8 2.5 79.0 2.8 12.1 –1.4 10.6 5.9 4.2 129.2 0.9 117.0 –2.5
1982 0.8 4.4 0.8 3.1 78.6 3.6 8.6 0.1 8.7 4.1 3.5 116.3 3.5 103.5 –3.6
1983 2.4 10.7 –0.8 3.5 78.5 4.2 8.9 –1.7 7.1 3.8 2.9 103.0 5.7 91.6 –1.0
1984 4.0 6.7 3.2 3.1 77.8 5.0 8.0 1.4 9.5 6.2 4.1 104.8 5.5 96.2 0.4
1985 2.1 2.2 4.0 2.9 78.7 4.4 7.4 0.2 7.5 3.8 3.7 103.7 1.5 99.2 –1.3
1986 1.1 3.3 1.7 2.7 79.1 4.0 4.3 3.0 7.4 3.9 3.5 101.5 –1.7 100.9 0.7
1987 2.4 2.5 3.7 1.9 80.1 3.8 4.2 2.2 6.5 — — 100.0 5.7 100.0 –0.3
1988 2.1 3.0 2.9 1.6 — 3.7 5.8 — 3.4 5.0 99.8 — 104.8 —
a
Employed as a percentage of population aged 15–64.
b
Persons affected by special labour market measures as a percentage of total labour force.
c
Relative unit labour costs.
Source: OECD (1973–88).
more than 7 per cent of GDP by 1982. After a short pause before the second
oil crisis, inflation reaccelerated and the rate of increase in consumer prices
reached 13 per cent in 1982. Only labour market indicators brought any
satisfaction; but they, too, took a turn for the worse. The rate of open
unemployment approached 3 per cent in 1982—despite the fact that about
3 per cent of GDP was used for special labour market measures. Total
employment, which, thanks to the rapid growth of the public services, had
still been rising in the late 1970s, fell.
However, the Swedish economy recovered rapidly after 1982. In 1984,
manufacturing output was already 12 per cent above its level in 1982. The
sector underwent a successful restructuring. Manufacturing investment
picked up after declining from 1976 to 1982. The rate of growth of labour
productivity increased. With continued growth of output in the service
sector, the rate of growth of real GDP soon reached that of the average of
European OECD countries. The rise in employment resumed after 1982.
This made it possible gradually to dismantle the manpower programmes and
subsidies while the rate of unemployment fell continually to its present level
of below 2 per cent. Among the macro-economic indicators, a clear improve-
ment has taken place in the fiscal balance of the public sector; the financial
deficit of the consolidated public sector was eliminated by 1987 and the
public sector resumed its traditional role as a financial surplus sector. The
foreign trade balance has also swung back into surplus, although the current
account is in deficit. Inflation started to abate after 1983 but it has persist-
ently run higher than on average in the OECD countries and is currently on
the increase. Indeed the inflationary pressures, intensified with the recent
overheating, combined with persistent slack in the productivity growth and
ongoing erosion of the institutional and political basis of the model, have
most recently led the Swedish model into a new crisis, which will be
considered towards the end of this section.
These changes in Sweden’s economic performance are reflected in do-
mestic and foreign assessments. Whereas in the early 1980s it was widely
cited as a warning example for the consequences of not combating imbal-
ances in time, the Swedish economy was later on as widely praised for its
success in combining restructuring and continuous full employment. Most
recently, however, the current problems of overheating have evoked increas-
ing scepticism.
As a matter of fact, the late 1970s represented the first large-scale attempt
in Sweden to overcome an international recession by ‘vulgar’ Keynesian
means. Was this flirtation a failure and was there a fundamental recon-
sideration of policies in 1982 which saved the Swedish economy from
ruination? It seems that a positive answer to this question has widely been
taken for granted. This has certainly been the case among those Nordic
economists who have criticized Swedish economic policies in the late 1970s
for accommodating public employment measures which have allowed trade
unions to pursue excessive real wage claims.11 An important exception is
formed by the corporatism literature, where the continuous maintenance of
full employment is linked to the corporatist features of Swedish society.
Andrew Glyn and Bob Rowthorn, for instance, interpret the period from the
first oil crisis to the mid-1980s rather as a continuous whole.12 The first
phase, extending to the early 1980s, consisted of job protection and the
expansion of public sector employment, which was partly financed by an
increase in the budget deficit and partly by increased taxes. In the second
phase the main emphasis was on the restructuring of the open sector, which
was helped by the international upswing.
I shall also argue for this continuity hypothesis but with some qualifica-
tions. It is doubtful whether there ever was any conscious two-stage strategy.
Rather, Swedish policy makers tried, often with much hesitation and
controversy, to cope with the situation as it evolved. While changes in
circumstances led to changes in policies, the continuity was represented by
the priorities, or the pressures the policy makers had to adjust to.
It is useful to distinguish between different types of continuity:
1. continuity of policy goals and priorities;
2. continuity of the range of feasible policy options as determined by
external and internal constraints;
3. continuity in targeting different policy tools to given objectives.
I shall argue that in the Swedish case the corporatist influences have worked
for continuity in priorities (1), which, in turn, as a reflection of changes in
constraints (2), has led to changes in policy tools (3). Changes in the
international economic environment were the main factor contributing to
the difficulties at the end of the 1970s as well as to the subsequent recovery.
This interpretation, which emphasizes the constraints that socially accept-
able goals set on admissible economic outcomes, is in line with the notion
of social corporatism as a non-exclusive system.13
The first signs of a deterioration in the relative performance of the Swedish
economy were already apparent in the late 1960s in the form of a weakened
external balance. In the early 1970s, the government pursued tight economic
policies in order to combat external deficits. Thus Sweden had a comfortable
11
See e.g. Söderström and Viotti (1979); Calmfors (1982). For a mild criticism of this view,
see Jackman (1989). The assessment by other social scientists may be more varied. Göran
Therborn (1986), for instance, seems to imply that the bourgeois governments in Sweden were
so tightly constrained by the ‘institutionalised commitment to full employment’ (p. 23) that they
had no alternative but to pursue full-employment policies.
12
Rowthorn and Glyn (1990).
13
Göran Therborn relies on a similar kind of explanation when he refers to ‘an in-
stitutionalised commitment to full employment’ (Therborn 1986). A problem in his interpreta-
tion is, however, that he to a large extent assumes this commitment to endure under all
circumstances and leaves its concrete implications somewhat vague.
surplus on current account until the first oil crisis. This was achieved at the
cost of a rise in unemployment to the highest level since the late 1950s.
Due partly to the fact that the resource-based export structure gained
from the commodity boom and partly to the accommodative reaction of
economic policies in the major OECD countries, the initial impact of the
first oil crisis was a boom in 1974. The inflationary thrust of this was
exceptional, as both export and import prices soared. The rise in export
prices was reflected in a profits explosion in 1973–4. This laid the ground
for a rapid increase in wages in the consequent years. In 1974 central
agreements were still aimed at moderate wage increases, but, as a reaction
to high profits and the booming economy, the negotiated wage increases
were overwhelmed by rapidly accelerating wage drift. With the public sector
leading the next wage round, even contractual wages were sharply increased
in the two-year agreements for 1975–6. In 1974–6 the average annual
increase in money wage costs in Swedish manufacturing amounted to nearly
19 per cent. Boosted by slow productivity growth and the appreciation of
the Swedish Krona,14 relative unit labour costs increased by almost 20 per
cent in 1974–6. This, together with the sudden slow-down in export prices,
was in turn reflected in sharply declining profit shares.
Under these circumstances, the continuity of the Swedish model was
represented by the priority given to low unemployment as a policy goal. The
approaches adopted by the policy makers in adjusting to internal and
external pressures without abandoning full employment can be divided into
three partly overlapping stages.
The first stage was the original bridging attempt after the first oil crisis.
Initially, it was expected to last over an international downswing of a normal
cyclical character only. Economic policies were expansionary and several
special measures, such as subsidies for stockpiling, were taken to keep the
export sector intact. When it became evident from 1976 onwards that the
international environment would remain gloomy for a long time and that the
structure of Swedish exports would make its effects still more damaging, the
bridging strategy was replaced in the second stage by attempts by the new
bourgeois government to shelter the economy more permanently from the
international crisis.15 While subsidies to ailing manufacturing industries
were continued and many firms facing bankruptcy were taken over by the
state, the sheltered sector, the public sector in particular, was now assigned
the main role in maintaining full employment. Notwithstanding the fact that
the budget deficit increased rapidly, taxes were increased to finance the
14
The Swedish Krona remained in the snake until 1977 and was thus tied to the German
Mark.
15
The distinction between the bridging and sheltering strategies has been made by Andersen
(1990). For the weakness of the export performance of Swedish manufacturing, see e.g. Erixon
(1984) and Horwitz (1982).
expansion of the public sector and to suppress the real disposable income of
those employed and thereby to reduce imports and the external dependency
of the economy. As Andrew Glyn’s calculations in his essay in this volume
show, there was in the social corporatist economies, and in Sweden in
particular, a marked decrease in the share of consumption out of labour
income in the market sector. Overlapping with these sheltering measures
Sweden abandoned the currency snake in 1977 and a series of devaluations
followed the two previous devaluations against the snake in 1976 and early
1977. The trade unions agreed to wage restraint and real wages declined.
The recovery of 1978–9 brought a short respite. But now the government
showed its indeterminacy by continuing expansionary policies. Inflation
accelerated. When the next international recession following the second oil
shock arrived, it seemed that the government had lost its faith. Internal
controversies among the bourgeois bloc increased. Relations between the
government and the trade unions were strained and there was a short general
strike in 1980. Unemployment started to increase. This, together with some
cuts in the welfare sector which acquired great symbolic significance, gave
the Social Democrats a case for claiming that the government was preparing
to abandon the Swedish model.
The third phase of adjustment, that of export-led growth, was launched with
the 1982 devaluation by the new Social Democratic government. The inter
national recovery was soon to follow. As was explained above, the recovery
that started in Sweden in 1983 was in many ways impressive. The main
problem confronted by the government now was how to check the inflation-
ary pressures.
I shall next deal more closely with two issues:
1. What is the overall assessment of the Swedish performance over the
post-1974 period as a whole?
2. What role has social corporatism played in it and what are its future
prospects?
1. As far as the assessment of Swedish performance is concerned, the above
interpretation amounts, at a general level, to a broad Keynesian argument
according to which investment in buying time, i.e. fighting the recession in
the 1970s by expanding the sheltered sector and by subsidizing the open
sector, turned out to be successful after all. This strategy was in many
respects similar to ‘Austro-Keynesianism’. But while the latter was praised
and even adopted as the standard for assessing the performance of other
countries in the early 1980s, the Swedish strategy was strongly criticized.16
Yet Austria was accumulating difficulties for the future while Sweden was
more successful in restructuring.
This consideration apart, the critics of the Swedish case maintain that the
16
See e.g. Scharpf (1987: 96).
sheltering attempts became futile and only prolonged the crisis when it
became clear, say, by 1977 that there was a structural crisis in Swedish
manufacturing. Two considerations are relevant in assessing this claim:
(a) What was the alternative to sheltering in the conditions of the late 1970s
and what would have been its consequences? What the critics have usually
had in mind would have been a more rapid restoration of external equi-
librium and budget balance. This would most likely have implied an increase
in unemployment, which, in turn, would have diminished domestic cost
pressures. I shall compare these two alternatives more closely below when
discussing the Finnish case, which comes closest in the Nordic context to
the kind of rapid adjustment of domestic demand proposed by the critics.
Did sheltering inhibit the necessary restructuring of Swedish manufac-
turing and prolong the crisis of the open sector? One can first point out that
the main phases of the Swedish adjustment are clearly delimited: rapid pub-
lic sector growth, subsidies, public indebtedness, and balance-of-payments
deficits until 1982 and a reversal thereafter. While productivity growth was
extremely low in the first period, it recovered in the latter. Second, the series
of discrete devaluations undertaken to restore the price competitiveness of
the Swedish open sector had been started quite early. The first devaluation
took place in 1976, and it was followed by two more in 1977 and one in
1981. The effective exchange rate of the Swedish Krona depreciated by
almost 20 per cent as a result of these measures alone. While the rise in
money wage costs was at the same time on a downward trend, reaching a
low point of 6 per cent in 1982, relative unit labour costs in Swedish
manufacturing started to fall in 1977 and the decrease amounted to about
20 per cent by 1979. After the devaluation of 16 per cent in the autumn of
1982, relative unit labour costs in Swedish manufacturing, measured in
common currency, were about 20 per cent lower in 1983 than in 1974 when
the deterioration of the relative cost situation had started. It is questionable
whether an alternative strategy would have brought better results in this
respect.
Third, considering the decade from the mid-1970s onwards as a whole,
the manufacturing sector in Sweden has maintained its position and its
structural transformation has been quite successful by international stand-
ards. This is generally recognized as far as the 1980s are concerned, but even
in the late 1970s Sweden’s record in restructuring compared well with other
countries.17 In fact, in Sweden certain policy measures associated with
sheltering, such as extensive manpower programmes, were targeted at re-
structuring. By the early 1980s, the number of people on special employment
schemes rose to about 5 per cent of the labour force. Had this number been
added to the unemployment pool, the rate of open unemployment in Sweden
17
See OECD (1987); Rivlin (1987: ch. 3), in which the structural change in Sweden is
compared with the USA; see also Landesmann in this volume.
would have amounted to more than 8 per cent, not much lower than in
Europe on average. The social and economic costs of such an open un-
employment alternative would have been high.18 Increasing open unemploy-
ment would probably have involved hysteresis effects so that restructuring
would hardly have been faster. Moreover, active manpower policies that
increase labour mobility can be regarded as a supply side precondition for
low wage dispersion in Sweden.
The above is not meant to imply that the Swedish strategy for fighting
unemployment was without errors. Critics have a point, e.g. as far as the
expansionary policies during the short upswing of 1978–9 are concerned;
they generated imbalances which worsened the situation in the next reces-
sion in the early 1980s. And, as pointed out by Michael Landesmann in this
volume, the explosion of subsidies in the late 1970s was to a large extent
defensive in character and did not contribute very much to restructuring in
any direct way. Much of the credit for the success of Swedish restructuring
must be given to the efficiency and flexibility of the corporate sector. Apart
from the bankrupt firms in the crisis sectors that fell into the hands of the
government, firms in the open sector pursued a restructuring process of their
own. The dominance of relatively big firms and the export orientation of the
Swedish manufacturing industry facilitated its adjustment to the new market
situation.
2. As far as the relevance of corporatism for Swedish performance is
concerned, it should first be recalled that, according to the interpretation
given above, external factors were the main factor behind both the rapid
recovery after 1982 and the difficulties preceding it. The influence of cor-
poratism should be traced to the ways the domestic economy managed these
changes in the environment. The discussion can be divided into two issues:
(a) Corporatist influences on economic policy priorities;
(b) the influence of corporatism on the behaviour of the private sector.
(a) The solidity of the tripartite bargaining system had, for several decades,
been based on a commitment by economic policy makers to maintain full
employment under all circumstances. This gave rise to internal pressures
under which economic policies were implemented. Initially, they led to
policies that, by relying on at times almost desperate bridging and sheltering
measures, did not allow unemployment to increase. Then, in the early 1980s,
18
Yet some scholars have claimed that the manpower programmes and subsidies, together
amounting to about 5% of GDP at their peak, did little to foster restructuring and were an
expensive way of maintaining employment. See Standing (1988). For a survey of studies of
Swedish labour market policies, see Bjorklund (1990), which comes to the conclusion that no
clear positive effects are indicated by the studies, except for intensified employment service, and
that the question of the efficacy of Swedish labour market policies remains open. A more positive
account is given e.g. by Jackman (1989).
when the bourgeois government started to have doubts about whether or not
to go on with sheltering, they led to a change in government.
(b) But had the corporatist influences not affected the behaviour of the
private sector and its readiness to share the burden of adjustment in an
equitable way, merely fixing the full-employment target might only have led
to an accumulation of difficulties over time. In fact, as has already been
indicated above, such influences can be discerned in the Swedish case. After
the overshooting of wage increases in 1974–6, the wage restraint imposed by
labour market organizations was effective. Real take-home pay in the private
sector decreased for four years in the period 1976–82 and there was still a
decrease in the pre-tax real wage in 1983. The impact of wage restraint on
the profitability of the open sector was enhanced by devaluations. In addition
to decreasing real wages, the real take-home pay of those employed was
further reduced by tax increases used to finance job creation in the public
sector (see also Andrew Glyn in this volume). This casts doubt on the
critique of accommodation, according to which a combination of trade
union monopoly power and full employment guaranteed by expansion of the
public sector leads to a crowding-out of the open sector through excessive
wage increases.19
Wage restraint is by now a classical economic argument for corporatism.
But the influence on behaviour of inherited corporatist institutions was
broader. As was indicated in section 10.2 above, inherited national policy
legacies have a momentum of their own. So far as they are embodied in
certain administrative capabilities and reaction patterns, they may facilitate
the readiness to react to certain kinds of shocks. The Swedish case provides
an example. The government was well equipped to make extensive use of
active labour market policies. Many other countries did not have such
readiness and were driven to open unemployment instead.20 Moreover, the
Nordic countries possessed some experience of the beneficial effects of devalua-
tions.21 There was also a more general legacy of an export-led growth
strategy to which these countries had been orientated. This was built into
the behaviour pattern of the private sector and fostered the restructuring
process of the corporate sector and the rapid investment reaction to the 1982
devaluation.22
19
See n. 11 above. For a further assessment of this argument see Matti Pohjola’s chapter in
this volume (ch. 3). It should be noted that some results reported in Calmfors (1990) find
support for the view that accommodating employment policies increase real labour costs
whereas changes in income tax rates seem to leave real labour costs for employers unchanged.
20
See Weir and Skocpol (1985); Therborn (1986); Jackman (1989).
21
The Swedish devaluation in 1949 was regarded as a success that rendered the Swedish
manufacturing industry highly competitive for two decades to come. Finland, in turn, had a
legacy of its own growth-boosting devaluations in 1957 and 1967.
22
From this point of view, contrasting the opening measures of the Swedish Social
Democrats with the strategy adopted by the French Socialists in 1981–2 is illuminating. The
French started their recovery programme with an expansion of domestic demand and delayed
What then has been the overall importance of the corporatist influences
for the success of Sweden? There is no simple answer to this question.
Rather, there has been a dynamic process involving the interaction of a
multitude of factors. In this process, success has fostered success. When full
employment under all conditions was given top priority in economic policy,
workers did not resist restructuring. With this taken for granted, the wage
restraint imposed by centralized wage agreements made firms ready to
respond to their enhanced competitiveness by investing more. Uncertainty
concerning die goals of economic policy and future bargaining did not cause
an adverse reaction of expectations to the extent which depressed private
demand, delayed recovery, and depressed investment in many other
European countries that were also successful in restoring price competitive-
ness and profit shares by the early 1980s.
My final comments on the Swedish case amount to some speculations
concerning its prospects. In a long historical perspective, a central feature of
the Swedish model has been the crucial and independent role the trade union
movement, more specifically the LO, has played in its formation and
implementation. Since the early 1950s, many of the new policy ideas which
subsequently turned out to be of lasting influence were first proposed by the
LO. It has also been an established tradition in Sweden that government
does not directly intervene in wage negotiations.23 The political and or-
ganizational risks inherent in wage restraint have been faced by the labour
market organizations. Collective wage restraint is a potential source of
conflict between the leadership and the rank and file in the trade union
movement. But it has been able to overcome this by insisting on the principle
of solidarity—full employment and equalization of wage differentials—and
on more far-reaching economic and social goals.
The traditional basis of the Swedish model is, for many reasons, under
pressure to change. The production structure and the composition of the
labour force have changed. While the overall degree of unionization has
remained high, the two central organizations of white-collar employees have
increased their share, and in addition within the LO there has been a shift in
the balance of power from the open sector towards those employed in the
closed sector. In the 1980s, these shifts have resulted in difficult intersectoral
wage disputes. The traditional model of wage determination in which the
open sector is supposed to play the determining role has come under doubt.
Combined with pressure from employers, structural change in the or-
ganization of labour has also fostered decentralization tendencies in wage
bargaining. These were accelerated by the profit boom after the 1982
the more significant devaluation until 1983. Furthermore, the private sector in France did not
have confidence in the Socialists’ policies. This was reflected in a capital flight.
23
For the Swedish bargaining system, see Elvander (1988a).
devaluation, which threatened to cause a breakdown in the centralized wage

negotiation system in 1983. It required a great effort by the government
before some form of central bargaining could be restored. Shifts between
more and less centralized bargaining have occurred before. In fact centraliza-
tion has often been a response to the acceleration of inter-union wage rivalry
under decentralized bargaining. Something like this has occurred during
more recent years.24
Thus recurrent decentralization tendencies do not necessarily mean that
the system has changed in a fundamental way. But changes in technology,
the organization of production, and the structure and skills of the labour
force may now be taking place that have put the negotiation system under
more irreversible pressure to change and as a consequence of which firm-
specific factors play a more prominent role in wage formation than before.
While local-level organization is strong in the Swedish trade union move-
ment and has had a say, for example, in the decisions concerning new
technology, these developments need not weaken the positions of trade
unions as such. But they may involve a shift in strategic power from the
central to the local level and at the same time increase the pressure for greater
flexibility of relative wages. Thus it is equality among wage earners and the
chances to countervail potential inequality tendencies through education,
labour mobility, and the like which are at stake.
The Swedish system has been unique in providing a framework for an
effective exchange between voluntary wage restraint and economic policies
aimed at full employment. But this sharing of responsibility is under pressure
to change. The temptation for economic policy makers to intervene in wage
bargaining has increased, as their room for manoeuvre in monetary and fiscal
policies has diminished and internal divisions in the trade union movement
have reduced its capability to impose wage restraint. In Sweden, government
involvement in wage negotiations increased after the 1982 devaluation and
at times there has not been much difference from formal incomes policies.
It remains to be seen whether the Swedish system is moving permanently
towards a more ‘statist’ version of corporatism. This might worsen the
free-rider dilemma inherent in collective wage restraint, as the government
could be blamed for the breakdown of wage restraint. It could also erode the
internal solidarity of the trade union movement and threaten the integrity of
the whole system of industrial relations.
The debate on wage earner funds may have resulted in a further weakening
of the role of the trade union movement in controlling the conflict over real
wages and the functional distribution of income in Sweden.25 By the late
1970s, the LO’s fund proposal had become a political burden to the Social
24
After all restrengthening centralized bargaining in the new environment was what the ‘new
EFO’ or ‘FOS’ project recently argued for. See Faxen (1989).
25
SeePontusson (1987).
Democratic Party. Although it never dissociated itself from the LO’s
proposal it played down its role in order not to endanger its election chances.
The party won the election and the retreat on the fund issue was widely
regarded as a loss to trade unions. One may question whether the increase
in relative power among the left towards the party will indirectly shift the
balance of power in the corporatist triangle towards government and in this
way affect the preconditions for social corporatism. Be that as it may, the
meagre outcome of the fund debate shows the limits of structural reforms in
a system where the mandate to make investment decisions is tied to owner-
ship of capital. This also sets limits on the corporatist agenda.
The current overheating of the economy has brought many of the under-
lying difficulties of the Swedish model into a sharp focus. A cyclical upswing
of long duration combined with an annual increase of nominal earnings of
well over 10 per cent and with a persistently low rate of growth of labour
productivity (in the range of 1–2 per cent annually in the 1980s) has sharply
intensified inflationary pressures and deteriorated the external balance.
Frustrated by the strong resistance among various pressure groups to its
attempts to tame the overheating, the government proposed a package of
direct control measures, including price controls and a ban on strikes, in
winter 1990.
This package represented a new phase in governments’ involvement in
wage settlements. It was widely rejected even among the Social Democrats.
The government resigned, and after certain manoeuvres, which damaged the
popularity of the Social Democrats, Ingvar Carlsson set up a new cabinet
with a new Finance Minister and the package was passed in a diluted form
in which it did not directly attack the conflict rights of the trade unions.
Although too much should not be made of the present crisis—the govern-
ment after all came out of it with a package largely involving the deflationary
thrust it had originally aimed at—it is nevertheless symbolic of the underlying
difficulties of the Swedish model. The traditional spheres of responsibility of
government and the trade unions in the implementation of the model are in
a state of flux, and the problems inherent in this are rendered more difficult
by the weak productivity performance of the economy.
10.4 NORWAY: SYMPTOMS OF FRAGILITY?
Traditionally, Norway has been regarded as lying second to Sweden among

the Scandinavian countries as far as the strength of its corporatism is
concerned. Norway has also been one of the most stable economies in the
OECD area in the post-war era. But, in the course of the 1980s, the
Norwegian economy has encountered difficulties (see the economic in-
dicators in Table 10.2). The oil sector exposed it to new types of external
TABLE 10.2 Economic indicators on Norway, 1973–1988
in real in real in real ment rate ment ratea market in con- Real Nominal Negoti- Wage ive in real account
GDP exports domestic meas- sumer hourly hourly atedwage drift exchange output per (% of GDP)
(%) (%) demand ures prices earnings earnings rate person
(%) (%) employed
1973 4.1 8.3 6.5 1.5 66.4 7.5
— 2.9 10.6 4.8 7.2 109.3 6.1 82.3 –1.8
1974 5.2 0.7 7.1 1.5 66.3 9.4 7.4 17.5 17.3 8.0 118.8 4.2 85.2 ^.8
— 20.3 119.9 –6.7 90.4 –8.5
1975 4.2 3.1 5.9 2.3 67.7 1.1 11.7 7.7 9.2 4.1
1976 6.8 11.3 7.4 1.7 70.6 0.6 9.1 7.3 17.1 9.6 5.8 122.6 –0.7 96.0 -11.9
1977 3.6 3.6 3.5 1.4 71.7 0.6 9.1 1.7 10.9 3.0 7.0 125.9 –0.6 100.7 -14.0
1978 4.5 8.4 –5.6 1.8 72.5 0.6 8.1 –0.1 8.0 2.1 3.9 116.4 1.0 95.5 –5.2
1979 5.1 2.6 3.9 2.0 72.7 0.7 4.8 -1.8 2.9 0.3 2.5 113.0 6.2 88.6 –2.2
1980 4.2 2.1 4.9 1.6 72.8 0.8 10.9 –0.9 9.9 5.1 8.3 113.6 –2.5 90.2 1.9
1981 0.9 1.4 0.9 2.0 73.4 0.9 13.7 –2.9 10.4 1.8 6.5 116.3 0.6 95.0 3.8
1982 0.3 –0.1 1.9 2.6 73.0 0.9 11.3 –0.9 10.4 5.5 6.3 119.8 2.6 98.3 1.1
1983 4.6 7.6 1.3 3.4 72.6 1.3 8.4 0.3 8.7 1.0 5.1 116.1 7.2 98.8 3.6
1984 5.7 8.2 6.0 3.1 73.1 1.8 6.3 2.1 8.5 3.0 5.7 113.1 3.6 97.3 5.3
1985 5.3 6.9 4.7 2.6 74.4 5.7 1.6 7.4 0.5 6.5 110.6 2.8 96.8 5.3
— 10.3 104.2 –2.5 97.1 –6.4
1986 4.3 1.9 8.1 2.0 76.5
— 7.2 2.9 3.7 8.4
1987 3.4 0.3 –1.0 2.1 77.3 8.7 6.8 16.1 0.2 7.5 100.0 3.0 100.0 –5.0
— 99.7 101.9
1988 2.6 5.9 –1.1 3.2 — — 6.7 — — 1.7 2.0 — —
a
b
c
pressures. Norway may also exemplify some of the problems attached to
corporatism in combining full employment and flexibility. To assess these
claims, I shall first describe the interplay of external shocks, policy reactions,
and wage bargaining since the 1970s.26 This is followed by a cautious
assessment of the performance of and future prospects for Norwegian
corporatism.
As in all three Nordic countries discussed here, the initial impact of the
first oil crisis was a boom, with soaring profits and increasing inflationary
pressures in 1974. In Norway, this overheating in an atmosphere of impend-
ing global slump was faced by a Social Democratic minority government,
the authority of which had been weakened by the 1972 referendum on the
EEC pact and by the party’s defeat in the parliamentary election in 1973.
Given the high priority placed on full employment and the prospect of
rapidly increasing revenues from North Sea oil, Norway, like Sweden,
adopted a bridging strategy, which turned out to be much more prolonged
than was originally expected.
While insisting on wage moderation, the government relied on expan-
sionary policies to protect Norway from the effects of the international
recession. Private consumption was boosted by real wage increases exceed-
ing on average 7 per cent annually in 1974–6. There were generous increases
in subsidies to ailing manufacturing industries and agriculture. New social
policy programmes were launched, and public expenditure increased
rapidly. The growth of domestic expenditure kept the labour market tight.
Deficits in the government’s budget, which had traditionally been in surplus,
and in the current account started to increase. But, with expectations of
increasing oil revenues and an end to the world recession, this was accepted.
In 1975, Norwegian exports turned down and the external equilibrium
started to deteriorate much more rapidly than had previously been expected.
In 1976–7, the cumulative current account deficit exceeded 25 per cent of
annual GDP. Profitability in the manufacturing industry declined sharply,
as the rise in domestic costs continued and adherence to the European snake
had resulted in an appreciation of the Norwegian Krone. Despite the
increasing economic difficulties, the Social Democrats succeeded in
strengthening their position in the 1977 election. As a matter of fact, their
policies were not much different from those of the bourgeois cabinets in
Sweden in the late 1970s. Although the main thrust of policies was on
sheltering and fiscal policies were accommodating, policies were gradually
tightened and the emphasis shifted to improving the profitability and com-
petitiveness of the open sector. A series of devaluations of the Norwegian
Krone were implemented when Norway left the currency snake in 1978.
With the prospect of increasing oil revenue to rely on, Norway was more
26
This account of the developments in Norway relies heavily on Fagerberg et al. (1989).
ready than Sweden to draw on its current account deficit. In Norway the
improvement in the price competitiveness of manufacturing industries was
nothing like as big as in Sweden. Nor was there any squeeze on real incomes
comparable with that in Sweden, although real earnings decreased some-
what after their record high increase in the mid-1970s. The expanding oil
sector substituted for a restructuring of the traditional open sector and the
existing structure was maintained with subsidies. Reflecting Norwegian
traditions, policy measures involved a regulatory thrust. The government
legislated a wage freeze to last until the beginning of 1980 and this was
combined with price controls. Control of the housing market was tightened.
Financial market controls were also maintained and the share of the state
banks in the total credit supply was further increased.
In many respects, Norway coped with the late 1970s quite well. Full
employment was maintained, inflation was brought under control, and the
relative unit labour costs of Norwegian manufacturing fell slightly. The state
of the economy was further enhanced by the international recovery in
1978–9. This made it possible for the government to start tightening its
fiscal stance. The economic outlook became even brighter with the rise in
oil prices in 1979. By 1980, the combined share of oil and natural gas in
total Norwegian exports exceeded 30 per cent. Revenue from oil in the form
of taxes and excise duties accounted for 30 per cent of the total revenue of
the central government in the early 1980s.
After the 1981 election a new minority government was formed by the
Conservatives, who had called for fundamental changes in the traditional
Norwegian model. The fact that Norway was the only one of the three
Nordic countries to become, for some time, a laboratory for new conser-
vatism was partly due to purely accidental questions of timing. While in
Sweden the bourgeois cabinets were in office during the most difficult
period of the late 1970s and the recession of the early 1980s, in Norway the
Social Democrats paid the political price for it. But the fact that the Conser-
vatives’ challenge was echoed so well among the electorate that their share
of total votes exceeded 30 per cent—almost 10 percentage points higher
than in the two other countries—was due to more fundamental factors. In
Norway, the strategy of the Social Democrats had differed from that in
Sweden in two respects. First, it had relied more on direct state control and
regulation in its policies on prices, incomes, housing, financial markets, etc.
Second, Norway’s Social Democrats did not manage to gain wide support
among the rapidly expanding group of middle-class voters. This constitu-
ency responded to the Conservatives’ call for lower taxes, in particular.
The Conservative government took office just at the time when Norway’s
oil revenues were accelerating sharply. The value of exports of oil and natural
gas almost doubled from 1981 to 1985. Thanks to oil, the surplus on the
current account amounted to roughly 5 per cent of GDP in the mid-1980s
despite the fact that traditional trade was heavily in deficit. But the effects of
the international recession were felt particularly strongly in the manufactur-
ing industries, where employment fell by more than 6 per cent in 1982–3.
While economic policies were tightened somewhat at the same time, the total
rate of unemployment exceeded 3 per cent in 1983, although an increasing
share of the slack in the labour market was concealed by the expanding
manpower programmes.
The weakening in economic performance was still modest by contem-
porary international standards. But combined with certain belt-tightening
measures interpreted as attacks on the welfare state, it was enough to evoke
a strong political reaction. In response to it, two other major non-socialist
parties agreed on a coalition government with the Conservatives in the spring
of 1983. The new coalition was driven to lax pro-cyclical policies in the
upswing of 1984–6. Domestic demand increased at an average annual rate
of 6 per cent in this period. The combination of international recovery and
domestic expansion generated a boom. The unemployment rate fell to 2 per
cent by 1986. But the expansion led to the accumulation of imbalances in
the form of inflation and a deteriorating external balance, which was further
aggravated by the oil price collapse. The current account deficit exceeded 5
per cent of GDP in 1986 and 1987.
The coalition government resigned in the spring of 1986 and was replaced
by a Labour minority government. In some respects, the Brundtland cabinet
returned to traditional Norwegian policies in the form of fairly heavy direct
state involvement. But now policies were geared towards fighting the ex-
ternal imbalance and enhancing the productivity and competitiveness of the
manufacturing sector. The Norwegian Krone was devalued in 1986. The
government legislated strict wage control. Wage increases started to slow.
Price inflation decelerated less rapidly than wage increases; this, together
with tightened fiscal measures, contributed to an almost 8 per cent decline
in the real disposable income of households in 1985–8.
The Brundtland cabinet resigned in the autumn of 1989 and was replaced
by a bourgeois coalition. It is still too early to assess its success in attempting
to balance the Norwegian economy. The current account deficit has been
corrected, the budget balance has been strengthened, profits have increased,
and the manufacturing industries have started to recover. But the other side
of the coin has been an increase in the unemployment rate to over 4 per
cent—the highest in Norway since the 1930s.
Both Sweden and Norway exemplify strong continuity implanted in policy
priorities by established corporatist structures.
From the current perspective it is very possible that the international
upswing will bring such relief that even the Norwegian case will provide one
more piece of evidence for the proposition mentioned in the context of
Sweden above, that those countries which, so to speak, bought time and,
under the pressure of inherited policy priorities, resisted a big increase in

unemployment in the late 1970s and early 1980s have fared relatively
well.
But the Norwegian case is weaker than the Swedish one. The restructuring
of the non-oil sector was delayed and the problems of the late 1980s were
caused by earlier lax policies. In addition to these policy failures, there are
some institutional and structural problems which cast a shadow over the
future of Norwegian corporatism.
First, in Norway the political and institutional basis of corporatism has
been weakened. In the political arena, the Social Democrats have lost their
former majority position and the non-socialist bloc, further confused by the
rapid rise of the Progress Party, is internally divided. The degree of unioniza-
tion is low by Nordic standards and the LO, which has traditionally played
the main role in wage settlements, has lost ground. Wage drift has started to
play a more important role for wage increases in the 1980s and wage
differentials have started to widen.
More generally, partly as a result of the previous policy failures, the
reliance on direct state intervention, traditionally heavily used in Norway,
has lost its credibility. But there has not necessarily been any general decrease
in government intervention. Its contents have changed, however. While the
government has lost some of its traditional policy options and at the same
time the accumulated imbalances have called for stronger intervention to
restore the external balance, greater use than before has been made of
incomes policies imposed by the government. Such a development may
increase resistance to incomes policies: there is a danger of the ‘British
disease’ where the corporatist alternative is criticized from the right and the
left and is in disarray because of a lack of political support.
The foregoing implies that social corporatism in the sense used in this
project may be eroding in Norway. This may be further exacerbated by the
fact that the structure of the Norwegian economy has changed in a way that
may undermine its traditional stability and thus complicate corporatist
settlements. The rise of the oil sector has increased external vulnerability. It
also sharpens the dualistic nature of the Norwegian economy. Apart from a
few isolated exceptions, e.g. in the electrical industries and industries con-
nected to offshore activities, no large-scale restructuring of the manufactur-
ing sector of’mainland Norway’ has taken place. Finally, deregulation of the
financial markets and the increasing importance of international capital flows
have weakened the traditional mechanisms of financial control. This change
has taken place more or less in all market economies, but its consequences
may be particularly acute in Norway because Norwegian economic policies
have traditionally relied on tight quantitative control of credit flows and the
disturbances emitted by the oil sector tend to be magnified by the reaction
of international capital movements.
10.5 FINLAND: STRENGTHENING CORPORATISM?
In the Nordic context, Finland provides an example of unsettled but gradual-

ly strengthening corporatism. Although Finnish corporatism in its present
form is of fairly recent origin, its roots reach further back in time. Various
attempts at corporatism have been made since the 1940s but these have been
repeatedly wrecked by external and internal pressures. The Finnish version
of corporatism also contains some elements of the traditional Finnish policy
model, which strongly reflected the interests of the export industries and
overlooked stabilization of domestic demand. This volatile historical back-
ground and mixture of old and new in its present version raise questions as
to the future sustainability of Finnish corporatism.
The strengthening of corporatism in Finland since the late 1960s is
evident by most indicators commonly used: the rate of unionization has risen
strongly; centralized wage negotiations have become the norm, with few
exceptions; these negotiations have often been connected with comprehens-
ive economic and social policy packages through which the government has
become an active partner in bargaining; strike activity has been in decline,
etc. This has coincided with a clear improvement in the relative performance
of the economy. Since the end of the 1970s, economic development in
Finland has been among the most stable in the OECD area and GDP has
persistently grown faster than in the European OECD countries on average.
Unemployment has decreased. There has not been as steep a slow-down in
productivity growth as in many other countries. Real wages have increased
without squeezing profits (see Table 10.3).
The combination of strengthening corporatism and improved economic
performance naturally raises the question as to whether there is a causal
relation between the two. Has corporatism led to better economic perform-
ance or has the latter kept the former intact? I shall argue that, apart from
strengthening corporatism, the improvement in relative economic perfor-
mance in Finland has been combined with many other factors, such as
diversification and diminishing external vulnerability of the economic struc-
ture, with a more stabilizing orientation of economic policies—and with
sheer good luck. In fact, to the extent that the strengthening of corporatism
has been a reaction to a more or less exogenous improvement in the
economy, the Finnish case rather illustrates the endogeneity of corporatism.
As far as the post-war period until the late 1970s is concerned, Finland’s
relative economic performance was characterized by the following features:
1. Over the period as a whole, output and productivity growth in Finland
roughly corresponded to the average of the OECD countries. During the
reconstruction period after the war, growth was relatively rapid, but it
stagnated in relative terms in the 1960s.
TABLE 10.3 Economic indicators on Finland, 1973–1988
in real in real in real ment rate ment ratea market in con- ive in real account
Real Nominal Negotiated Wage
GDP exports domestic meas- summer exchange output per (% of GDP)
hourly hourly wage drift
(%) (%) demand uresb prices rate person
earnings earnings
(%) (%) employed
1973 6.7 7.3 7.1 2.3 68.8 1.3 10.7 5.5 16.8 11.1 8.0 108.2 2.3 91.3 –1.9
1974 3.0 –0.6 7.5 1.7 70.3 1.3 16.9 4.7 22.3 11.3 13.7 111.6 1.1 98.7 –4.9
1975 1.2 –14.0 2.1 2.2 69.6 1.3 17.9 2.7 21.1 8.7 10.2 110.0 –3.8 105,8 –7.5
1976 0.3 12.8 –4.3 3.8 71.2 1.1 14.4 0.3 14.7 7.2 6.7 110.3 2.6 115.8 –3.7
1977 0.1 15.7 –0.9 5.8 69.5 1.1 12.6 –3.2 9.0 5.5 3.2 105.7 1.2 108.9 –0.3
1978 2.2 8.9 –0.3 7.2 68.3 1.8 7.8 –0.4 7.4 5.4 3.6 95.2 7.2 92.2 1.9
1979 7.3 8.8 9.2 5.9 69.7 2.3 7.5 3.7 11.5 4.9 3.1 94.7 7.1 90.5 –0.3
1980 5.4 8.4 5.5 4.6 71.6 2.4 11.6 1.1 12.8 11.3 4.6 97.5 2.8 92.1 –2.7
1981 1.6 4.9 –0.4 4.8 71.9 2.1 12.0 0.7 12.8 7.0 4.0 101.5 3.9 97.2 –0.8
1982 3.6 –1.1 4.3 5.3 72.2 2.3 9.6 0.9 10.5 8.4 3.1 103.8 3.3 101.2 –1.6
1983 3.0 2.5 2.3 5.4 72.0 2.3 8.3 1.1 9.5 5.6 3.6 99.7 5.4 99.3 –1.9
1984 3.3 5.4 2.0 5.2 72.2 2.5 7.1 3.1 10.4 4.0 3.8 101.9 5.3 103.4 0.0
1985 3.5 1.2 2.9 5.0 72.6 2.2 5.9 1.7 7.7 3.8 3.3 101.5 5.4 103.5 –1.3
1986 2.3 1.3 3.1 5.3 72.5 2.0 2.9 3.2 6.2 2.8 3.4 99.5 5.8 99.7 –1.1
1987 3.8 1.7 4.7 5.0 72.1 2.2 4.1 4.6 8.9 3.4 3.6 100.0 6.4 100.0 –2.2
1988 4.8 4.2 6.4 4.5 — 2.2 5.1 — — 5.3 4.0 101.2 — 104.3 —
a
b
c
2. The rate of structural change was very rapid. In the early 1950s, no less
than half the population was employed in primary industries. Since then, the
share of the agricultural labour force has decreased rapidly, to about 20 per
cent in 1970 and less than 10 per cent at present. As a reflection partly of
this, and partly of the high capital intensity of the main export industries, the
investment rate in Finland (like that in Norway) has been very high by
international standards—averaging about 25 per cent of GDP.
3. Cyclically, the economy was one of the most unstable in the OECD
area. Although the measured rate of unemployment remained on average
between 2 and 3 per cent, there were periods of high unemployment in the
late 1950s and late 1960s and continuous disguised unemployment in rural
areas before the jump in the unemployment rate to more than 7 per cent in
1978. Inflation was higher than in the OECD countries in general and there
was a persistent tendency of trade balance deficits. Both these problems
reached their climax in the mid-1970s: in 1975 the rate of increase in
consumer prices amounted to nearly 18 per cent while the current account
deficit rose to 7.5 per cent of GDP.
This mixed performance can be partly attributed to structural factors. The
markets for forestry products, the combined share of which in total exports
was still almost 80 per cent in the early 1950s, were cyclically unstable.
Contrary to, for example, the Norwegian case, which has also had a highly
unstable export sector, the import content of Finnish exports was fairly low
and the multiplier effects of cyclical export shocks were big. But the struc-
tural factors notwithstanding, cyclical swings were amplified by two factors
pertaining to economic policies and wage bargaining:
1. Reflecting the combined influence of the traditional fiscal orthodoxy,
aversion to government borrowing and the politically motivated attempt to
restrict the growth of the public sector, the reactions of fiscal policies were
highly pro-cyclical. Neither did monetary policy effectively prevent cyclical
swings in domestic demand.
2. The relations between unemployment, inflation, wages, profits, and
economic policies displayed regularities that extended over the normal trade
cycle. At roughly ten-year intervals (1949, 1957, 1968, and 1977–8), after
a prolonged period of high unemployment, a major devaluation of the
Finnish Markka took place just towards the end of a deep recession. These
devaluations were preceded by contractionary policies. The devaluations
changed the distribution of income towards profits, and various attempts at
wage restraint were taken to maintain this shift. But when labour markets
became tighter during the following international upswing, wage restraint
did not hold. Domestic inflation accelerated and the competitiveness of the
economy deteriorated again. This paved the way for a new devaluation after
a couple of normal export cycles.
In the Finnish discussion, these regularities have been referred to as the

devaluation cycle.27 They can be interpreted in terms of a Kaleckian political
business cycle reflecting the fact that the class relations of post-war Finland
persistently prevented the combining of low unemployment, low inflation,
and a stable functional distribution of income. Economic policies were
targeted at high investment and rapid structural change. This required a
favourable outlook for profits and devaluations were used to provide this.
But in order to be effective, it had to be ensured that the advantages they
brought were not immediately eroded by increasing wages. To prevent this,
before a devaluation, unemployment was deliberately used as a means to
provide workers with a lesson of the consequences of excessive wage claims
and to give the economy room to expand. After devaluation wage restraint
was striven for so as to preserve its distributive effects. Attempts at proced-
ures akin to corporatism were made during this phase of the cycle.
Given the prevailing institutions of industrial relations, policy makers had
limited power to influence wages. Attempts at extensive agreements on
wages and prices did not bring lasting results. There was a bitter general
strike in 1956. Shortly thereafter, the blue-collar trade union movement was
split into two central organizations. Union enrolment was on the decline. In
these conditions, the sound policy for unions, unable to behave collectively,
seemed to be to gain as much in wages as they could. In the vicious spiral of
devaluation, inflation, and unemployment, short-sightedness was institu-
tionally imposed.
After the failure of the 1950s, there were renewed attempts at incomes
policies in the 1960s. As one obstacle to this was formed by the split in the
trade union movement, the state and employers backed the negotiation
process which led to the reunification of the SAK (‘the Finnish LO’) in 1967.
A network of tripartite institutions between trade unions, employers, and
government was formed to convey information and to co-ordinate economic
and social policies with incomes policies. Agricultural prices and farm
incomes were also integrated with incomes policies. Comprehensive pack-
ages of centralized wage agreements and economic policies and social
policies have subsequently been quite common in Finland.
The new system of comprehensive incomes policies was created under
strong economic pressures. The devaluation cycle had reached the phase
where the currency was overvalued, profits were down, and unemployment
was high. The Finnish Markka was devalued by 31 per cent in 1967 and a
comprehensive incomes policy package was agreed thereafter. Thanks to
the boost given by devaluation and by the international recovery soon to
follow, the economy recovered rapidly towards the end of the 1960s.
27
The term and the formal presentation of the idea are due to Korkman (1978). See also
Pekkarinen(1989).
Unemployment decreased and profits soared. Wage restraint started to
erode, as was shown by, for example, a strike in the metal and engineering
industries in 1970. The failure of this phase of Finnish incomes policies was
made evident by the strong upswing in the early 1970s. The Finnish
economy was, like Sweden and Norway, in the peculiar position that the
impact of the 1973 oil shock was expansionary and strongly inflationary—
even more dramatically so in the Finnish case because the forest industries
accounted for a bigger share of Finnish exports. Moreover, during the
upswing preceding the first oil crisis, economic policies had been expansion-
ary in Finland, including a further depreciation of the currency in terms of
the trade-weighted basket during the slide of the dollar in 1972. Expansion-
ary policies were continued until the first half of 1975. While other areas of
economic policies were biased in the opposite direction, much emphasis in
the control of inflation was laid on incomes policies. But not unexpectedly,
wage control did not succeed under the conditions of tight labour market
and accelerating inflation. The centralized wage agreements were over-
whelmed by wage drift that accounted for the major part of the total increase
of more than 20 per cent in nominal wages both in 1974 and in 1975. In
1975, export revenues dropped sharply while the expansion of domestic
demand continued to imports and wage and price increases accelerated. As
a result, the competitiveness and profitability of the export industries rapidly
deteriorated and the current account deficit increased sharply.
In this situation, economic policies in Finland took a sharply different
course from those in Sweden and Norway. While the two other countries
insisted on keeping unemployment rates low and adopted bridging strategies,
reducing inflation and restoring external balance were given primacy in
Finland and economic policies shifted in a sharply restrictive direction.
Policy makers’ interest in incomes policies waned and unemployment was
deliberately used as a means to bring down inflation.28 Monetary policy was
tightened in 1975. Fiscal policy followed suit in the following year when the
discrete (structural) surplus in the budget of central government increased
by about 4 per cent of GDP. Tight policies were continued until the end of
1977 when the first devaluation took place in a series of three devaluations
in 1977–8, amounting in all to a 19 per cent increase in the effective exchange
rate. Before the relief brought by the devaluations and the international
recovery from the end of 1978 onwards, the economy was thrown into its
worst recession since the 1930s. Only when it became obvious that the
unions were ready to accept tight wage constraint were incomes policies
given a central role again in 1978. Trade unions accepted the shift in income
distribution. Once again boosted by devaluation and international recovery,
the Finnish economy embarked on a rapid upswing in 1978, and the
28
The practice of centralized wage agreements was not abandoned, however.
following year total output increased by more than 7 per cent. Unemploy-
ment started to fall.
In the 1980s, economic development in Finland has been extraordinarily
stable at a persistently higher level than in the European OECD countries
on average. Industrial relations have been more peaceful than before, and
certain tendencies toward more decentralized bargaining notwithstanding,
no breakdown in incomes policies has occurred.29 Is this achievement due
to strengthening corporatism, to wise policies, or to a favourable external en-
vironment? One can point out significant changes in all three sets of factors.
1. As far as the structure of the economy is concerned, there have been
several changes exerting a stabilizing influence. First, the Finnish economy
provides a good example of successful catching-up. The average annual rate
of growth of total factor productivity in Finland decelerated by less than half
a per cent from the 1960s to the early 1980s, which is exceptional by OECD
standards. Export diversification has continued. While forest-based products
still made up more than half of Finnish exports in the early 1970s, its share
has now fallen below 40 per cent and, apart from cyclical changes, has been
exceeded by that of the metal and engineering industries. Swings in these
two main groups of Finnish exports offset each other to some extent so that
exports as a whole have become more stable. A significant upgrading has also
taken place in the internal composition of both main export branches.
Finnish export industries, which like their Swedish counterparts are domin-
ated by relatively big firms, have succeeded quite well in diversifying.
Second, thanks to the geographical pattern of Finnish foreign trade, the
impact of the second oil shock was different in Finland from that in a typical
non-oil-producing OECD country. Finland imports most of her oil from the
Soviet Union. Trade between these two countries has been based on bilateral
agreements aimed at balancing trade over five-year periods. Consequently,
when the increase in the price of oil increased Finland’s import bill from
1979 onwards, exports to the Soviet Union also increased. This cushioned
the adverse developments in exports to Western markets.
Third, as Juhana Vartiainen and Michael Landesmann point out else-
where in this volume, the shift from the conflict-ridden devaluation cycle to
a more co-operative mood of play in industrial relations may reflect a change
in the structural conditions under which the distribution of income between
wages and profits is determined. Until the 1970s, Finland fitted the charac-
terization of a labour surplus economy with a reserve army of labour in
agriculture. This made the policy of big devaluations work by making real
wages responsive to them. Profits were increased and firms were able to
expand production by relying on workers migrating from the primary sector.
In these conditions, economic instability aggravated by economic policies
29
Wage agreements have been reached three times at industrial union level in the 1980s.
may have accelerated long-run growth by diminishing the power of organized
labour and thus increasing the average share of profits in industry. When
surplus labour in agriculture was exhausted, this conflictive accumulation
mechanism no longer generated growth but rather inflation and instability.
This may have paved the way for a transition from a political business cycle
to more co-operative procedures.
2. Economic policies in Finland have become more stabilizing during the
last ten years. Paradoxically enough, while the OECD countries in general
have become more wary of fiscal policies, Finland, which had previously
resisted fiscal activism, has just begun to make use of the idea of compens-
atory fiscal policy. The timing of fiscal measures has not been perfect, but at
least the clearly pro-cyclical fiscal shocks that destabilized the economy
before have been avoided. Moreover, the expansion of the public sector has
been rapid on average, reflecting the fact that many welfare programmes had
been started relatively late in Finland and no cuts have been implemented.
Exchange rate policies have tried to distance themselves from the devalu-
ationary policies pursued earlier. The Markka was devalued in two steps by
11 per cent just before and after the Swedish devaluation in 1982. On the
other hand, modest revaluations were implemented in 1979, 1983, and
1989.
3. Finally, industrial relations in Finland have been more peaceful than
earlier. Real wage increases have remained broadly within the limits set by
productivity growth. Temporary oscillations aside, the distribution of in-
come between profits and wages has been relatively stable. Stabilizing
changes in economic policies and industrial relations in Finland can partly
be traced to the strengthening of corporatism. But as far as the significance
of the latter for the relative improvement in economic performance is
concerned, the Finnish experience tentatively suggests that a reform of the
institutions pertaining to industrial relations alone does not guarantee
economic stability. This seems to be one lesson from Finland’s history of
devaluation cycles. The institutional framework for comprehensive incomes
policies was created in the late 1960s and its roots go back further in time.
Yet this structure did not prevent the accumulation of the imbalances caused
by economic policies and external shocks in the early 1970s. Other factors
are needed to explain the persistence of the co-operative mood of play since
the end of the 1970s. As was indicated above, such stabilizing changes have
taken place both in economic structure and in economic policies.
The post-war Finnish experience illuminates the endogeneity of cor-
poratism.30 It is also an interesting feature of the Finnish devaluation cycle
that attempts at corporatism seem to occur during upswings in times of
relatively high profits, i.e. after devaluations, while they are abandoned
30
In his essay in this volume, Matti Pohjola puts forward an analytical argument for the
endogeneity of corporatism. See also Kaitala and Pohjola (1990).
during the periods of unemployment preceding devaluations. This runs

counter to the argument frequently made which states that corporatism
makes a real difference during periods of economic difficulties.
In conclusion, comparing the Finnish case with the Swedish and Nor-
wegian cases suggests some comments:
1. An interesting difference between Finland on the one hand and Sweden
and Norway on the other is to be seen in the economic policy reactions in
1975–7. While the latter countries opted for the bridging strategy, the
Finnish economy was driven by sharply contractionary policies into a deep
recession. An argument—quite the opposite to the one for social corporatism
presented above—sometimes put forward is that Finland’s good perform-
ance since the late 1970s is due to the fact that the imbalances were corrected
at an early stage.31 The current account deficits were restored to a sustainable
level and the imbalances were not carried forward to a time when greater
adjustments would have been needed. Trade unions and employers learned
that accommodative policies were not to be expected and inflationary
expectations were defeated. These arguments bear general resemblance to
the orthodox claim for credible rules of economic policy. In fact, this was
frequently used in the opposite direction, e.g. in the Swedish debate, to point
out the destabilizing consequences of economic policies that give a full-
employment guarantee to the private sector.32 Yet, according to the inter-
pretation given above, the later Swedish, and with more reservations also the
Norwegian, experience shows that preserving full employment by bridging
and sheltering did not necessarily lead to accumulation of greater difficulties
in the long run. In Sweden, real wages were flexible and successful restruc-
turing took place.
2. The fact that the unemployment rate has been higher in Finland than
in Sweden or Norway exemplifies the country specificity of the priorities
connected with corporatist institutions. These possess a continuity of their
own. In a sense, Finland is a good example of a case where corporatism has
traditionally been conceived as rather constrained, or conditioned, i.e.
mainly as a means to adjust the goals of different groups to the economic
necessities largely seen to be determined by the external economic environ-
ment. It renders the adjustment more manageable.
This characterization of Finnish corporatism is different from the social
corporatism of the Swedish type. The former bears some resemblance to the
consensus version of corporatism discussed in the introduction to this book.
The fact that it has not been very quick to adherence to full employment in
the past can be seen as an indication of exclusion. But on the other hand the
organization of political and economic interests is firmly class-based in
Finland and the conflict rights of interest organizations are acknowledged.
31 32
See e.g. Calmfors (1984). See literature in n. 9 above.
Moreover, the recent increase in unionization has implied a change in the
underlying balance of forces in Finnish corporatism that may underline the
inclusive principle. Finnish corporatism is an interesting mixture of social
corporatism and consensus.
Finally, what implications do the endogeneity and mixed nature of Finnish
corporatism have for its future prospects? Are there good reasons for relying
on its stability in future or will it break down soon after economic cir-
cumstances change? The new pattern of policies and industrial relations has
so far been tested by relatively stable external conditions. The old instabilities
may well return in the event of serious external shocks. The underlying
principles of policy need not have changed much from the days of the
devaluation cycle to have resulted in stable policies in a situation where the
external balance has not been fatally disturbed. But there have occurred
certain countervailing, partly irreversible changes, such as the diversification
of the economic structure, ‘institutional learning’ in economic policies, and
an increase in the strength of the trade unions. Only time will show whether
these changes are enough to bring about a permanent shift in the priorities
of economic management and to preserve Finnish corporatism over extens-
ive periods of external pressures.
In the 1980s, decentralization tendencies in wage bargaining have been
prevalent in Finland as in many other countries. On two occasions no central
agreement was reached. In 1988, in turn, no less than half the unions rejected
the central agreement. The role of the government has become more visible
in shaping and enforcing the central agreement. In 1989, the government
announced in advance that taxes would be increased if the central agreement
was not accepted by the unions, and the signing of the central agreement was
postponed until it was known whether it had been accepted by individual
unions. These decentralization tendencies and the government’s reaction to
them have tended to diminish the authority of central organizations. If they
continue, they will undermine the present wage negotiation system. As
government-imposed incomes policies in the presence of weak central or-
ganizations have in many countries turned out to be an unviable system,
more decentralized bargaining would be a more likely outcome of such a
dissolution of the existing system.
10.6 CONCLUSIONS
A relatively good economic performance has been a common characteristic

of the three Nordic countries discussed in this chapter. In conclusion, I shall
summarize the factors behind this by classifying them into economic struc-
ture, economic policies, and institutions.
As far as the economic structure of the three countries is concerned, Norway
had its newly found oil resources to rely on during the critical years of the
late 1970s and the early 1980s. Finland benefited from an increase in its
exports to the Soviet Union just when its Western markets were in recession
in the early 1980s; furthermore, the Finnish economy was undergoing a
favourable diversification phase as part of its catching-up process. The
course of the Swedish economy, too, seems to a large extent to have been
determined by external impulses; the structural problems of the Swedish
manufacturing industries exacerbated the problems in the late 1970s while
later on relief was brought by the international recovery.
This chapter has emphasized the interaction between economic policies and
institutions, including corporatist structures. It is useful in conclusion to divide
the linkages between corporatism and economic policies into two groups.
1. The supply side linkages that have figured in the foregoing analysis
mainly consist of the real wage and price competitiveness (the real exchange
rate or relative unit labour costs in common currency). The fact that Sweden,
Norway, and Finland have made use of devaluations instead of a prolonged
deflation in the attempts to restore the competitiveness of their open sectors
suggests priorities between inflation and unemployment different from those
countries who have been prepared to take the hard-currency option even at
the cost of persistent unemployment. The relatively high degree of flexibility
of real wages in these countries, which is shown by several empirical studies,
suggests, in turn, a readiness on the part of the Nordic trade unions to accept
the distributive impact of devaluations.
Although most accounts of corporatism have relied on it, the link between
real wages and employment is not the sole explanation for the good relative
performance of the Nordic countries. As a matter of fact, since the late 1970s
and early 1980s, the Nordic countries have not performed very well as far as
the evolution of their relative unit labour costs in common currency is
concerned.
2. Aggregate demand forms a second link between economic performance
and corporatism. International differences in labour market performance
stem to a great extent from the first half of the 1980s. Empirical studies, in
turn, suggest that differences in the stance of fiscal and monetary policies
were important in explaining differences in economic performance during
this period.
The three Nordic countries avoided contractionary policies for different
reasons, reflecting differences in their corporatism. Owing to their strong
commitment to full employment this option was excluded in Sweden and
Norway. Governments which were preparing for austerity measures in the
early 1980s were defeated in elections. In Finland, on the other hand, there
was no need for contraction; the foreign balance and government finances
were in good shape and the impact of the deep recession in the West was
relatively modest thanks to the increase in exports to the Soviet Union.
This chapter has also dealt with corporatism as a partly endogenous factor
dependent on economic policies and economic performance themselves.
This has been discussed as a problem of the stability of corporatism. The
three countries can also be compared from this point of view. The Finnish
experience since the late 1970s illuminates the case of strengthening cor-
poratism. But it is an open question how resistant it will be to unfavourable
changes in the external environment. Sweden, in turn, represents a case of
fairly stable corporatism which has succeeded in combining continuous full
employment and structural transformation. Certain doubts can in turn be
raised concerning the stability of Norwegian corporatism as restructuring of
the traditional industries has been delayed and certain institutional and
political preconditions for corporatism also show some symptoms of weak-
ness. But from today’s perspective the Norwegian system seems to be
recovering quite well. Corporatism is thus conditional on the economic
success which it may, in turn, foster. But what causal power is then left for
corporatism? I have tried to shed light on the interaction between cor-
poratism and economic policies, as reflected in economic policy priorities
and in the room for manoeuvre in policy. But behind this macro-economic
context there is a broader and, in the last analysis, perhaps more important
question: how do the social, political, and economic forces pertaining to
corporatism influence the capacity of the economy to adjust to changing
circumstances at the micro level? This question opens up a wide range of
issues from the educational system and industrial policies to corporate
strategies and the role of the trade unions at the work-place which have been
briefly hinted at in this chapter.
The three Nordic cases represent somewhat different varieties of cor-
poratism. But one can also argue that there have occurred certain shifts in
each of them, as a result of which the three cases have approached one
another. While Finland has moved in the direction of more full-blown
corporatism, some ‘Finlandization’ has on the other hand taken place in the
Swedish economic policy strategy after 1982 and in the Norwegian one after
1986. Namely, quite in accordance with the traditional Finnish economic
policy priorities, emphasis in these two countries has been shifted towards
enhancing profitability. In its extreme form, such a policy strategy involves
an unconditional profit guarantee and represents a withdrawal from the
principles of the Rehn-Meidner model where profit squeeze is assigned the
role of speeding up structural change and productivity growth.
It has been argued in the Swedish debate that some consequences of this
withdrawal are already to be seen in the fact that Swedish exports have
actually gained in relatively low-technology industries in the 1980s (cf.
Erixon 1989). Moreover, as the Finnish experience of the devaluation cycle
exemplifies, from the point of view of stabilization, policies targeted at
enhancing profits involve the risk of a wage explosion which in turn paves
the way for future devaluations. One may question whether the present
difficulties in Sweden testify to the relevance of such risks.
Finally, what are the prospects for the three Nordic cases of corporatism?
One factor that may bear upon the future prospects is the shift which has
taken place in the pattern of government involvement. Partly as a result of
increasing integration, government intervention outside the labour market
has become more difficult. Financial integration has diminished the auto-
nomy of monetary policies which have become increasingly targeted at the
balance of payments. It is doubtful whether increased financial integration
any longer allows for successful discrete exchange rate adjustments.33 Also,
fiscal policies are for various reasons considered to be subjected to tighter
constraints than before. Under these conditions, macro-economic wage
adjustment has become more important since it has been required to
compensate for the loss of the freedom of manoeuvre elsewhere. This, in turn,
has increased the temptation to make more extensive use of government
intervention in the negotiations.
Government intervention in wage negotiations has also been stimulated
by a tendency toward greater decentralization of wage bargaining and the
increasing difficulties of the labour market organizations to reach a central
agreement on their own. These tendencies may well be different from the
previous cycles between centralized and decentralized bargaining ex-
perienced in all Nordic countries. First, owing to the ongoing change in the
industrial and occupational structure of the labour force, the craft-based
professional unions have become stronger at the cost of the industry-based
blue-collar workers’ unions. The former, while strongest in the sheltered
sector, are in a better position to push wage increases across industries which,
in turn, evokes compensation demands among the industrial unions in
manufacturing: there is a danger of a wage-wage spiral which is initiated in
the sheltered sector and in which organizational disputes play a large role.
Second, the role of firm-specific factors in wage determination is increasing
as is shown for example by the increasing importance of different types of
bonus systems. Third, as has been emphasized, for example, in the vast
literature on post-Fordism, the change from mass production towards more
differentiated products has, among certain sections of the labour force, been
connected with greater autonomy at work and differentiation of skills and
work careers. It has been claimed that trade unions have difficulties in
redefining their role to meet the expectations of this new labour force.
These changes may have weakened the preconditions for voluntary wage
restraint imposed by the central organizations. While at the same time the
33
In this respect, there is a difference between the typical Nordic view, which is described
by Sixten Korkman in this volume, and the view expressed by many foreign, especially
American, observers of the Nordic economies, like the Brookings Report on Sweden and
Cooper (1987).
importance of wage restraint in economic policy has increased, a con-
sequence has been more active direct government intervention in wage
formation. In a way, decentralization tendencies have been counteracted by
shifting the management of wage restraint away from the labour market
organizations to the government. It thereby becomes more politicized. The
experience of many countries suggests that such a system is fragile. It is
likely to decrease the autonomy of interest organizations and to endanger the
basis of corporatism by shifting the burden of the control of conflict away
from a struggle between capital and labour to that between the government
and unions and to internal disputes among the interest organizations them-
selves.
It is not only on the wage front that Nordic corporatism is under pressure
to change. Difficulties in maintaining public sector saving against a back-
ground of resistance to taxation and expanding expenditure requirements,
combined with the prospect of a diminishing financial surplus in the social
security funds sector, lead one to ask whether the Nordic countries face a
more acute shortage of saving than before or whether the fall in public saving
can be compensated for by an increase in private saving through tax con-
cessions or otherwise. Should the increase in private saving take place in the
household or in the corporate sector? Household saving has recently been
declining sharply in the Nordic countries and there are demographic and
other reasons for believing its longer-term prospects may not be good. Apart
from this, the former alternative would intensify pressures for greater income
differentials in household incomes than before. The latter alternative, in
turn, poses the question: on what conditions will workers be willing to accept
the increase in corporate profits it involves and what will be its implications
for the distribution of wealth. The outcome of the fight over wage earner
funds in Sweden suggests that Nordic societies are not yet ready to solve this
basic conflict.
The problems of redesigning the institutions influencing saving and wage
bargaining both face the uneasy trade-off between equality and economic
incentives. As such they pose a challenge to the Nordic types of corporatism,
which have so far succeeded in combining these two objectives in a way that
has maintained good economic performance and preserved the cohesion of
the existing institutions with their great emphasis on equality.
In many ways, the new tendencies evident in the nature of work, the
organization of labour, and government’s interest in wage restraint and in
the patterns of saving bring into focus the future role of the trade unions.
The Nordic trade unions are in a relatively good position to meet the future
challenges by new organizational and strategic solutions. One of their
greatest assets, provided by the experience of corporatism, is their estab-
lished status as partners in economic and social management at the local,
industry, and national levels. These types of basic institutional patterns are
largely irreversible and to a large extent set the landscape in which solutions
to new problems will be sought.
REFERENCES
ANDERSEN, TORBEN M. (1990), ‘Stabilization policies towards internal and external

balance: theory and some Nordic experiences’, in Calmfors (1990).
BJÖRKLUND, A. (1990), ‘Evaluations of Swedish labour market policy’, Finnish
Economic Papers, 3: 3–13.
BRAUN, A. R. (1986), Wage Determination and Incomes Policy in Open Economies,
Washington, DC: International Monetary Fund.
CALMFORS, L. (1982), ‘Employment policies, wage formation and trade union
behaviour in a small open economy’, Scandinavian Journal of Economics, 84:
345–73.
(1984), ‘Sverige och Finland: en jamforelse av stabiliseringspolitiska strategiar’,
Skandinaviska Enskilda Banken Kvartalsskrift, 13: 117–22.
(1990) (ed.), Wage Formation and Macroeconomic Policy in the Nordic Countries,
COOPER, R. N. (1987), ‘External constraints on European growth’, in Robert Z.
Lawrence and Charles L. Schultze (eds.), Barriers to European Growth, Washington,
DC: Brookings Institution.
CZADA, R. (1987), ‘The impact of interest politics in flexible adjustment policies’, in
H. Keman, H. Paloheimo, and P. F. Whiteley (eds.), Coping with the Economic
Crisis, London: Sage.
ELVANDER, N. (1988a), Den svenska modellen: Löneförhandlingar och inkomstpolitik,
Stockholm: Allmänna förlaget.
(1988b) (ed.), Fbrhandlingssystem, inkomstpolitik och arbetskonflikter i Norden,
Stockholm: Norstedts.
ERIXON, L. (1984), ‘Den svenska modellen i motgång: En analys av dess effekter och
förändrade förutsättningar under perioden 1974–1984’, Nordisk Tidsksnft för
Politisk Ekonomi, 15–16: 109–58.
(1989), ‘Den tredje vägen: inlåsning eller förnyelse?’, Ekonomisk Debatt, 3:
181–95.
ESPING-ANDERSEN, G. (1985), Politics against Markets: The Social-Democratic Road
to, Power, Princeton, NJ: Princeton University Press.
FAGERBERG, J., CAPPELEN, Å., MJOSET, L., and SKARSTEIN, R. (1989), ‘The decline
of Norwegian Social-Democratic state capitalism’, in P. Camiller (ed.), The Left in
Western Europe, London: Verso.
FAXEN, K. O., ODHNER, C.-E., and SPÅNT, R. (1989), Lönebildningen in 90–talets
samhällsekonomi, Stockholm: Raben & Sjögren.
Stabilization and Incomes Policies: European Experience, Washington, DC: Brookings
Institution.
GRUBB, D., JACKMAN, R., and LAYARD, R. (1983), ‘Wage rigidity and unemploy-
ment in OECD countries’, European Economic Review, 21:11–39.
HORWITZ, E.-C. (1982), ‘Export performance of the Nordic countries 1970–83’, in
Economic Growth in a Nordic Perspective, Helsinki: ETLA.
JACKMAN, R. (1989), ‘Where corporatism works’, LSE Quarterly, Autumn: 213–35.
KAHN, G. A. (1984), ‘International differences in wage behavior: real, nominal, or
exaggerated?’, Research Division, Federal Reserve Bank of Kansas City: Research
Working Paper 84–01.
redistribution in capitalism: efficient game equilibria in a two-class neoclassical
KORKMAN, S. (1978), ‘Exchange rate policy, employment and external balance’,
Bank of Finland B:33, Helsinki.
LEHMBRUCH, G. (1982), ‘Introduction: neo-corporatism in comparative
perspective’, in G. Lehmbruch and Ph. C. Schmitter (eds.), Patterns of Corporatist
Policy Making, London: Sage.
MCCALLUM, J. (1986), ‘Unemployment in OECD countries in the 1980s’, Economic
fournal, 96: 942–60.
MARKLUND, S. (1988), ‘Welfare, state policies in the tripolar class model of
Scandinavia’, Politics & Society, 16: 469–86.
OECD (1973–88), Economic Outlook, Paris.
----- (1987), Country Study: Sweden, Paris.
PEKKARINEN, J. (1989), ‘Keynesianism and the Scandinavian models of economic
policy’, in Peter A. Hall (ed.), The Political Power of Economic Ideas: Keynesianism
across Nations, Princeton, NJ: Princeton University Press.
PONTUSSON, J. (1987), ‘Radicahzation and retreat in Swedish democracy’, New Left
Review, 165: 5–33.
RiVLiN, A. B. (1987) (ed.), The Brookings Report on Sweden, Washington, DC:
Brookings Institution.
ROWTHORN, R., and GLYN, A. (1990), ‘The diversity of unemployment experience
SCHARPF, F. W. (1987), SozialdemokratischeKrisenpolitik inEuropa, Frankfurt: Cam-
pus Verlag.
SÖDERSTRÖM, H. T., and VIOTTI, S. (1979), ‘Money wage disturbances and the
endogeneity of the public sector in an open economy’, in Assar Lindbeck (ed.),
Inflation and Employment in Open Economies, Amsterdam: North-Holland.
STANDING, I. (1988), ‘Training, flexibility and Swedish full employment’, Oxford
Review of Economic Policy, 4: 94–107.
THERBORN, GÖRAN (1986), Why Some People Are More Unemployed Than Others,
London: Verso.
WEIR, M., and SKOCPOL, T. (1985), ‘State structures and the possibilities for
“Keynesian” responses to the great depression in Sweden, Britain, and the United
States’, in P. B. Evans, D. Rueschemeyer, and T. Skocpol (eds.), Bringing the State
Back In, Cambridge: Cambridge University Press.
11
Corporatism: Success or Failure? Austrian
Experiences
Alois Guger
11.1 INTRODUCTION
In the decade after the first oil shock the world economy experienced a marked
decline in output and productivity growth and a substantial rise in inflation
and unemployment.
All industrialized countries were confronted with these important changes
in the economic environment more or less to the same degree. However, we
can find a considerable increase in the variance of the performance measures
of individual countries since the early 1970s. Looking at Western Europe
even some highly open economies like the Scandinavian countries or Austria
have been much less prone to stagflationary processes than others.
In recent years, the economic policies and institutional arrangements of
these so-called ‘success’ countries have attracted growing interest among
social scientists. The institutional structure of a country promotes or restricts
strategic actions and determines the range of feasible policy options. On the
one hand, the degree of centralization in the bargaining process of a country
appears to foster the degree of co-operation between trade unions, em-
ployers, and government; and on the other hand, the political structure of a
society, i.e. the relative power positions of labour and business and the state,
exert a crucial constraint on the rank order of political priorities. Thus, it
seems that countries with powerful centralized trade unions and business
organizations, with a strong labour movement, and with highly centralized
bargaining arrangements have steered their economies significantly better
through the turbulent years since the first oil crisis than other countries.1
Since Austria has been considered as a paradigm of corporatism in all its
various classifications the performance of the Austrian economy in the
period of crisis since the early 1970s is of special interest.
The subject-matter of this chapter is the analysis of the Austrian economy
with respect to its institutional arrangements and economic policy. The
chapter is organized in five sections: first, we shall briefly comment on the
role of corporatist institutions in economics; second, we shall give a brief
The author wishes to thank Adalbert Knobl (IMF), Wolfgang Pollan (WIFO), Kurt W.
Rothschild (WIFO), Josef Steindl (WIFO), and Ewald Walterskirchen (WIFO) for comments.
1
Scharpf (1984), Bruno and Sachs (1985), Blaas and Guger (1985) written in 1977.
CORPORATISM IN AUSTRIA 339
account of the development of the Austrian economy during the period of
crisis and its aftermath; in section 11.4 a description of the Austrian
economic strategy of this time—the so-called ‘Austro-Keynesianism’—and
its underlying institutional arrangements and policy strategies will be given;
section 11.5 will deal with wage and labour market flexibility, the main
concern of prevailing orthodox economics in employment policy; and in
section 11.6 we shall, finally, discuss some major problems of this approach
and give an evaluation of the Austrian model and its prospects.
11.2 CORPORATISM AND ECONOMIC POLICY
The recent interest of economists in corporatism stems from the role that
mainstream economics attaches to labour markets in employment matters.
As in the period of mass unemployment in the 1930s, neo-classical eco-
nomists have again diagnosed inflexible labour markets, i.e. too rigid real
wages and low labour mobility, as the main reason for the high unemploy-
ment over the last fifteen years.
Some of the main elements of corporatism, such as powerful, centralized
trade unions and nation-wide bargaining systems, used to be criticized by
‘free-market economists’ as monopolistic elements and hence impediments
to the smooth functioning of competitive markets. But since in recent years
a number of studies seem to indicate that such corporatist features foster
smooth real wage adjustments, corporatist institutions have more and more
been discussed as a possible instrument to improve labour market perform-
ance, at least for countries with powerful trade unions.
From a Keynesian perspective, however, the meaning of corporatist
institutions goes beyond the labour market. As far as the labour market is
concerned, trade unions can only act upon nominal wages, not real wages.
In this model, inflation is determined by wage bargaining and ‘incomes
policies’ are seen as necessary elements of viable full-employment strategies:
when the growth of money wages exceeds productivity growth the price
level will increase. Since trade union power increases with rising employ-
ment, bargaining institutions and trade union co-operation are crucial in
this model to control inflation at full employment.2
But the effects of corporatist institutions go beyond the labour market.
2
Keynes and his group in Cambridge had clearly seen the problem of accelerating inflation
already at the time of the publication of the General Theory: their position is probably best
summarized in Joan Robinson’s statement in 1943: ‘Unemployment in a private enterprise
economy has not only the function of preserving discipline in industry, but also indirectly the
function of preserving the value of money. If free wage bargaining, as we have known hitherto,
is continued in conditions of full employment, there would be a constant upward pressure upon
money wage-rates ... In peace time the vicious spiral of wages and prices might become
chronic’
340 ALOIS GUGER
Industrial relations shape expectations of the private sector. A co-operative

mode of bargaining between business, labour, and government with con-
sensual outcomes leads to positive business expectations (‘animal spirits’)
and, hence, fosters private investment and growth.
The consequences of free bargaining at high-employment levels under
adverse politico-institutional conditions have been convincingly demon-
strated by Great Britain’s ‘stop-go policy’ in the post-war period and its
harmful effects on business expectations, investment, and growth.
The Keynesian position is best summed up by Kalecki’s conclusion (1943):
‘Full employment capitalism’ will have . . . to develop new social and political
institutions which will reflect the increased power of the working class. If capitalism
can adjust itself to full employment a fundamental reform will have been incorporated
in it. If not, it will show itself an outmoded system which must be scrapped.
If corporatist arrangements can be seen as such new social and political
institutions which facilitate the pursuit of long-run full-employment strat-
egies, some form of corporatism may prove a (necessary) precondition for
maintaining full employment in the long run, at least in societies with a
strong labour movement.
11.3 GROWTH AND LABOUR MARKET PERFORMANCE OF THE

AUSTRIAN ECONOMY SINCE THE EARLY 1970S
Starting from a low level of productivity the post-war development of the

Austrian economy has been a more or less continuous catching-up process
(Table 11.1): in 1970, Austria’s GDP per capita was still about 5 per cent
lower than in OECD Europe on average and 20 per cent lower than in
Germany. In the 1970s and early 1980s, however, Austria’s economy could
improve its relative position significantly. In 1985, Austria’s GDP per capita
was about 10 per cent higher than in OECD Europe and 12 per cent lower
than Germany’s.3
After the war, rapid productivity growth had been the result of the
reconstruction boom and of bridging the technological gap, but soon a
deliberate policy enhanced growth. After breaking the post-war price-wage
spiral in 1953, economic growth received the highest priority in Austria’s
economic policy. Since growth was limited by labour shortages throughout
the 1960s a comprehensive system of investment promotion was introduced
to foster capital accumulation and productivity growth. Furthermore, the
labour market was opened to foreign workers. Their share in the labour
3
According to purchasing power parities (OECD 1989a) in industry, labour productivity
was still 30% below the German level in 1976, but by 1985 this gap had shrunk to 13% (Guger
1988).
TABLE 11.1 Growth of real GDP per capita
Average annual increase (%)
1960–8 1968–73 1973–9 1979–87
Austria 3.6 5.4 3.0 1.6
Germany 3.2 4.0 2.5 1.5
Switzerland 2.7 3.4 –0.1 1.5
Sweden 3.6 3.1 1.5 1.6
Finland 3.3 6.5 2.0 2.9
Norway 3.6 3.3 4.4 2.9
The Netherlands 3.5 3.7 1.9 0.6
Denmark 3.8 3.3 1.6 1.9
OECD Europe 3.6 4.0 1.9 1.3
OECD 3.9 3.6 1.8 1.8
SMECa 3.9 4.2 1.4 1.2
a
Smaller European countries.
force increased from 2 per cent in the late 1960s to 7 per cent in 1973/4
(Fig. 11.1).
In the aftermath of the first oil crisis, Austria responded to the recession
of 1975 with an expansionary strategy and gave full employment highest
priority. In fact, labour market participation and employment increased,
and unemployment rates remained under 2 per cent despite a considerable
labour force growth till 1980 (Table 11.2).
The employment policy of the government was also supported by some
labour hoarding in the nationalized industries in the late 1970s. In fact,
private firms were also rather cautious in their lay-off policies, since un-
employment remained at a very low level and business expectations were
fairly optimistic4 throughout the 1970s. Furthermore, there were supply
side measures on the labour market: a reduction of the legal working time,
in 1975, and a more restrictive foreign labour policy (Fig. 11.1).
After the second oil crisis, the policy response was different. All main
trading partners, and Germany in particular, followed a more deflationary
course than in the 1970s; and Austrian policy makers moved further and
further away from the idea of ‘diving through the recession’.
A number of factors—some hard facts and some ideological or theoretical
reasons—which coincided in the early 1980s led to this change in policy.
In the aftermath of the second oil shock, the effects of the world-wide
4
Austria’s high investment rate may be mentioned as an indicator: gross fixed capital
formation was 26.4% of GDP on average between 1974 and 1979 (OECD 1987). Only Norway,
Japan, Iceland, and Finland show higher figures in the OECD area. Austria’s relative position
was significantly higher in this period than in any period before and after since 1960.
342 ALOIS GUGER
FIG. 11.1 Foreign workers and unemployed in Austria

Source: WIFO.
recession were reinforced by a substantial effective appreciation of the

Schilling. Rapidly rising unemployment and the structural crisis in the
nationalized iron and steel industries put heavy strains on the federal budget
(Fig. 11.2).5 At the same time, an enormous rise in interest rates increased
the costs of financing the budget deficit considerably. In this environment,
the anti-state and anti-activist ideologies of the time found readier accept-
ance than in the 1970s. Furthermore, the increasing share of interest
payments in government expenditures and the deflationary international
environment gave some substance to these objections by weakening the
multiplier effects of expenditures. Finally, when the Socialist Party under
Bruno Kreisky lost its absolute majority in 1983, the coalition governments
that followed moved gradually away from full employment as an absolute
priority.
Hence, in the 1980s, the development of the Austrian economy indicates
a gradual abandonment of the policy of the 1970s. While the hard-currency
approach and moderate incomes policies have been maintained, fiscal policy
5
Employment programmes and claims on the social budget had to be financed; and the net
borrowing of the federal budget in percentage of GDP increased from 2.6% in 1981 to 5.5%
in 1983.
TABLE 11.2 Employment, unemployment, and labour market

participation in Austria
(a)
1960–8 1968– –73 1973–9 1979–87
Growth of
Population (15–64) –0.1 0.3 0.4 0.8
Total labour force –0.7 0.6 0.6 0.3
Total employment –0.6 0.8 0.5 –0.1
(b)
Average (%)
1960–7 1968– –73 1974–9 1980–7
Rates of
Participationa 70.1 67.6 68.9 66.7
Male 90.4 86.4 85.0 80.2
Female 52.1 50.5 53.9 53.6
Employment/
population ratiob 68.7 66.7 67.8 64.3
Foreign labour c 1.1 4.4 5.8 4.6
Unemploymentc 2.0 1.4 1.6 3.5
a
Total labour force as % of population of working age (15–64).
b
Total employment as % of population of working age (15–64).
c
Foreign labour/unemployed in % of total labour force.
has been significantly less expansionary or even contractionary. The

‘consolidation’ of the federal budget has gained priority over full employ-
ment, and the unemployment rate increased from 1.6 per cent in 1980 to
4.9 per cent in 1987.
The rise in unemployment may seem moderate by international stand-
ards, in particular when we take into account seasonal unemployment which
is high and rising with the growing tourist sector. But the unemployment
figures do not reflect the labour market situation quite correctly; on the one
hand, early retirement schemes, special allowances for particular groups of
older workers, and a lower recruitment of foreign labour reduced open un-
employment. On the other hand, in some regions, the structural crisis of the
large iron and steel industries has led to much higher unemployment figures.
As the figures in Table 11.2 indicate, labour market participation has
traditionally been low in Austria. This is mainly due to pension laws and
low female participation rates. The legal retirement age is 60 for women
344 ALOIS GUGER
FIG. 11.2 Employment in nationalized industries in Austria

Source: WIFO.
and 65 for men, but people can and normally do retire earlier if they have
worked for 35 years. Actually, in 1987, at the age of 55–60 only two-thirds
of men and one-quarter of women held a job, and at the age of 60–5 only
16 per cent of men and 7 per cent of women.6 The participation rates of
older people are lower in Austria than in all other European countries but
The Netherlands.
Religion has an important influence on women’s attitude to employment
and family; in Catholic countries women’s participation rates are sig-
nificantly lower than in, for example, Protestant countries.7 Participation
rates of women in the prime age groups8 are also low since a substantial
proportion leave their jobs to take care of children. In this respect, limited
kindergarten facilities, almost no all-day public schools, and few pan-time
employment possibilities have also to be mentioned. Part-time jobs are not
very popular with trade unions in Austria. In particular, female trade union
leaders argue that part-time jobs do not offer equal career prospects.
While in many European countries part-time employment has had im-
portant effects on labour market performance, in Austria the share of
part-time work in total employment grew only from 6.3 per cent in 1973
6 7 8
Walterskirchen(1990:54). Rothschild (1989). Biffl (1988: 42 ff.).
to 7.3 per cent in 1987. Over the same period, the share of part-time work
increased by more than 20 percentage points in The Netherlands and by 7
points (to a level of more than 25 per cent) in Sweden.9
Summing up, the growth and labour market performance of the Austrian
economy were quite remarkable in the 1970s. In the 1980s, unemployment
figures increased substantially until 1987, though they have remained well
below the average European rates. In the favourable international envir-
onment of the last two years, unemployment rates have also been declining
in Austria. So, at the moment, total unemployment is not of primary con-
cern but long-term unemployment has been rising.
The performance of Austria’s economy in the 1970s has often been at-
tributed to its institutions and policy; the question arises, therefore, whether
the deterioration of the 1980s is just caused by the abandonment of the former
strategy or whether the strategy itself was not viable in the longer run?
11.4 THE AUSTRO-KEYNESIAN MODEL: INSTITUTIONS AND POLICY
In the late 1970s, Austria’s economic institutions and policy approach

attracted widespread interest.10 In the aftermath of the first oil crisis, Austria
managed to accomplish low unemployment with moderate inflation and a
remarkable rate of growth. While most industrialized countries gave highest
priority to the fight against inflation, Austria pursued a Keynesian full-
employment strategy, despite inflationary pressures from abroad.
The main characteristics of the Austrian economic strategy, which has ex
post proudly been labelled as ‘Austro-Keynesianism’ (Seidel 1982), are:
• the social partnership as a form of voluntary, informal, and permanent
incomes policy to control wages and prices and international compet-
itiveness;
• an expansionary fiscal policy to foster growth and employment;
• a hard-currency policy to fight inflation; and
• absolute priority of full employment and growth.
This Austro-Keynesian model differs from standard Keynesianism; it is
not confined to anti-cyclical demand management, but it represents a
long-term device to stabilize business expectations with the help of incomes
and exchange rate policies, thus fostering investment and growth; in fact,
Austro-Keynesianism focuses extensively on stabilizing the private sector.11
Austro-Keynesianism is based, on the one hand, on rather fundamental
Keynesian notions such as uncertainty and volatile business expectations,12
9 10
Walterskirchen (1990: 22 ff.). Arndt (1982); Flanagan et al. (1983).
11
In his discussions of the economic plans for the post-war period in the Treasury, Keynes
stressed the importance of this point (Guger and Walterskirchen 1988).
12
Tichy(1984).
346 ALOIS GUGER
and on the other hand, on the corporatist institutions of the Austrian social
partnership.
Furthermore, there is a specific assignment of instruments and goals that
takes into account the openness of the country, its institutional situation,
and the inflationary environment of the time. In the period of high imported
inflation, the traditional assignment did not seem adequate for a small open
economy. While the traditional fiscal policy of compensating for fluctua-
tions in demand was extended to foster growth by investment promotion,
exchange rate policy was assigned to price stability,13 and incomes policy
to external equilibrium by checking the price-wage spiral and preserving
international competitiveness. Thus, moderate incomes policies had to
soften the combined effects of hard-currency and expansionary fiscal policy
on the balance of payments.
As Austria’s institutions have been widely discussed in the international
literature on corporatism and economic policy14 we shall confine ourselves
to a rather short summary and raise some special issues.
11.4.1 The social partnership

15
The social partnership has been widely considered as the centrepiece of
Austria’s post-war economic policy. Bruno Kreisky used to describe it as
‘the sublimation of the basic capitalist class conflict’, meaning a more civil-
ized form of class struggle. Its scope is much wider than an institution of
wage bargaining; it is a system of institutionalized co-operation between
labour, business, and government which is involved in all important aspects
of economic and social policy.
The Austrian social partnership was formed on a voluntary and informal16
basis by the Austrian trade union federation and the Chambers of Agricul-
ture, Commerce, and Labour to control post-war inflation in the early
1950s. Although the immediate reasons mainly concerned the shortcomings
in the public administration of the immediate post-war years, it is deeply
rooted in the history of Austria. Firstly, from an institutional perspective,
there is the Austrian chamber system as the highest level of interest repres-
entation with compulsory membership whose origin dates back to the last
13
Helmut Frisch’s (1976) application of the Scandinavian model to Austria may have acted
as a rationale for this concept; cf. Handler (1982).
14
Katzenstein (1984); Scharpf (1984); Goldthorpe (1984).
15
Suppanz and Robinson (1972); Marin (1982); Gerlich etal. (1985).
16
The voluntary and informal nature of the Austrian approach has been imposed by the
Austrian Constitution. In 1952, the Supreme Court of Austria ruled that a (cabinet) commission
to co-ordinate decisions on price controls, rationing, etc. was unconstitutional since the
authority of a cabinet minister cannot be superseded by collective decisions of a commission or
the cabinet as a whole (Flanagan et al. 1983: 58). A decision that proved wise, as Josef Steindl
pointed out to me, since it pushed back the considerable influence of lawyers in Austrian politics
and strengthened the impact of economic reasoning in Austrian economic policy.
century. The chambers are legal representatives of their members in the
legislative process, i.e. the government has to obtain the approval of the
chambers on draft legislation before it is submitted to parliament. There
are chambers for various professions, but only the Chamber of Commerce,
the Chamber of Labour, and the Chamber of Agriculture are of political
importance from a macro-economic point of view and, hence, are repres-
ented in the Parity Commission. The chambers play an essential part as
intellectual brain trusts for the two large political parties in the general
process of policy and strategy formulation. At the same time, they con-
stitute, together with the Austrian trade union federation (OGB), the
nation-wide, top-level bargaining institutions for wage and price policies.17
Secondly, from the perspective of political history, the tragic experience
of irresolvable class conflicts and civil war in the inter-war period and the
German occupation encouraged a co-operative mode of relationship between
the social classes.18 After World War II, the need for such co-operation was
reinforced by the desire for quick independence from the allied occupational
forces.19 The achievement of political consensus was reflected in a long-
lasting coalition government between the Austrian People’s Party (OVP)
and the Austrian Socialist Party (SPO), both representing the major interest
groups and together representing about 90 per cent of the electorate.
Thirdly, while the trade unions in the First Republic (1918–38) had
been fragmented along craft and political lines, their institutional structure
was changed dramatically in 1945 by the formation of the Austrian trade
union federation. In general, this structural reorganization not only in-
creased the power of the labour movement by centralizing authority,
finance, and bargaining but, in leaving less room to inter-union competition,
placed a higher priority on long-run and nation-wide macro-economic goals
and promoted a co-operative mode of trade union contact with employers.
Today, the trade union federation represents some 60 per cent of all
Austrian employees, and in large industrial companies this percentage is
much higher.
The Parity Commission for Wages and Prices is the most important
formal institution of social partnership. It was founded in 1957 and is
formed by the Chambers of Commerce, Labour, and Agriculture, together
with the Austrian trade union federation (OGB) and the relevant ministers
on a voluntary basis. Although the Federal Chancellor acts as chairman,
members of government have no right to vote. All decisions have to be
unanimous. The Parity Commission works on a voluntary basis and has no
17 18
Biffl etal. (1987). Matzner (1974).
19
At the beginning, the intellectual left was rather sceptical as this co-operation resembled
the ideology of Austro-Fascism and right-wing Catholicism (‘Standestaat’). The chambers,
which in the 1930s had been instruments to expel the socialists and trade unions, were now
supposed to integrate labour.
348 ALOIS GUGER
legal authority nor means of applying direct sanctions. The threat to impose
sanctions is left to the government, but has hardly ever been used.
The Parity Commission is the top-level bargaining institution of incomes
policy in Austria; it was set up to control price and wage trends. Thanks to
its tripartite structure, however, it has also become ‘an instrument which
gives the employers’ and workers’ organizations a voice in government
economic policy in general and conversely enables the Government to make
sure of co-operation from these organizations in its economic policy
measures’.20
The Parity Commission has formed subcommittees on prices and wages
to handle its task of curbing these. The Prices Subcommittee authorizes
price increases which have to be justified by substantial cost increases and
applied for by individual firms or branches. Since the Subcommittee has
no authority to examine firms’ accounts, its effects are limited and more or
less confined to postponements of price increases. In the late 1970s, the
Prices Subcommittee covered about 20 per cent of consumer prices and
about 50 per cent of industrial prices; officially regulated prices and tariffs,
which cover another 15 to 20 per cent of consumer expenditures,21 and
import prices are excluded. Today, the coverage is still lower and by and
large confined to certain basic foods and energy supplies.
The Wages Subcommittee exercises its control on wages by approving
or refusing the opening of wage negotiations. Its task is to combine both
wage-bargaining autonomy at the branch level and the introduction of
macro-economic considerations in the wage formation process. To start
wage negotiations, individual trade unions have to apply through the trade
union federation. Thus, although wage negotiations are conducted by the
subordinate sectoral trade union bodies, the federation has a voice in fixing
the dates and co-ordinating the individual wage claims. Actually, there is a
Working Party on Wage Policy at the trade union headquarters to formulate
common objectives of wage policy. Although individual trade unions are
autonomous in their actual wage negotiations, bargaining processes have to
be authorized and are co-ordinated by the trade union federation and the
Wages Subcommittee.22
In this context, it seems to me that the power of the Austrian trade union
federation vis-d-vis its subordinate sectoral trade unions—at least some of
them—is somewhat overestimated in international studies on wage bargaining
and corporatism. For one thing, the high and growing wage differentials
indicate that one needs to be careful not to put too much emphasis on the
high degree of centralization of bargaining in Austria. By having control
over finances, the trade union federation is potentially very powerful, but
has given much autonomy to the individual unions.
20 21
Suppanz and Robinson (1972: 17). Nowotny (1989: 138).
22
Suppanz and Robinson (1972: 11 ff.).
In 1963, the Parity Commission set up a third permanent Subcommittee,
the Economic and Social Advisory Board, to extend the activities of the
Commission beyond incomes policies and to broaden the scientific basis of
economic policy. The Advisory Board is composed of representatives of the
social partners and ministries as well as of experts of the Austrian Institute
of Economic Research and universities. Its task is to study important
economic and social questions and prepare a scientific basis for policy
recommendations of the social partners to the government. Compared to
similar institutions abroad, the Advisory Board is ‘both a group of expert
advisers and a body for resolving conflicts of interest’.23 It has to be seen,
on the one hand, as an instrument of finding consensus by providing a
common assessment of important features of the economy but, on the other
hand, it has also served to suppress the discussion of crucial issues such as
the problems of environmental protection.
In addition, there are councils to various ministers where the social
partners play a leading role. They are composed of representatives of the
social partners and of the public administration, and experts from research
institutions, and their important task is to advise the minister in special
fields.24
In short, the Austrian social partnership, whose institutions and climate
of consensus grew out of the history of this country, can be characterized
by a rather high degree of centralization, a wide scope of policy involvement,
a co-operative mode of industrial relations, the absence of direct govern-
ment involvement, and by informality.
11.4.2 Incomes policy: wages and wage flexibility

In many countries, governments resort to incomes policies as short-term
manoeuvres to cope with accelerating inflation; incomes policy is normally
confined to wage and price regulation and is backed up by legal force. In
contrast, the Austrian social partnership represents a long-term, voluntary
strategy incorporating virtually all aspects of economic policy. This means
that incomes policy in Austria is embedded in a national economic strategy
and co-ordinated with monetary and fiscal policy.
23
Ibid. 21.
24
Most of these councils have been formed since the late 1960s when the need for a more
scientific basis of policy-making became apparent. As a rule, they have a legal basis like the
Councils for Living Standard Adjustments of Pensions (Pensionsanpassungsgesetz 1965), the
Council for Labour Market Policy (Arbeitsmarktforderungsgesetz 1968) or the Council for
Foreign Labour Policy (Auslanderbeschaftigungsgesetz 1975) which advise the Minister for
Labour and Social Affairs (Butschek 1973; Frank 1975). From a political perspective, as Felix
Butschek pointed out to me, the Council for Labour Market Policy might also be seen as an
instrument of the first single-party (conservative) government (1966–70) after the war to
integrate the trade unions and the Chamber of Labour into the administration of their traditional
domain.
350 ALOIS GUGER
(a) Wage policy

The Austrian trade unions have no explicit distributive goals: their policy
is based on a firm growth ideology. Trade union leaders argue that, since
Austria’s productivity level is still low, wage policy is not an effective redis-
tributive instrument; it would lead to a lower investment rate and reduced
long-term productivity growth and, consequently, to lower real wage in-
creases.
Progressive taxes and high public expenditure are considered as appropri-
ate redistributive instruments.25 Since the 1970s trade union leaders and
the politicians of the labour movement have emphasized uniformly that
maintaining full employment is the most effective distributive policy.
Generally, the wage policy of the trade union federation has been orient-
ated both at compensating for inflation and at long-run productivity growth.
In the period since 1960 both real wages and productivity have on average
been growing by about 3.5% per year. In the 1960s, the unions pursued
an explicit counter-cyclical wage policy, which implied rather moderate
wage claims during business upswings and some pressure during a reces-
sion. Hence, wage policy had overall stabilizing effects.26
In the 1970s, the trade unions aimed at promoting growth and full em-
ployment and formulated as a new guideline the so-called ‘Benya-formula’,
which aimed at a long-run real wage growth of 3 per cent. Since productivity
growth ranged between 4 and 6 per cent, at the time, this implied that firms
were widely provided with funds for investment (Table 11.3).
Since it soon became clear in the 1970s that a unilateral stimulation of
demand would clash with external equilibrium, current account considera-
tions were more and more taken into account in wage claims.27 Actually,
the wage explosion of 1975, which was mainly due to far too optimistic
forecasts, was soon corrected by moderate wage agreements in the following
years, when even trade union leaders spoke of the ‘overdrawn’ wage round
of 1974/5 (Fig. 11.3). Also, the figures in Table 11.3 show that the inter-
national cost competitiveness of manufacturing has improved substantially
since the mid-1970s, although the Austrian iron and steel industries were
in a deep structural crisis in the mid-1980s.
(b) Wage flexibility

Despite the corporatist institutional setting, which has often been con-
sidered as an obstacle to flexible market adjustments, nominal and real wage
flexibility in Austria is relatively high by international standards. Taking
25
In fact, the overall tax system is not progressive in Austria; because of a large share of
indirect taxes it is rather proportional, cf. Guger (1987).
26
Pollan (1984).
27
According to wage equations for the period 1964–84, an increase of the ratio of the current
account deficit to GDP of 1 percentage point dampened the growth of contractual wage rates
by half a percentage point; cf. Biffl et al. (1987).
TABLE 11.3 Labour costs and productivity in industry, Austria

(a)
1964–73 1973–9 1979–88
Total wage costs per hour 10.7 11.0 5.7
Productivity per hour 6.7 5.9 4.2
Unit labour costs 3.7 4.8 1.5
Exchange rate:
Weighted trading partners –0.5 –3.1 –1.5
Germany 1.4 –0.1 –0.4
(b)
Unit labour costs of main trading

partners in. Austrian schillings
1964–73 1973–9 1979–88
Germany 5.8 3.8 2.2
Switzerland 3.5 8.2 2.3
Italy 3.9 3.3 2.9
Sweden 1.8 4.8 –0.2
Weighted trading partners 3.8 4.5 2.1
Austria 3.7 4.8 1.5
Note: (a) - denotes a revaluation.

Source: Austrian Institute of Economic Research (WIFO); OECD, Main
Economic Indicators.
FIG. 11.3 Growth of nominal and real wages per capita in Austria
352 ALOIS GUGER
into account our foregoing analysis, Austria’s corporatist institutions have,

however, to be seen as an important reason for this high wage flexibility.
Wage equations show that Austria is one of the few countries where the
Phillips relation still holds. The unemployment elasticity of nominal wages
comes close to that of Japan and has been one of the highest in the OECD
countries. Since price increases have not been fully compensated and nom-
inal wages react flexibly to changes in unemployment real wage flexibility
has also been high.28
A comparison between the 1970s and 1980s also reveals that wage
flexibility did not decrease in the 1980s when unemployment started to rise
significantly. On the contrary, the unemployment flexibility of wages seems
rather to have risen in the 1980s, since the metal workers’ union, which acts
almost as wage leader, was hit most severely by the crisis in the nationalized
industries.
11.4.3 Demand management

In the 1950s and 1960s the promotion of economic growth had highest
priority. Hence, fiscal and monetary policy started from a long-run growth
perspective. When, in the 1970s, the maintenance of full employment
became the prime objective, their role was no longer confined to mitigating
cyclical business fluctuations but was committed to fostering investment
through an extensive system of tax incentives and direct subsidies (acceler-
ated depreciation, tax credits, low-interest loans).
Since fiscal and monetary policy was co-ordinated with incomes policy
through the institutions of social partnership, an expansionary fiscal strategy
which was accommodated by monetary policy could be pursued despite
inflationary pressures. The inflation of the 1970s was clearly conceived as
cost inflation that could not be fought effectively by monetary instruments.
In Austria, the main demand management instrument is fiscal policy.
Monetary policy was committed to accommodating fiscal policy. In this
context, the Austrian National Bank is, on the one hand, obliged by law
‘to consider the economic policy of the Government’ and is, on the other
hand, also part of the corporatist networks of social partnership through
overlapping memberships. The main objective of the National Bank was to
smooth the waves of financial speculation and to stabilize nominal interest
rates. But since 1976 the hard-currency option has tied the Austrian Schill-
ing to the German Mark, and the room for action on monetary policy has
been seriously limited and to a high degree exogenously determined by the
Bundesbank.
28
The elasticity of wages with respect to prices is somewhere between 0.5 and 0.9 and with
respect to unemployment 1.3; Biffl et al. (1987), appendix, tables 5 and 6.
In Austria, the mere size of the public sector gives great importance to
fiscal policy; in 1983, the share of public expenditures in GDP amounted
to 55 per cent. Furthermore, built-in stabilizers play an important part due
to a well-established system of social security; hence, budget balances are
highly sensitive to output and employment fluctuations. Although these
automatic stabilizers may have been of great importance, there were also
significant discretionary fiscal actions; full-employment budget measures
indicate a clear change to a more expansionary fiscal stance after 1975.29
Austria’s growth advantage in most of the period 1974–82 was related to
its expansionary fiscal policy and could be termed ‘government-led growth’
(Walterskirchen 1990). All in all, the government’s expansionary fiscal
stance and straightforward commitment to full employment and growth gave
business confidence and stabilized the private sector. The policy response
to the second oil crisis was less clear as is aptly illustrated in Table 11.4 by
Hans Seidel (1987).
TABLE 11.4 Demand management and macro-economic performance after

OPEC I and OPEC II in Austria
% increase over 3 years
1974–7 1980–3
Structural budget deficit (% GDP) +3.5 +0.6
M1 in real terms +9.2 +1.9
Real domestic demand +11.2 +0.2
Current account balance (% GDP) –2.4 +2.3
GDP in real terms +8.7 +3.0
Employment +0.6 –2.5
Rate of unemployment (% point change) +0.5 +2.6
Source: Seidel (1987).
The response to OPEC II was still expansionary, but it was rather mod-
erate compared to xht policy in xhe 1970s. The much higher public debt
at the beginning of the second period and the high levels of interest rates
were certainly important causes for this policy change. Since the mid-1980s
the federal government has given high priority to budget consolidation;
there have been cuts in public consumption and partial privatizations of
nationalized industries. The overall impacts on effective demand were
slightly restrictive in the last years. The 1989 budget had, however, mod-
erate expansive effects due to the revenue losses caused by the income tax
reform.
29
Guger (1978); Frisch (1982).
354 ALOIS GUGER
11.4.4 Hard-currency policy

Price stability has traditionally been assigned to monetary policy. Taking
into account the conditions of a small open economy and the origin of
inflation, Austria has assigned price stability to the exchange rate and
external equilibrium to incomes policy since the 1970s.
Since 1976 the Austrian Schilling has been closely tied to the German
Mark; the consequence has been an effective appreciation of the Schilling.
This policy has been widely considered as a precondition for the expansion-
ary demand policy in the face of high inflationary pressures from abroad.
Interestingly enough, this policy was suggested by the trade unions: on
the one hand, because it would slow down the increase in import prices, but
also because it may have made it easier for the trade unions to explain their
moderate wage policy to their members by exchange rate policy reasons.30
In the early 1970s fiscal and monetary policy had been well co-ordinated
in following a rather expansive course. The National Bank aimed at stable
nominal interest rates as well as a stable price level and exchange rate. Then,
since the mid-1970s the Austrian Schilling has been tied to the German
Mark. Yet, mainly due to capital market imperfections, Austrian monetary
policy managed to combine its policy of stable interest rates with a roughly
constant exchange rate to the German Mark till 1979. Then, with growing
capital market integration and rising German interest rates this policy
became unsustainable and the National Bank gave up its interest policy.
The decision to keep to the fixed Schilling-DM rate has since narrowed
the room for manoeuvre in monetary policy, and Austrian monetary policy
has to a high degree been determined by both the Deutsche Bundesbank
and the difference in inflation expectations between Austria and Germany.
Nevertheless, there is no indication that Austrian monetary policy has had
particularly restrictive effects on the real economy.31
11.5 PROBLEMS OF CORPORATISM IN AUSTRIA
In the 1980s, it became more difficult to pursue expansionary policies,

particularly in the restrictive environment of the big neighbour Germany;
so Austro-Keynesianism has lost much of its appeal. Unemployment started
to rise and public debt and interest payments on that debt increased rapidly.
When, in the mid-1980s, the deep crisis in the nationalized industries also
became a big burden for the federal budget there was a rather radical change
in economic policy. Since then budget consolidation has had high priority
in fiscal policy and federal budgets have been rather restrictive32 and have
30 31
Winckler(1988). Handler ; 1989).
32
Although, in 1989, the income tax reform which led to a big loss in tax revenues bolstered
disposable incomes substantially.
no longer been an instrument promoting growth and employment. In fact,
expansionary fiscal policy, a crucial element of the policy of the 1970s, has
been given up.
Although the restrictive international environment and the burden of
public debt may have been the actual reason for these important changes,
the question remains whether endogenous flaws in Austria’s policy of the
1970s may have reduced its long-run viability. Actually, for some time it
has been contended that Austro-Keynesianism suffers from allocative ineffi-
ciencies.33 And Tichy writes: ‘If Austro-Keynesianism should fail in the
longer term, it will probably be because of the fact that industrial policy is
definitely conserving old and outdated structures.’34
11.5.1 Allocative inefficiencies

The Austro-Keynesian policy of the 1970s was essentially a macro-economic
strategy which aimed at growth and full employment. First, there was a
crucial lack of allocative aspects in this strategy. Industrial policy was more
or less confined to indirect investment promotions such as accelerated depreci-
ation, investment allowances, and low-interest credits affecting all industries
across the board and favouring particularly capital-intensive industries.
While this policy fitted in with the macro-economic strategy of fostering
overall growth and a high level of employment, it more or less ignored the
long-run allocative effects and hampered the process of structural adjust-
ment and technological progress.35 Besides, there is evidence that in the
nationalized industries redundant labour was hoarded to support the full-
employment strategy in the late 1970s.36
Second, there is another important aspect of inefficiency in Austrian cor-
poratism. There are numerous regulations, such as the Gewerbeordnung,
protecting substantial parts of goods’ and services’ producers, in particular
the professions, against new competitors.37 Hence, although the wage level
is relatively low by international standards in Austria services are rather
expensive.
Furthermore, in agriculture and food-processing industries we can find
a sort of paradigmatic case of a ‘distributional coalition’ (in the sense of
Mancur Olson) between food-processing industries, farmers, and the food
workers’ trade union, which leads to both relatively high profit margins
and wages in these industries and high food prices for farmers and con-
sumers.38
33
Guger (1978: 19 ff.); Walther (1989); Winckler (1985).
34
Tichy (1984: 382).
35
For a more comprehensive treatment of Austrian industrial policy, see Landesmann in
mis volume.
36 37 38
Nowotny( 1979), cf. also Fig. 11.4. Szopo (1986). Guger (1988).
356 ALOIS GUGER
11.5.2 High wage differentials and slow structural change

Due to an extensive social system and large public expenditure on social
affairs, Austria has often been compared with Sweden as a country of a high
degree of equality. But the structure of primary incomes in Austria differs
substantially from other corporatist economies: wage differentials are very
high in Austria. According to Rowthorn’s (1990) data in this volume,
Austria shows the highest inter-industry wage differentials in Europe.
Among industrialized countries, inter-industrial wage dispersion is only
higher in Japan, the USA, and Canada.
Segmentation in the Austrian labour market is generally relatively high;
also, wage differentials according to social status and qualifications are
much higher than in the Scandinavian countries and higher than in Ger-
many: in the early 1980s, the ratio between industrial earnings of manual
and non-manual workers was 61 per cent in Austria and 69 per cent in
Germany (Table 11.5).39 Earnings data in industry for April 1989 show
also that the ratio of low-skilled blue-collar worker earnings to earnings of
their high-skilled colleagues is 73 per cent in Austria and 81 per cent in
Germany.
The Austrian wage structure has been rather sensitive to market forces.40
In fact, two elements seem to have been of outstanding importance; the
general labour market conditions and the opening of the labour market for
foreign labour. In the last thirty years, inter-industrial wage differentials
have widened substantially as the changes in the coefficients of variation in
Table 11.5 indicate. In the early 1960s, the inter-industrial wage dispersion
of manual and non-manual workers had been quite similar, i.e. less than
15 per cent of average income. Then, in the late 1960s and early 1970s,
when labour demand increased rapidly, the dispersion of blue-collar earn-
ings widened with the influx of cheap foreign labour (Fig. 11.1), while the
structure of salaries remained stable.
From the mid-1970s to the late 1980s, the dispersion of both wages and
salaries increased. Over this period, overall labour market conditions deteri-
orated and general wage policy followed a rather moderate course. Both
factors may have been of importance for the explanation of the tendency
to rising inequality; since low contractual wage increases have left ample
room for wage drift in industries either where demand conditions and
profitability have been better or where the trade union organization has had
a stronger position, as it had in the nationalized industries till the mid-
1980s. Furthermore, in Austria, the concept of ‘solidaristic wage policy’
has more or less been confined to broadly equal increases in contractual
wage rates. The actual development of effective earnings has been left to
branch- or firm-specific negotiations. In fact, till the end of the 1980s, there
39 40
Guger (1989a). Pollan (1980).
TABLE 11.5 Wage differentials in Austrian industry
Inter-industrial wage differentials Monthly earnings Monthly earnings3 Hourly wages3 (blue-
(coefficient of variation) (industrial (blue-collar as %
CORPORATION IN AUSTRIA
collar: lowest skill
average = 100) of white-collar group as % of highest
workers) skill group)
Monthly earnings Hourly Oil Clothing
Employees Blue-collar White- wages industry industry
collar
1961 14.6 14.7 13.9 — 128.7 76.7 57.7 62.8

1971 19.8 20.9 13.8 16.5 150.9 62.2 60.3 63.4
1981 22.8 24.0 16.2 19.0 165.5 58.4 61.1 66.5
1987 24.7 25.1 17.8 21.8 173.4 55.4 59.2 68.6
a
Industrial average.
Source: Guger (1989k 186).
358 ALOIS GUGER
was hardly any attempt to reduce absolute wage differentials. However, in

recent wage rounds there have already been some activities in reducing wage
dispersion within certain industrial wage contracts.
While wide wage differentials and a flexible wage structure have often
been considered as instrumental to labour market flexibility and produc-
tivity growth, the Austrian experience teaches rather the contrary. The large
and widening inter-industrial wage differentials have rather hampered the
process of structural adjustment; first in the 1970s, by keeping labour and
capital too long in marginal production, and second in the mid-1980s, by
impeding the process of reallocating labour from high-wage but money-
losing (nationalized) industries to profitable industries paying lower wages.
In addition, there were also for too long public subsidies41 to ailing indus-
tries which have weakened competitive pressures and slowed down the
process of restructuring.
A more solidaristic wage policy with higher wage increases in the low-wage
sectors would probably have led to higher wage and labour market rigidity,
but certainly also to higher capital mobility and, thus, to faster structural
adjustments in the period of high employment till 1981.
11.6 SUMMARY AND PROSPECTS
Corporatism is an important feature of the political economy of Austria and

her economic performance. Austrian social partnership represents a volun-
tary, long-term device of co-operation between labour, capital, and the
government in virtually all aspects of economic and social policy.
From a macro-economic perspective, corporatism has been rather suc-
cessful in Austria, despite some adversities. In the 1950s and 1960s, the
promotion of economic growth and, in the 1970s up to 1983, full employ-
ment had highest priority in the national economic strategy. Accordingly, an
expansionary policy, in the short run through the budget and in the long run
through an extensive system of tax incentives and subsidized credits, was
complemented by a hard-currency policy, directed towards holding back
financial speculation and stabilizing interest rates. Unions aimed at stabiliza-
tion of employment and prices and not so much at changes in the functional
distribution of income. Since the rapid acceleration of imported inflation in
1973, Austria has tied the Schilling first to a currency basket and since 1976
to the German Mark. The hard-currency policy was supported by the
Austrian trade union federation, which shows the high degree of trade union
co-operation. Also, through the institution of social partnership incomes
41
However, in the 1970s, subsidies to nationalized iron and steel works were low compared
to the subsidies per ton in the EEC. But, in the mid-1980s, the financial help to the nationalized
industries was quite substantial.
policy is embedded in a national economic strategy and co-ordinated with
monetary and fiscal policy.
In the increasingly deflationary environment of the 1980s, when public
debts and interest payments increased rapidly, the fiscal policy changed
from an expansive course to budget consolidation as primary objective.
Mainly due to faster-than-expected international expansion, the consolida-
tion process is well under way, despite a major income tax reform that
implied rather large revenue losses. Also, unemployment has been declining.
From a micro-economic perspective, Austrian corporatism has been less
successful. There has been a lack of allocative efficiency; on the one hand,
there are numerous regulations in the interest of specific sectors and groups
as well as sector-specific distributional coalitions at the expense of con-
sumers. On the other hand, the process of structural adjustment has been
hampered by an industrial policy which favoured large capital-intensive
enterprises rather than small businesses, as well as by high and growing wage
differentials which kept capital and labour too long in marginal production.
In recent years, some major improvements on the supply side of the
economy have been initiated; first, the restructuring of the nationalized
industries is well under way; second, the new tax system favours the tech-
nological improvement of the capital stock and the enlargement of a firm’s
capital base; third, there have also been moderate steps to reduce subsidies
and a process of some deregulation has been going on in the agricultural
and food-production sector quite recently. Finally, the trade unions them-
selves have become aware of growing inequalities and tend to narrow wage
dispersion in most of the new wage contracts.
Austria’s economic performance has been creditable again in the last
years. Corporatism and its institutions, however, have been losing influence
and popularity in the last decade; for one thing, due to the changes in the
employment structure, i.e. shifts from large-scale manufacturing production
to services, and the crisis in nationalized industry organized labour has been
gradually losing some of its authority and influence. For another, the social
partnership has also failed to get rid of the reputation of a ‘production
coalition’ able to take up new problems of general interest, such as environ-
mental protection.
REFERENCES
ABELE:, H., NOWOTNY, E., SCHLEICHER, S., and WINCKLER, G. (1989) (eds.),
Handbuch der osterreichischen Wirtschaftspolitik, 3rd edn., Vienna: Manz.
ARNDT, S. W. (1982) (ed.), The Political Economy of Austria, Washington, DC:
American Enterprise Institute for Public Policy Research.
360 ALOIS GUGER
BIFFL, G. (1988), Arbeitsmarkt 2000, Vienna: Bundesministerium fiir Arbeit und

Soziales.
——, GUGER, A., and POLLAN, W. (1987), ‘The causes of low unemployment in
Austria’, Occasional Papers in Employment Studies No. 7, University of Buck-
ingham.
BLAAS, W., and GUGER, A. (1985), ‘Arbeitsbeziehungen und makroökonomische
Stabilität im internationalen Vergleich’, in Gerlich et ah (1985).
Basil Blackwell.
BUTSCHEK, F. (1973), ‘Arbeitsmarktforschung und Arbeitsmarktpolitik in Öster-
reich’, in H. Seidel and H. Kramer (eds.), Wirtschaftsforschungin Österreich, Vienna:
WIFO.
CAMERON, D. R. (1984), ‘Social Democracy, corporatism, labour quiescence and
the representation of economic interest in advanced capitalist society’, in Gold-
thorpe(1984).
Stabilization and Incomes Policies: European Experience, Studies in Wage–Price
Policy, Washington, DC: Brookings Institution.
FRANK, R. (1975), ‘Träger und Instrumente der Arbeitsmarktpolitik’, in F. Butschek
(ed.), Die ökonomischen Aspekte der Arbeitsmarktpolitik, Vienna: Bohmann Verlag.
FRISCH, H. (1976), ‘Eine Verallgemeinerung des Scandinavischen Modells der
Inflation: mit einer empirischen Analyse für Österreich’, Empirica, 2: 197–218.
—— (1982), ‘Macroeconomic adjustment in small open economies’, in Arndt (1982).
GERLICH, P., GRANDE, E., and MULLER, W. C. (1985) (eds.), Sozialpartnerschaft in
derKrise: Leistungen und Grenzen desNeokorporatismus in Österreich, Vienna: Böhlau.
Clarendon Press.
GUGER, A. (1978), ‘1st die Beschäftigungspolitik am Ende?’, Wirtschaft und Gesell-
schaft, 4: 9–26.
—— (1987) (ed.), Umverteilung durch öffentliche Haushalte in Österreich, Vienna:
WIFO-Gutachten, WIFO.
—— (1988), ‘Produktivität und Lohnkosten in der Industrie: Ein Vergleich zwischen
Österreich und der Bundesrepublik Deutschland’, WIFO-Monatsberichte, 9: 506–
13.
—— (1989a), ‘The distribution of household income in Germany’, WIFO Working
Paper 35.
—— (19896), ‘Einkommensverteilung und Verteilungspolitik in Austria’, in Abele et
al. (1989).
—— and WALTERSKIRCHEN, E. (1988), ‘Fiscal and monetary policy in the Keynes-
Kalecki tradition’, in J. A. Kregel, E. Matzner, and A. Roncaglia (eds.), Barriers to
Full Employment, London: Macmillan.
HANDLER, H. (1982), ‘Die österreichische Hartwährungspolitik’, in H. Abele,
E. Nowotny, S. Schleicher, and G. Winckler (eds.), Handbuch der österreichischen
Wirtschaftspolitik, 1st edn., Vienna: Manz.
—— (1989), Grundlagen der österreichischen Härtwahrungspolitik, Vienna: Manz.

KALECKI, M. (1943), ‘Political aspects of full employment’, Political Quarterly, 14:
322–31.
KATZENSTEIN, P. J. (1984), Corporatism and Change: Austria, Switzerland and the
Politics of Industry, Ithaca, NY: Cornell University Press.
MARIN, B. (1982), Die Paritätische Kommission: Aufgeklärter Technokorporatismus in
Österreich, Vienna: Internazionale Publikationen Gesellschaft m.b.H.
MATZNER, E. (1974), ‘Funktionen der Sozialpartnerschaft’, in H. Fischer (ed.), Das
politische System Österreichs, Vienna: EVA.
NOWOTNY, E. (1979), ‘Verstaatlichte Unternehmen als Instrument der Beschäftig-
ungspolitik in Österreich’, Zeitschrift für öffentliche und gemeinwirtschaftliche Unter-
nehmen, 2: 252–76.
—— (1989), ‘Institutionelle Grundlagen, Akteure und Entscheidungsverhältnisse in
der österreichischen Wirtschaftspolitik’, in Abele et al. (1989).
OECD (1987), Historical Statistics 1960–85, Paris.
—— (1989a), National Accounts, i, Paris.
—— (19896), Historical Statistics 1960–87, Paris.
POLLAN, W. (1980), ‘Wage rigidity and the structure of the Austrian manufacturing
industry: an econometric analysis of relative wages’, Weltwirtschaftliches Archiv,
116: 697–728.
—— (1984), ‘Lohnpolitik und Einkommensverteilung’, in H. Abele, E. Nowotny, S.
Schleicher, and G. Winckler (eds.), Handbuch der österreichischen Wirtschaftspolitik,
2nd edn., Vienna: Manz.
ROBINSON, J. (1943), ‘Planning full employment’, The Times, 22 and 23 Jan.; repr.
in her Collected Economic Papers, i, Oxford: Basil Blackwell (1966).
ROTHSCHILD, K. W. (1989), ‘Some reflections on the growth of female labour supply
and the tertiary sector’, International Review of Applied Economics, 3: 233–42.
SCHARPF, F. W. (1984), ‘Economic and institutional constraints of full-employment
strategies: Sweden, Austria and West Germany 1973–1982’, in Goldthorpe
(1984).
SEIDEL,H. (1982), ‘DerAustro-Keynesianismus’, Wirtschaftspolitische Blatter, 29: 11–15.
—— (1987), ‘Austria’s macro-economic policies in the last decade’, in C. Saunders
(ed.), Management of the Economy and of the Enterprises in East and West, London:
Macmillan.
SUPPANZ, H., and ROBINSON, D. (1972), Prices and Incomes Policy: The Austrian
Experience, Paris: OECD.
SZOPO, P. (1986), ‘Deregulierung in Österreich: Problematik und Ansätze’, Wirt-
schaft und Gesellschaft, 12: 43–61.
TiCHY, G. (1984), ‘Strategy and implementation of employment policy in Austria’,
Kyklos, 37: 363–86.
WALTERSKIRCHEN, E. (1991), Unemployment and labour market flexibility: Austria,
Geneva: ILO
WALTHER, H. (1989), ‘Einige mikro und makroökonomische Aspekte der Preisrege-
lung durch die Paritätische Kommission’, in H. Abele, E. Nowotny, S. Schleicher,
and G. Winckler (eds.) Handbuch der österreichischen Wirtschaftspolitik, 3rd edn.,
Vienna: Manz.
362 ALOIS GUGER
WINCKLER, G. (1985), ‘Sozialpartnerschaft und ökonomische Effizienz’, in Gerlich

et al. (1985).
—— (1988), ‘Härtwahrungspolitik und Sozialpartnerschaft’, Wirtschaftspolitische
Blätter, 35: 58–67.
12
The Swiss Model: Corporatism or Liberal
Capitalism?
Wolfgang Blaas
12.1 INTRODUCTION
Switzerland is often, if not always, considered a special case in terms of its

macro-economic performance. Labour unions and Social Democracy are
weak, wages are high, and inflation is low; monetary policy is restrictive and
fiscal policy rather cautious, but there is full employment. It is widely held
that the case of Switzerland does not fit into any textbook model and is
therefore difficult to understand.
In this country study we will show that Switzerland is neither a special
case nor is its macro-economic performance difficult to explain—if due account
is taken of the country’s economic, institutional, and political endowment.
The chapter is divided into six sections. The next section summarizes
the most important macro-economic indicators in order to show why it is
the conventional wisdom to talk about the ‘Swiss success story’. Leaving
the surface level of long-term averages, a look at the year-to-year figures for
employment and unemployment in section 12.3 reveals a seemingly para-
doxical picture of employment crisis without rising unemployment in two
recessions. Section 12.4 analyses the Swiss labour market and contains the
main arguments explaining this seemingly paradoxical scenario. Out of these
arguments follow some considerations about the limits of the Swiss model
in section 12.5. Section 12.6 summarizes the chapter.
12.2 THE SWISS SUCCESS STORY
This section is deliberately kept short since it is well known that Switzerland
has done extremely well according to most macro-economic indicators. To
summarize the most important facts, long-term average rates of change in
real growth, consumer prices, and employment are shown in Table 12.1
along with long-term figures for unemployment, the current balance, and
currency appreciation. To provide a point of reference, the corresponding
figures for Austria and the OECD countries on average are also given.1
1
Reference is made throughout the chapter to ‘WIFO Data Bank’ as a source for empirical
data. This is the data bank system of the Austrian Institute for Economic Research. Further-
more, since Switzerland and Austria are quite comparable in size (Switzerland has a population
364 WOLFGANG BLAAS
TABLE 12.1 Macro-economic performance: Switzerland in

international comparisona
Switzerland Austria OECD
(%) (%) (%)
Real growth of GDP 2.6 3.6 3.7
Growth of total employment 0.7 0.0 1.0
Labour productivity 1.9 3.6 2.7
Unemployment rate 0.3 2.1 4.9
Inflation15 4.0 4.8 6.1
Current balance0 1.9 –0.4 0.0
Currency (vs. US$) 143.0 70.0 —
a
b
Average rates of change 1960–86.
c
Consumer prices.
Current balance as percentage of GDP, 1960–86 average.
Sources: All figures are taken from OECD (1988b), except for unemployment
figures: Schmidt (1985); WIFO Data Bank.
As the table shows, Swiss unemployment and inflation rates remained at

very low levels throughout the twenty-five years examined, and the current
balance was generally positive, being one reason why the strong Swiss Franc
appreciated by more than 140 per cent against the US Dollar during that
period. Long-term labour productivity growth was significantly below the
OECD average, so that even with a growth rate of real GDP lower than
comparable countries like Austria and lower than the OECD countries on
average, there was long-run employment growth. This increase in employ-
ment, although modest at 0.7 per cent, may have contributed to the ex-
tremely low unemployment levels. The unemployment rates are particularly
interesting, since the developments in them in the 1970s stand in sharp
contrast to the experience in most other Western countries (see Fig. 12.1).
Looking at the overall figures for economic performance, it is certainly justi-
fied to talk about the ‘Swiss success story’ (Danthine and Lambelet 1987).
12.3 EMPLOYMENT CRISIS WITHOUT UNEMPLOYMENT

IN TWO RECESSIONS
Let us consider in more detail what has happened to the Swiss labour
market since the beginning of the 1970s. To begin with, we look at employ-
ment performance. As can be seen from Fig. 12.2, by OECD standards (or
of 6.5 million and Austria a population of 7.5 million) and labour force (Switzerland 3.2 million
and Austria 3.4 million), the Swiss figures are often compared to those for Austrian and/or the
OECD on average.
CORPORATISM IN SWITZERLAND 365
FIG. 12.1 Unemployment rates

Sourer. WIFO Data Bank.
FIG. 12.2 Total employment

Source: WIFO Data Bank.
in relation to neighbouring Austria) developments in employment were

significantly worse in Switzerland during the 1970s. There was a severe
employment crisis in mid-decade; the number of jobs decreased by more
than 10 per cent between 1974 and 1976. The loss of 330,000 jobs during
this period indicates that, in terms of employment, Switzerland was hit
366 WOLFGANG BLAAS
much harder by the world recession of the mid-1970s than other European
countries and the OECD on average (Danthine and Lambelet 1987: 163).
Some two-thirds of these jobs were held by foreign guest workers, and about
50,000 of the remaining 100,000 jobs which vanished in the mid-1970s
recession were held by women (Schmidt 1985: 30). The rest of the unem-
ployed included a number of older workers who retired (before the normal
retirement age) and also some younger persons, now lagging behind in their
entry into the labour market.
The employment crisis had quite significant effects on the supply side of
the Swiss labour market in the 1974–6 recession. During this period 245,000
foreign workers and 60,000 Swiss workers (mainly women) withdrew from
the market, leaving only 25,000 unemployed persons still in the market
and an unemployment rate of no more than 0.8 per cent (see Fig. 12.1). Thus,
despite a deep employment crisis there was almost no (open) unemployment.
During the recovery of 1976–81, the number of new jobs grew rapidly
again. A total of 125,000 new jobs was created, eliminating unemployment
almost completely and providing 100,000 jobs for foreign workers. The
1981–3 recession differed significantly from that in the mid-1970s, because
the loss of jobs was spread much more evenly among Swiss and foreign
workers. During that period, ‘only’ 65,000 employed lost their jobs. Since
20,000 Swiss workers and 20,000 foreign workers withdrew from the labour
market, only 25,000 workers were unemployed, again giving rise to an
unemployment rate of just 0.8 per cent.
Thus, an extremely high elasticity of labour supply was responsible for a
swift adjustment to rapidly changing labour demands, equilibrating the
Swiss labour market rather smoothly. In order to gain a more comprehens-
ive picture of the nature and causes of that unique achievement, we will
take a closer look at Switzerland’s labour market and its institutions.
12.4 THE SWISS LABOUR MARKET
In the following we examine those characteristics of the Swiss labour market

which are both peculiar to the Swiss case and essential for an explanation
of full employment in Switzerland.
12.4.1 Institutional background

(a) Minimalistic state and market dominance
The Swiss economic system is frequently referred to as one of genuine liberal
capitalism. Swiss economic culture is certainly highly liberal, exhibiting a
strong confidence in private business and its ability to manage economic
affairs of all kinds (Breuss 1988: 239). A small public sector hardly interferes
at all in the business sector, which, in turn, is fully aware that it can expect
little in the way of public help in the event of any economic difficulties
(Danthine and Lambelet 1987: 171).
Furthermore, the composition of members in the Swiss parliament is
strikingly biased towards business and business interests as regards the
professions and occupations of its members. This is quite different from,
for example, Austria where the parliament consists mainly of professional
politicians, officials from different interest organizations, and civil servants
(Breuss 1988: 233).
Hence, as might be expected, the state plays but a minor role in the Swiss
economy. In 1986, government expenditure as a percentage of GDP
amounted to 30.4 per cent in Switzerland, as compared to 51.9 per cent in
Austria, 63.5 per cent in Sweden, and 40.2 per cent in the OECD countries
on average (OECD 19886: 64).
(b) Social security

Social security has traditionally been a matter of private initiative. After
World War II, when other European countries enlarged their public welfare
systems, Switzerland retained her system of voluntary unemployment insu-
rance. Thus, when the recession hit Switzerland in the mid-1970s, only 22
per cent of workers were insured against unemployment (Schmidt 1985:
112). Since they were not entitled to receive unemployment benefits, the
majority of workers had no incentive to register as unemployed persons.
The statistical picture of unemployment is therefore definitely biased
downwards. In an empirical study on the relation between registered and
self-declared unemployment in Switzerland, Stolz estimates that this ratio
was 1.4 per cent (!) in 1970. Ten years later, after the introduction of a
statutory unemployment insurance scheme in 1977, the ratio was 22.7 per
cent (Stolz 1985: 403). Registeredunemployment has risen quite significant-
ly since then. A comparison of average unemployment rates before and after
1977 shows an increase by a factor of 6 in Switzerland, compared to a factor
of 1.7 in Austria and a factor of 1.8 in the OECD area on average.
(c) Labour market policy

It is also consistent with the general principle of non-intervention in eco-
nomic affairs that labour market policy and labour market instruments are
rather underdeveloped in Switzerland, at least in comparison to other
European countries.2
There is, however, one important exception. The state interferes heavily
2
In 1987, no more than 0.4% of Swiss GDP was publicly spent on labour market policies,
whereas in Austria the comparable figure was 1.48%, in the UK 2.57%, and in Sweden 2.66%.
Furthermore, income maintenance absorbed more than half of these outlays in Switzerland,
compared to only 30% in Sweden. OECD (1988a: 6).
368 WOLFGANG BLAAS
in the market by closely regulating the (in)flow of foreign workers.3 Until

very recently, labour market policy essentially meant foreign worker policy,
based on the priority principle of protection of native workers (Trioritarer
Schutz der einheimischen Arbeitskrafte’). According to this principle, which
is supported by a large social and political consensus, foreign workers are
the ‘first out’ and the ‘last in’ when employment is changing. Since foreign
worker policy plays a prominent role in Swiss economic policy, we will take
a closer look at foreign workers in Switzerland (see below).
(d) Business dominance in industrial relations

Social peace in Switzerland dates back to the famous Peace Agreement
(Friedensabkommen) in the 1930s, the origin of a distinctive ‘ideology of
consensus’, as all comparisons of strike figures quite clearly indicate: during
the period 1965–80, there were, on average, only 0.8 minutes of strike per
worker in Switzerland, compared to 6.5 minutes in Austria, 17.4 minutes
in the Federal Republic of Germany, 96.6 minutes in Belgium, and 226.8
minutes in Finland (Blaas 1984: 166).
In contrast to Austria, for example, the Swiss business community is
centralized and strong whereas labour is rather weak. Between 1960 and
1985, about 30 per cent of the labour force was unionized in Switzerland.4
However, there is no peak organization of labour. The Swiss trade union
confederation SGB (Schweizerischer Gewerkschaftsbund) comprises about
fifteen unions, organized by industries or groups of different jobs. But, in
addition to the SGB, there are also confessional unions, such as the CNG
(Christlichnationaler Gewerkschaftsbund), and white-collar unions, such as
the VSA (Vereinigung der Angestelltenverbande).5 The national employers’
organization (Zentralverband der schweizerischen ArbeitgeberOrganisauonen)
embraces thirty-six local and partial organizations and the member com-
panies employ more than one million people (1976). Hence, the degree of
organization is higher among employers than among labour.
The centralization of capital does not mean, however, that labour-capital bar-
gaining takes place at the national level. In Switzerland bargaining is a matter
for the individual union or institutions at company or even plant level. The
central organizations focus their effort on contribution in political institutions
(parliament). Business and labour deal with each other on questions of wages
and employment and the state steps in only in the event of a serious crisis.6
Investment issues are dealt with between industry and the financial sector.
3
Apart from a protectionist agricultural policy, Swiss foreign worker policy seems to be the
only area where Switzerland does not adhere to the principle of non-intervention.
4
Schweizerische Bankgesellschaft (1987: 53).
5
In 1985, 51% of unionized labour was organized in the SGB, 12% in the CNG, and 17%
in the VSA. Schweizerische Bankgesellschaft (1987: 53).
6
In this respect, Switzerland constitutes a striking exception among the small European
countries (Katzenstein 1985: 91).
To sum up, business is in a strong position to control both employment
and investment decisions and this is accepted by labour as an intrinsic
feature of the capitalist system, whose prevalence is not questioned. As a
consequence of comparatively weak labour organizations, workers do not
enjoy such far-reaching protection against dismissal as is usual in other
European countries, so that hiring and firing costs are relatively low.
Hence, changes in the business cycle have much more immediate effects
on employment.
Furthermore, the powerful and therefore dominant business interests
make for an explicitly conservative attitude towards foreign workers and the
role of women in society at large. This attitude clearly supported and
favoured the withdrawal from the labour market by foreigners and women
in the 1974–6 recession.7
The Swiss economic system is sometimes classified as a ‘liberal corporat-
ism’.8 If corporatism is defined as a pluridimensional concept,9 embracing
centralization of government, employers, and unions, public involvement
in relations between the government and the two classes, and class co-
operation, then it seems more appropriate to think of the Swiss system as
liberal capitalism. In the Swiss model, centralization of unions is absent and
centralization of government is only weak; furthermore, public involvement
is essentially lacking in matters of class conflicts. The Swiss model more
closely resembles paternalistic-liberal capitalism of the Japanese type and
may not therefore be considered a corporatist economy as such.
12.4.2 Labour demand

Although there were cautious employment programmes in 1975/6 and
1982/3, the expectations of managers about fiscal behaviour in the reces-
sion have certainly been formed by the declared abstinence from interven-
tion.10 It is explicitly stated and fully in accordance with the principles of
non-intervention and market dominance that the state has no responsibility
to compensate for deficient effective demand during a downswing in the
business cycle.11
7
Cf. section 12.4.3(d).
8
For Katzenstein, Switzerland is a liberal variant of democratic corporatism. Katzenstein
(1985: 105).
9
See R. Archer in this volume.
10
The employment programme of 1975/6 is estimated to have created or maintained 8,600
jobs in 1975, 24,900 in 1976, 25,000 in 1977, and 8,400 in 1978. This implies a reduction in
the potential unemployment rate of 0.3 percentage points in 1975, 0.8 percentage points in
1976 and 1977, and 0.3 percentage points in 1978. If one also takes into account cumulative
employment effects, the potential unemployment rate was reduced by a maximum of 2.2
percentage points (Schmidt 1985: 93–4).
11
‘Es kann und darf niemals Aufgabe des Staates sein, mit einem konjunkturellen
Stutzungsprogramm die fehlende Nachfrage voll zu ersetzen.’ Bundesrat (1976: 17).
370 WOLFGANG BLAAS
It comes as no surprise then that real investment in machinery and

equipment declined strongly in 1975 (–17.7 per cent) and 1976 (–13.9 per
cent), while the European OECD economies on average experienced a
decline in real investment of 6.5 per cent in 1975 and a rise of 5.4 per cent
in 1976. Furthermore, private consumption is bound to be a more procyc-
lical demand element in a system without statutory unemployment in-
surance12 and with (comparatively) low social security benefits, so that
automatic stabilizers can yield only minor compensating effects. In addition,
Switzerland’s extremely hard currency has not been without heavy costs for
export business. In 1975, the volume of exports and services decreased by
6.6 per cent as compared to only 3.7 per cent in OECD Europe (Austria:
–2.4 per cent). In the recession at the beginning of the 1980s, exports
declined by 3 per cent in 1982, whereas exports rose in the European OECD
countries on average by 1.2 per cent (Austria: +1.9 per cent).
Consequently, industrial production fell dramatically in the 1974–6
recession and subsequently recovered very sluggishly. From 1974 to 1976,
the index of total industrial production decreased by 12 per cent, approach-
ing the level of 1974 again in 1980 and 1981. This is in stark contrast to
what happened, for example, in Austria, where industrial production shrank
by 6 per cent from 1974 to 1975, and recovered immediately in 1976 to
the level of 1974. From 1977 onwards the index of industrial production
was higher than in 1974 in every succeeding year. By contrast, it was not
until 1985 that Swiss industrial production exceeded the 1974 level (WIFO
Data Bank).
12.4.3 Labour supply
(a) Foreign workers

Switzerland, like most other central European and Nordic countries, suf-
fered from labour shortage in the long prosperous period after World War
II and therefore, like these countries, attracted workers from other—mainly
southern European13—countries. There is, however, one outstanding fea-
ture of the Swiss foreign labour force: its order of magnitude compared to
the native labour force. As early as 1965, when the share of foreigners in
the labour force was 6.5 per cent in Belgium, 5.7 per cent in Germany, 0.8
per cent in Denmark, 1.8 per cent in The Netherlands, 1.6 per cent in
Austria, and 4.6 per cent in Sweden, the Swiss ratio was 23.7 per cent, i.e.
almost one quarter of jobs were held by foreign workers.14 Again, as in most
12
It has already been mentioned that unemployment insurance was not made compulsory
until 1977.
13
For the composition of countries of origin of Swiss foreign labour see Biffl (1985: 485).
14
Ibid. 481.
comparable countries, the share of foreign workers reached its peak during
the early 1970s (26 per cent in Switzerland, as compared to e.g. 11 per
cent in Germany, 9 per cent in Austria), and has been declining ever since.
Hence, the very number of foreign workers made foreign worker policy
more important and also more powerful than in any other comparable
country.
On the other hand, increasing uneasiness among the native population
(‘overforeignization’) led to growing pressure on politicians to impose a
ceiling on the total number of foreign workers. Such a ceiling was eventu-
ally introduced (in the 1970s) at 25 per cent of the labour force, which is
still regarded as politically and socially acceptable. It is important to note
that while the mid-1970s recession provided economic reasons for reducing
the foreign labour force, there were also good political and social reasons,
at least from the viewpoint of Swiss politicians, for repatriating foreigners.
(b) The structure of foreign labour

To comprehend the options of Swiss foreign worker policy, it is important
to know that there are four different groups of foreign labour:15 first,
workers with a permanent resident status; second, border commuters,
travelling daily between their jobs and their homes; third, annual workers,
who have to renew their work permits every year; and fourth, seasonal
workers, having the right to work up to nine months out of twelve. Fig.
12.3 shows the distribution of these four categories in the period from 1970
to 1988.
At die beginning of die 1970s, the major part of the foreign labour force
(70 per cent) consisted of annual and seasonal workers. Therefore, the
inflow of foreign workers could be managed quite efficiently by controlling
work permits for diose having to reapply on a yearly or seasonal basis.
Indeed this was what happened in the 1974–6 recession. As Fig. 12.3 shows,
more than 240,000 foreign workers lost dieir jobs in the recession, and the
main burden was borne by annual and seasonal workers, whereas the
number of border commuters decreased only slightly and die workers with
resident status increased more or less steadily.
The figure also shows quite clearly that die structure of foreign labour
has changed dramatically since the early 1970s. Today, foreign workers
with resident status and border commuters take up the large and dominant
share of the foreign labour force which annual and seasonal workers used
to have in die early 1970s. Consequently, die options of foreign worker
policy today must be quite different from what tiiey were then. We will
come back to this point in section 12.5.
15
Recently, a fifth group of workers with short-term work permits (‘Kurzaufenthalter’) has
been introduced.
372 WOLFGANG BLAAS
FIG. 12.3 Foreign labour in Switzerland

Source: Die Volkswirtschaft (Nov. issues).
(c) ‘Natural’ turnover rate of foreign labour

Foreign workers may have reasons other than economic ones for not re-
applying for work permits or for leaving the host country. The wish to return
to their family and origins or the discrimination experienced in the host
country may also be responsible for the decision to leave the labour market
voluntarily.
It is estimated that, prior to the 1974–6 recession, only two-thirds of the
seasonal workers in Switzerland used to reapply for a contract in the
following year. Furthermore, on average about 50,000 annual workers and
workers with permanent resident status left Switzerland every year between
1970 and 1974 (Danthine and Lambelet 1987: 163). This large ‘natural’
turnover rate in the 1974–6 recession added to the then much welcomed
downward supply flexibility of the labour market.
(d) Participation rates

We have already mentioned that institutional factors in general and the
underdeveloped social security system may have contributed to the vanish-
ing of the unemployed from the labour market. This is confirmed quite
convincingly by labour participation figures.
The total labour force participation rate decreased from 77.4 per cent
in 1974 to 73.6 per cent in 1976, i.e. by almost 4 percentage points. In the
European OECD countries on average, the participation rate fell by less
than half a percentage point, whereas in countries like Austria, Denmark,
and Finland, the participation rate actually rose (OECD 1988b: 34).
Further, if despite slow growth and a bad employment performance after

the 1974–6 recession (see Fig. 12.2) there was no significant rise in unem-
ployment in Switzerland, a declining participation rate was mainly respons-
ible for full employment.
Admittedly, the labour participation rate also declined in the European
OECD countries between 1974 and 1986 from 67.4 per cent to 65.4 per
cent on average. However, if Switzerland had shared in the European trend
(i.e. a decline in the labour participation rate of 2 percentage points), there
would have been approximately 188,000 additional unemployed persons in
1986 resulting in an unemployment rate of 6.6 per cent (0.8 per cent
registered unemployment plus additional 5.8 per cent unemployment).
In particular, female participation in the labour market stagnated from
1974 to 1986, whereas it rose by 3 percentage points on average in OECD
Europe, and by almost 14 percentage points in Sweden (OECD 19886:
35). Male participation fell in Switzerland as in OECD Europe, but much
faster: the rate dropped by almost 11 percentage points (1974–86) com-
pared to a fall of only 7 percentage points in OECD Europe.
12.4.4 The explanation summarized

The main thrust of our argument has been that rigorous management of
labour supply has been predominantly responsible for low unemployment.
This is in line with the fact that Switzerland does not fare particularly well
in a cross-country comparison in terms of real wage flexibility, although the
latter has been somewhat higher than in comparable countries.16 Also in
terms of wage dispersion Switzerland holds an average position among
OECD countries, but shows definitely more wage equality than for instance
Austria.17 Labour demand has been shown to be rather volatile since
business expectations are not stabilized by Keynesian-type fiscal and/or
monetary policy and automatic stabilizers are relatively small.
12.5 LIMITS TO THE SWISS MODEL
The above account of the Swiss labour market and its recent developments
reveals a fair number of problems. These problems may constitute some
limits to the Swiss model, and we examine each of them in turn.
First, an extremely high labour supply elasticity, which was quite essential
for maintaining full employment in the mid-1970s recession, may have been
16
Real wage flexibility, as measured by the coefficient of variation in real wage changes, is
slightly higher in Switzerland than e.g. in Austria. Sources: Schweizerische Bankgesellschaft
(1987); WIFO Data Bank.
17
See Rowthorn in this volume (Table 4.2 and Fig. 4.3).
374 WOLFGANG BLAAS
counter-productive to the Swiss economy in the past. Comparatively low

costs of hiring and firing contribute to low overall costs of labour, reducing
the incentive to restructure the economy and slowing down the pace of
innovation.
The failure of Switzerland, strongly criticized by the OECD,18 to restruc-
ture its industry by developing new products or shifting into new activities
with long-term market potential was an important reason for the stagnation
of industrial output and employment up to the mid-1980s. The predomin-
antly private Swiss business community preferred to restructure by cutting
back and rationalizing old industries.
Whereas the Swiss type of labour relations with weak unions but an
implicit job guarantee for native workers and foreign workers with per-
manent resident status would have allowed for innovation and restructuring
without much friction, there was not enough cost pressure to pursue that
particular kind of adjustment.19
Second, labour market flexibility has, on the other hand, been significantly
reduced by structural changes in the foreign labour force in the last twenty
years. A large share of foreign workers today holds the status of permanent
residents, having essentially the same privileges as native Swiss workers.
Therefore, if there was a recession today, one might expect that foreign
labour supply would react in a far less elastic way than it did in the
mid-1970s. This might also be true in the near future for other social groups,
e.g. female workers. In a more modern society, their participation rate may
not be as easy to reduce as it was in the past.
What would be the consequences of greatly reduced labour market flex-
ibility for the unemployment figures? Suppose that overall labour participa-
tion had declined in line with trends in OECD Europe between 1974 and
1986 and, in addition, that the number of foreign workers had remained at
its 1973 level. Then Switzerland would have suffered from mass unemploy-
ment of 9.8 per cent in 1986.20
Third, on the demand side there seem to be clear institutional limits to
any significant expansionary fiscal impulse. According to the principle of a
minimalistic central state there is a widely shared distaste for Keynesian
intervention to foster effective demand. Although there were small and
limited employment programmes in 1975/6 and 1982/3, it seems obvious
18
OECD (1984). See also Rowthorn and Glyn (1990).
19
At the time of writing (1989), the pressure to innovate and rationalize is coming
increasingly from labour shortages, as the overall ceiling on the maximum number of foreign
workers has been reached.
20
Schmidt estimates the rate of unemployment for 1976 to be between 4.2 and 7.9%, if the
stock of foreign labour is kept at 1973 levels (Schmidt 1985:63–4). Rowthorn and Glyn estimate
the unemployment rate in 1983 to be 15% under the assumption that the number of foreign
workers is kept constant from 1973 onwards and furthermore the overall participation rate is
kept at its 1973 level (Rowthorn and Glyn 1990).
by the mere size of the public sector21 that only minor effects on the business
cycle are to be expected.
Furthermore, and probably even more important, monetary and exchange
rate policy quite obviously reflect the interests of the financial community.
Contrary to the otherwise cherished principle of non-intervention in market
affairs, monetary authorities pursue an extreme hard-currency policy in
favour of rentiers and banks at the cost of industrial production and employ-
ment. The economic power hierarchy, characterized by superiority of finan-
cial business over industrial business, and by superiority of business over
labour, puts the burden of (cyclical) adjustment mainly on to labour. In the
past, this did not set any severe limits to the system, since those who carried
the costs were predominantly without much political weight. However, this
may change in the near future, as we have tried to demonstrate.
12.6 CONCLUSIONS
In section 12.2 we started out from the ‘Swiss success story’, by looking at
the well-known and impressive macro-economic indicators. When we went
behind these figures to try to gain a deeper understanding of this success
story, the picture changed gradually and in a number of different ways. In
the end, we found that the ‘Swiss success story’ not only looks much more
like the performance of other highly developed nations, but also displays
some remarkable failures.
The outstanding achievement of the Swiss economy is full employment.
However, as we have shown, it is more appropriate to talk about full
employment for (male) Swiss citizens. There is no full employment with a
constant labour force.
Under Swiss conditions, high labour market flexibility is the counterpart
of capital immobility. Slow growth of labour productivity and a low innova-
tion rate follow as negative consequences, though mitigated by high rates
of research and development expenditures of Swiss multinational com-
panies.
The Swiss model of social consensus about market dominance, non-
intervention, and class co-operation assigns the burden of economic dis-
equilibria quite clearly to the lowest level in the power hierarchy: foreign
workers (without permanent resident status) and women are the residual
entity. But Switzerland could perform much better if these power rigidities
could be broken up, and labour, industrial, and financial interests were to
co-operate as real partners.
21
Current government disbursements as a percentage of GDP were 29% (on average
1974–9) and 30% (1980–6) in Switzerland as compared to 40% (1974–9) and 45% (1980–6)
in OECD Europe. OECD (1988b: 63).
376 WOLFGANG BLAAS
REFERENCES
BIFFL, G. (1985), ‘Die Entwicklung der Ausländerbeschäftigung in den wichtigsten

europäischen Industriestaaten’, WIFO-Monatsberichte, 58: 480–93.
BLAAS, W. (1984), ‘Stabilisierungspolitik: Zur politischen Ökonomie markt-
wirtschaftlicher Instabilität’, Der öffentliche Sektor: Forschungsmemoranden 10/1–2:
1–238.
BREUSS, F. (1988), ‘Die wirtschaftlichen Beziehungen zwischen Österreich und der
Schweiz: zwei SMOPECS als Konkurrenten und Partner’, in Thalberg (1988).
Bundesrat (1976), Botschaft des Bundesrates über weitere Arbeitsbeschaffungsma.fi-
nahmen, Berne.
DANTHINE, J.-P., and LAMBELET, J.-CH. (1987), ‘The Swiss recipe: conservative
policies ain’t enough’, Economic Policy, 5: 149–79.
OECD (1984), Economic Survey of Switzerland, Paris.
—— (1988a), Employment Outlook, Paris.
—— (19886), Historical Statistics 1960–1986, Paris.
ROWTHORN, B., and GLYN, A. (1990), ‘The diversity of unemployment experience
SCHMIDT, M. G. (1985), Der schweizerische Wegzur Vollbeschäftigung: eine Bilanz der
Beschäftigung, der Arbeitslosigkeit und der Arbeitsmarktpolitik, Frankfurt: Campus
Verlag.
Schweizerische Bankgesellschaft (1987) (ed.), Die Schweizer Wirtschaft 1946–1986:
Daten, Fakten, Analysen, Zürich.
STOLZ, P. (1985), ‘Empirische Untersuchungen über das Verhaltnis von registrierter
zu selbstdeklarierter Arbeitslosigkeit in der Schweiz’, Schweizerische Zeitschrift für
Volkswirtschaft und Statistik, 121: 391–406.
THALBERG, H. (1988) (ed.), Österreich—Schweiz: Nachbam, Konkurrenten, Partner,
Vienna: Braumüller-Vlg.
13
The Unexpected Emergence of Australian
Corporatism
Robin Archer
13.1 INTRODUCTION
Throughout the 1980s corporatism has been in retreat. The trend has been
the same in one country after another. In Britain a weak legacy of corporat-
ism was defeated; in Sweden a strong legacy was weakened. Indeed amongst
OECD countries the trend has been almost universal. Almost, but not quite.
For just when the rest of the world was moving away from corporatism,
Australia was firmly embracing it.
By ‘corporatism’ I mean to refer to that phenomenon which has been
variously labelled ‘neo-corporatism’, ‘liberal corporatism’, ‘societal corpor-
atism’, and ‘social corporatism’. The same phenomenon has at least as many
definitions. Why this is so only becomes clear once we recognize that corpor-
atism is essentially a system of industrial relations.
The industrial relations systems of the advanced capitalist countries all
have three main actors: the government, the employers, and the unions. Re-
lations both within and between these actors can be organized in different
ways. And the different combinations of these intra- and inter-actor relations
define different industrial relations systems. Fig. 13.1 illustrates the six di-
mensions which combine to define an industrial relations system. Three of
these dimensions concern the degree of centralization of each of the actors.
The fourth and fifth concern the degree of public involvement in relations be-
tween the government and the two class actors. Public involvement concerns
both class involvement in public affairs and public (i.e. government) involve-
ment in class affairs. The sixth dimension concerns the degree of class co-
operation. Corporatism is an industrial relations system which combines a
high degree of all of these: centralization, public involvement, and class co-
operation.1
Because an industrial relations system can be made or broken by the way
1
Each of the better-known definitions of corporatism deals with one or more of these factors.
Schmitter (1981: 284) equates corporatism with centralization, Maier (1984: 40) emphasizes
class co-operation, Offe (1981: 136) defines it as public involvement, as does Maier (1981: 56)
in an earlier essay, Regini (1984) combines public involvement and centralization, and Crouch
(1979: 187) refers to centralization and consensus which he relates to both class co-operation
and public involvement. But only Lehmbruch (1979: 5; 1984: 61) recognizes that in fact
corporatism is a ‘pluridimensional concept’ which combines them all.
378 ROBIN ARCHER
FIG. 13.1 The six dimensions of an industrial relations system

it deals with wage demands, centralized wage fixation has often been seen as
the most important manifestation of a corporatist system. However unions
do have other bargaining chips. In Australia, for example, attention has
recently shifted from wages to working practices and the organization of
work. And these can play an important role in sustaining a corporatist
system. So while centralized wage-fixing systems are likely to be corporatist,
not all corporatist systems must feature centralized wage fixation.
Since the first election of the Hawke government in 1983, the Australian
labour movement has purposefully gone about building up a corporatist
system. The system which has been established is based on the Statement of
Accord by the Australian Labor Party and the Australian Council of Trade Unions
Regarding Economic Policy (ALP/ACTU 1983). The ‘Accord’ as it came to
be known was indeed centred on an incomes policy. However by now its
original prescriptions have been altered so many times that the government
and the unions refer instead to the ‘Accord process’. The bipartite origins of
this process have ensured that a high degree of public involvement and a
highly centralized union movement have become the most distinctive fea-
tures of Australian corporatism. Both formal and informal negotiations
between the government and the ACTU play a fundamental role in setting
macro-economic policy. However this process has in turn brought about
greater class co-operation, leading, for example, to a radical reduction in
strike activity. More recently, explicitly negotiated class co-operation be-
tween unions and employers has also begun to emerge.
In the early years of the Labor government many commentators dismissed
the emergence of corporatism in Australia as an aberration which would
soon fall into line with international trends. But when the Labor Party was
re-elected for an unprecedented third term in July 1987 it became clear that
the Australian labour movement had genuinely bucked the international
trend and established a lasting corporatism in the 1980s.
THE EMERGENCE OF AUSTRALIAN CORPORATISM 379
This chapter is divided into four more sections. Each is designed to answer
a specific question about the evolution of Australian corporatism. But each
is also designed to shed light on one of the more general problems which our
(i.e. WIDER’s) study of corporatism has raised. Section 13.2 looks at how
the Accord came about, and, in particular, at whether or not it is a product
of uniquely Australian traditions. It aims to address the general question of
whether different societies are free to accept or reject a corporatist system,
or whether they are bound by tradition to their respective industrial relations
systems. Section 13.3 considers whether the Accord achieved the goals that
it originally set itself. It adds this evidence to the general debate about the
effect of corporatist institutions on macro-economic performance in a period
of stagflation. Section 13.4 considers whether the Accord can be successfully
adapted to achieve a new set of goals. It questions the generally held belief
that corporatism is incompatible with the demands of micro-economic
restructuring and labour market flexibility. Finally, section 13.5 looks at the
Accord’s future prospects. It considers whether the conditions underpinning
the Accord are strengthening or weakening and speculates about the ends to
which a continuing Australian corporatism could be put.
13.2 CHOOSING CORPORATISM
Is it possible for a society to choose its industrial relations system? And in

particular is it possible for a country without a tradition of corporatism to
opt for a corporatist system? For a study which is based on a comparison of
the experiences of different countries these are fundamental questions. If we
answer ‘no’ to them, then our work can be of no more than scholarly interest.
From the point of view of the policy maker, there would be little point in
undertaking a comparison of different industrial relations systems if the basic
characteristics of each are pre-given and unchangeable.
For this reason there is an implicit assumption in our project that ad-
vanced capitalist countries can learn from each other. But is this assumption
tenable? The work of some industrial sociologists suggests that it is not.
Colin Crouch, for example, argues that, in Europe, corporatism can only be
successfully established in countries which have a tradition of sharing polit-
ical space, and that these traditions were established ‘some time in the
eighteenth or nineteenth centuries’ (1986: 177). Drawing on the work of
Charles Maier (1984: 29), he highlights the importance of the ‘liberal
parenthesis’ that existed between the era of the guilds and the standestaat,
and the era of organized capitalism and organized class interests. During this
parenthesis parliamentary states attempted to monopolize political space.2
2
The liberal parenthesis is better known as the era of laissez-faire capitalism. As Crouch
points out state attempts to monopolize political authority were directly linked to the defence
380 ROBIN ARCHER
In some countries the parenthesis lasted longer than in others. And in some
countries the political monopoly that inaugurated it proved more difficult to
establish than in others.3 According to Crouch these two factors established
a tradition which continues to determine whether a country has a liberal or
a corporatist industrial relations system. His central thesis is that ‘the longer
the interval, or the sharper the breach, between the destruction of ancient
guild and standestaat institutions and the construction of typically “modern”
interest organisations, the more committed did the state become to liberal
modes of interest representation, and the less likely to tolerate sharing po-
litical space’ (1986: 182). So whether a country is prepared to share political
space, and hence whether it is prepared to countenance corporatism, is a
function, not of contemporary choices, but of long-established traditions. I
will refer to this type of claim as the ‘tradition-boundness thesis’.
Now I have been dwelling on this point because, prima facie at least,
Australia seems to provide a decisive counter-example. Of course it is not a
counter-example of Crouch’s particular thesis since in the history of white
Australia there is no pre-parenthesis period. But it does seem to provide a
counter-example to the general line of argument that corporatism can only
arise as a function of long-established tradition. For Australia’s long-
established industrial relations tradition is generally thought of as a variant
of the British liberal tradition. This is reflected in almost every attempt to
rank countries according to their degree of corporatism. Australia is typically
grouped near the bottom of the list with the other Anglo-Saxon countries—
Britain, the United States, Canada, and New Zealand.4 But if this is so, the
emergence of corporatism in the 1980s can hardly be a product of tradition.
The liberal industrial relations tradition of the Anglo-Saxon world is widely
seen as the antithesis of a corporatist tradition.
In response it might be said that white Australia does not provide a
counter-example to the tradition-boundness thesis because it is simply so
young that it has no traditions and hence should be ruled out as an excep-
tional case. We should be wary of this argument. As with most advanced
capitalist countries, Australia has a tradition of organized industrial relations
dating back well over a century. The tradition is a product of both its British
of a laissez-faire market. To defend a laissez-faire market the state must not only keep itself out
of the economic system; it must also monopolize political space to keep economic actors out of
the political system, for this would also have the effect of politicizing the market.
3
The difficulty was especially great in those countries where the Catholic Church had a
long-established claim to political space.
4
See e.g. Cameron (1984: 165), Crouch (1985: 119), Bruno and Sachs (1985: 226), and
Schott (1984: 2). Schmitter (1981: 294) does not include Australia but as Crouch (1985: 118)
points out Schmitter’s criteria would place it ‘at or near the bottom of his ranking’. The one
exception is Calmfors and Driffill (1988: 17–20) who give Australia an intermediate ranking.
This is partly because of their attempt to take into account the developments of the early years
of the Accord.
inheritance and a reaction against it. If the tradition-boundness thesis is true,

the emergence of Australian corporatism must be a product of this tradition.
To see if Australia is really a counter-example we will have to identify the
elements of this tradition and assess whether they helped or hindered the rise
of corporatism.
Two ideologies dominated the first century of white settlement in
Australia. The first was a version of English radical liberalism. The second
was the distinctively Australian ethos of ‘mateship’.
The Australian colonies were among the world’s first democracies. Be-
tween 1850 and 1858 the main points of the constitutional programme of
English radicalism were put into effect. Manhood suffrage and vote by ballot
were adopted a generation before the third Reform Act brought them to
England. Economically the radicals focused on the question of land distribu-
tion. By promoting the development of a class of small independent farmer-
settlers they hoped to provide the new democracy with a solid foundation. This
aspiration was shared by many miners who began to look for alternative ways
of safeguarding their independence once the gold rush of the early 1850s
began to wane (Gollan 1960: 15–16). The radicals won the legal battle with
the passage of the Land Acts of 1860–1 in the colonies of New South Wales
and Victoria. But the squattocracy (named both for the way in which they
originally acquired their vast land holdings and for their aristocratic pre-
tensions) retained overall control of the pastoral industry, and it became
increasingly clear that wage-labour rather than self-labour would be the
norm. So lucrative was the pastoral industry, however, that in the early
twentieth century Australia had the highest standard of living in the world.
Alongside these radical liberal initiatives a different ideology was emerging
amongst the bush workers of the outback: the ethos of mateship. Mateship
grew up amongst the semi-nomadic drovers, shepherds, shearers, bullock-
drivers, stockmen, boundary-riders, and station-hands who were collectively
known as bushmen. These men did not deny the value of individual inde-
pendence stressed by the radical liberals. But they placed a special emphasis
on egalitarian values and on the importance of solidarity or fellowship. Mates
were equals, and mateship was a fellowship of equals. These values were
rooted both in the hardship and loneliness of life in the outback and in the
convict origins of the earliest bushmen (Ward 1965: 2, 10, 27, 78). Although
it seems likely that ‘the main features of the new tradition were already fixed
before [the beginning of the goldrush in] 1851’ (Ward 1965: 11), it was in
the 1880s that they came to have enormous national significance when a
movement of writers and poets began to promote mateship and the values
of the bush as definitive of a new Australian national identity.5
5
One of the most forceful and influential advocates of this new identity, the Bulletin, defined
it like this: ‘By the term Australian we mean not those who have been merely born in Australia.
All white men who come to these shores—with a clean record—and who leave behind them the
382 ROBIN ARCHER
The point to note is that neither radical liberalism nor mateship is straight-
forwardly propitious for the development of corporatism. The first is clearly
hostile, placing a premium on individual independence; while the second is
more ambiguous, valuing both a collective and an individual identity.6
But ideologies are not the only manifestations of traditions. Traditions
are also carried by long-established institutions. To get a clearer picture of
Australia’s industrial relations tradition we need to look at each of the
institutions that has played a significant role in establishing it. To assess the
tradition-boundness thesis we will have to consider whether these institu-
tions have helped or hindered the emergence of corporatism. I will first
consider the institutions of the labour movement (the unions and the Labor
Party), then those of the employers, and finally those of the state.
Embryonic workers’ organizations began to appear in the 1830s, but it was
only in the 1850s, after gold brought new prosperity to the colonies, that
stable unions were formed. These were unions of skilled workers based on
craft divisions and they set a pattern of craft unionism which continues to
exist. They were joined by new unions of unskilled and semi-skilled miners,
shearers, and waterside workers in the 1880s. These unions drew strength
from the mateship ethos of the bush. According to the first president of the
shearers’ union, W. G. Spence:
Unionism came to the Australian bushman as a religion. It came bringing salvation
from years of tyranny. It had in it that feeling of mateship which he understood
already, and which always characterised the action of one ‘white man’ to another.
Unionism extended the idea, so a man’s character was gauged by whether he stood
true to Union rules or ‘scabbed’ it on his fellows. (Crowley 1980: 78)
The new unions faced their first trial of strength during the depression of
the early 1890s. Under the banner of ‘freedom of contract’ employers,
supported by police and troopers, took on the seaman, the shearers, and the
miners. In each case the workers were decisively defeated.
These defeats encouraged unions to consider the benefits of direct
involvement in parliamentary politics. From 1890 onwards the Trades and
memory of the class distinctions and the religious differences of the old world; all men who place
the happiness, the prosperity, the advancement of their adopted country before the interests of
Imperialism, are Australian. In this regard all men who leave the tyrant-ridden lands of Europe
for freedom of speech and the right of personal liberty are Australians before they set foot on
the ship which brings them hither. Those who fly from an odious military conscription; those
who leave their fatherland because they cannot swallow the worm-eaten lie of the divine right
of kings to murder peasants, are Australians by instinct—Australians and Republicans are
synonymous. No nigger, no Chinaman, no Kanaka, no purveyor of cheap coloured labour is an
Australian.’ Bulletin, 2 July 1987 (Clark 1955). Australians combined a vituperative racism with
an image of themselves as egalitarian, prosperous, and free.
6
Bushmen exhibited a ‘ “manly independence” whose obverse side was a levelling, egalit-
arian collectivism, and whose sum was comprised in the concept of mateship’ (Ward 1965: 167).
Labour Councils that co-ordinated union activity in the various colonies

began to form parties to contest seats in their respective parliaments. By
1884 the New South Wales Labor Party held the balance of power. And in
1899 the world’s first labour government was formed in Queensland al-
though it was out of power by the end of the week. However there was little
to distinguish Labor parliamentarians from other progressives. The repres-
entatives of labour were deeply committed to liberalism and parliamentary
government, and only a minority saw socialist or collectivist ideas as distin-
guishing them from other liberals (Gollan 1975: 206). Those socialist ideas
that did exist were heavily influenced by Edward Bellamy’s justly forgotten
Looking Backward which mingled them with a populist belief in a semi-
mystical entity called ‘the money power’ which conspired against ‘the
people’. In general, however, there was little attention to doctrine of any kind.
Anticipating federation in 1901, the colonial Labor parties met the year
before to establish the Australian Labor Party (ALP). Its objectives were
established at its first full-scale national conference in 1905. By 1910
majority Labor governments were in power both federally and in New South
Wales.
Neither the unions nor the Labor Party were developing in a way that was
propitious for corporatism.
I will begin with the unions. The one factor which is almost universally
agreed upon as a precondition for corporatism is a strong centralized union
movement. Indeed a number of writers see the centralization of organized
labour as coextensive with corporatism.7 Centralization has three dimensions:
membership density, organizational unity, and concentration of power.8
These are illustrated in Fig. 13.2. Only on the first of these dimensions do
Australian unions score well prior to the 1980s.
Membership density refers to the extent to which unions have succeeded
in organizing workers. In the first decades of this century, Australia became
the most unionized country in the world. By 1920 over 53 per cent of all
employees belonged to a union. During the depression this dropped to a low
point of 42.6 per cent in 1934 and then steadily rose to an all-time high of
63 per cent in 1953. During the 1960s and 1970s membership density
settled down to around the 55 per cent mark.
Organizational unity refers to the extent to which unions are themselves
organized into peak councils. The lowest scores go to countries with a mul-
tiplicity of peak councils and numerous unions. The craft origins of Austra-
lian unionism led to the formation of a plethora of unions each jealously
protective of its craft base. There have been well over 300 unions for most
7
This is especially true of the economists who have recently attempted to quantify cor-
poratism. See e.g. Calmfors and Driffill (1988).
8
Cf. Cameron’s (1984: 164–6) ‘organisational power of labour’, and Schmitter’s (1981:
294) ‘societal corporatism’.
384 ROBIN ARCHER
FIG. 13.2 Union centralization

Note: In Australia unions are members of both peak federations and state Trades and
Labour Councils (TLCs).
of this century. In 1983 the figure still stood at 319.9 After a number of
unsuccessful attempts many of these unions joined to form the Australian
Council of Trade Unions in 1927. But its coverage was far from universal.
The Australian Workers’ Union, built on the base of the old shearers’ union,
claimed the status of an alternative peak council and refused to join until
1967. Moreover white-collar and public sector unions were organized into
two separate peak councils. These merged with the ACTU in 1979 and 1981
respectively. The formation of a single peak council has certainly helped the
development of corporatism. But with over 300 unions and 3 or 4 peak
councils until the beginning of the 1980s, organizational unity can hardly be
considered part of the industrial relations tradition.
Concentration of power refers to the extent to which a peak council is able
to control the industrial activities of its affiliates. In Australia intrastate
industrial disputes are managed either by the union concerned or by the
state’s Trades and Labour Council. Although the Trades and Labour
Councils are branches of the ACTU, they retain significant independent
powers reflecting their early growth and importance during the colonial
period. Formally, the ACTU has the power to take over the management of
interstate industrial disputes. But in reality, unless its assistance is sought,
disputes remain outside its control. It is the affiliates that control strike funds
and, on the whole, it is the affiliates that initiate and manage disputes.
What about the Labor Party? There are three factors concerning socialist
9
Of these the largest 14 unions (each with more than 50,000 members) represented 44.6%
of unionists while the smallest 105 (each with less than 500 members) represented just 0.6C/3
(Hancock Committee 1985: 23).
parties which have been thought to help the development of corporatism.

None of them seems to apply in Australia.
The first concerns the original relationship between the party and the
unions. Corporatism has prospered where parties and unions developed at
the same pace. But it has proved difficult to establish either where parties
reached maturity before unionism became firmly established, as in France
and Italy, or where unionism preceded party formation, as in Britain (Maier
1984: 54–5). We have seen that it is this latter, British-style relationship that
pertains in Australia. The unions established the Labor Party to put their
delegates in parliament and they considered their influence over the party a
matter of right rather than one of negotiation.
Second, corporatism has been most sustained in those countries in which
socialist parties have played a dominant role in government (Schmitter
1981: 313–14; Panitch 1977). Clearly, ‘unions will be more prepared to
enter into political bargaining, and will have greater confidence in eventual
gains from it, where they possess close ideological as well as organisational
ties with the ruling party’ (Maier 1984: 49). However in Australia, despite
its early success, the Labor Party did not become a natural party of govern-
ment, as in Sweden, or even a natural partner of government, as in Austria.
The 1949 defeat of the Labor government which had led the country through
the Second World War inaugurated twenty-three years of continuous Con-
servative rule. In 1972 Labor regained government for just three years before
the Conservatives were returned to office where they remained until 1983.
Third, it seems that it is not just the predominance of socialist parties that
matters; it is the predominance of socialist parties of a certain type. In
particular, corporatism seems to be best fostered by German-style Social
Democratic parties with a Marxist-influenced past (Goldthorpe 19846:
327). Having resigned themselves to operating within a capitalist economy,
the Marxist heritage of these parties fostered class-wide corporate identites
and a more political unionism. We have seen that the Australian Labor Party
was built on essentially liberal ideological foundations. The speed with which
the newly formed Labor parties moved into a position of power forced them
to turn their attention immediately to the exigencies of running a capitalist
economy and left little room for the development of radical political ideas.
Marxist ideas, in particular, were thin on the ground. According to a con-
temporary commentator, Labor leaders ‘know little or nothing of Marxian
theories. Few of them know even by tide the principle text books of contin-
ental socialism.’10 It is true that in 1921 the ALP adopted ‘the socialisation
of industry, production, distribution, and exchange’ as its objective, but, by
then, ‘socialism without doctrine’ was already well entrenched.11
1`0
Clark (1906). Cited in Pons (1984: 118).
11
In this regard it is interesting to note the role of Communists in providing the legitimizing
framework for the Accord within the union movement. Australia’s most prominent Communist
386 ROBIN ARCHER
In summary, neither of the two main institutions of the Australian labour

movement carried traditions which were propitious to the development of
corporatism. The unions did have a high membership but were otherwise
weakly centralized. They were established well before the Labor Party, and
the party itself failed to predominate. Marxist ideas had very little influence.
In each of these respects the Australian labour movement looked very similar
to its British counterpart, and a similar industrial relations tradition could
be expected.
Employer organizations have also been unfavourable to the development
of corporatism. The employers have even less of a tradition of strong
centralized organization than the unions. Their early organizations were
either trade associations or were formed for specific purposes such as
combatting the 1890 strikes. When national organizations began to emerge
after federation their role was restricted to making legal challenges to the
emerging industrial relations system. Ultimately two stable organizations
took root: one co-ordinated employers’ federations, the other co-ordinated
chambers of commerce. Employers remained deeply divided between the
two until the 1960s. A merger was finally achieved when the Confederation
of Australian Industry was formed in 1977. But the unity was short lived. In
1983 a second national employer organization—the Business Council of
Australia—was formed. And in the latter half of the 1980s the CAFs claim
to be a national peak council was further undermined as one after another
of its main industry groups chose to operate independently.
Finally I want to look at the institutions of the state. Australia’s origins as
a penal colony bequeathed it centralized and interventionist administrative
institutions. Following self-government these institutions were turned to
entrepreneurial tasks in order to promote rapid colonial development. Large-
scale assisted immigration programmes were launched to import labour
from Britain (de Lepervanche 1975), and the colonial governments raised
huge capital loans on the London Bond Market which they invested in
railways, roads, and bridges (H. V. Evatt Research Centre 1988: 17). So
widespread and so widely supported was this public economic intervention
that it is often referred to as a kind of ‘colonial socialism’. After federation
in 1901 an interventionist approach was also adopted by the federal govern-
ment. It was most clearly articulated in the ‘New Protection’ policy which
restricted immigration to whites, and imposed tariffs and bounties to protect
local manufacturers, and some agricultural producers, from overseas com-
petition. However in the realm of industrial relations it was much more
difficult for the federal government to act.
union official, Laurie Carmichael, repeatedly emphasized that the Accord was a class policy
designed to protect the interests of all workers, not just the sectional interests of those who
happened to have industrial muscle.
The new Australian Constitution specified the powers of the Common-

wealth (or federal government) and left all other powers with the states
(which took over from the old colonial governments). In the area of in-
dustrial relations the Commonwealth’s powers are severely restricted. Sec-
tion 51 (xxxv) of the Constitution empowers it: ‘to make laws for the peace,
order and good government of the Commonwealth with respect to . . .
Conciliation and Arbitration for the prevention and settlement of industrial
disputes extending beyond the limits of any one State’. The power to
legislate about incomes, working conditions, and all other industrial rela-
tions matters rests with the states.12 With such limited powers the federal
government does not seem well equipped to foster corporatism. In fact,
however, this peculiar constitutional situation led to the establishment of a
unique conciliation and arbitration system which some have seen as a form
of proto-corporatism. Anyone wanting to claim that the Accord is a product
of Australian traditions would emphasize its continuity with the arbitration
system. Since the arbitration system provides defenders of the tradition-
boundness thesis with their strongest argument, I want to consider it in more
detail.
The outcome of the industrial struggles of the 1890s convinced the
labour movement that some sort of state machinery was necessary to enforce
union recognition and fair wages and conditions. In 1904, with Labor Party
backing, the federal government used its limited constitutional power to
establish a Court of Conciliation and Arbitration which was later renamed
the Conciliation and Arbitration Commission. When a dispute occurs, the
Commission is first required to encourage conciliation between the parties.
If the parties manage to reach an agreement, the Commission will usually
register it as a ‘consent award’. But if the parties fail to agree arbitration is
compulsory and the Commission must convene a hearing and make an
award itself. Either way the outcome is an award which is legally binding on
both parties. Penalties can be imposed on those who fail to comply with an
award, but in reality these are only used against employers.13
By 1912 similar industrial tribunals had been established in all the states.
Gradually the federal tribunal became predominant. In the 1907 Harvester
case, the founder of the federal arbitration system, Justice Higgins, estab-
lished a needs-based criterion for setting the basic (or minimum) wage that
should be paid to an unskilled worker. Skilled workers would be awarded
‘margins’ over and above the basic wage. Arguing that an average man must
support himself, his wife, and three children, Higgins set the basic wage at
42 shillings a week. Since this was above the prevailing minimum rate being
12
Wartime, and the Commonwealth’s own employee, are exceptions to this.
13
Attempts to fine striking unionists in the 1960s led to die jailing of the tram workers’ leader
Clarrie O’Shea. The unions responded with a national strike and refused to pay all further fines.
Since then penalties for direct action by unionists have not been imposed.
388 ROBIN ARCHER
offered by the state tribunals, many unions sought to be covered by a federal

award. Ultimately almost the entire work-force was covered by either federal
or state awards.14
There is no doubt that the arbitration system ‘is deeply entrenched in
[Australian] society’. These are the words of the Hancock Committee
(1985: ii. 4) which was established in 1983 by the Hawke government to
conduct the first comprehensive review of the industrial relations system
since its inception. The Hancock Committee essentially recommended a
continuation of the existing system, and public opinion seems to support this
recommendation.15
Now there are a number of objections to the claim that the Arbitration
Commission has fostered the development of corporatism. Firstly, it can be
argued that, both by taking on the mantle of the public interest, and by
protecting the existence of a plethora of small craft-based unions, the Com-
mission has hindered the development of an encompassing strategically
thinking union movement. Why should unions consider the public interest
when the Commission is there to mediate between it and their claims?
Secondly, some economists have suggested that the Commission has no
long-term effect at all. This is a controversial point but those who support it
argue that Commission decisions simply mimic the changing market
strength of employers and unions.16 Thirdly, it is often emphasized that it is
a mistake to juxtapose the arbitration system to free collective bargaining
since one of the main aims of the system’s founders was to achieve recogni-
tion of collective bargaining. In particular ‘it is wrong to equate “conciliation
and arbitration” and centralisation and to equate “collective bargaining” and
decentralisation ... it is possible to have highly centralised collective
bargaining systems and highly decentralised arbitral systems’ (Hancock
Committee 1985: ii. 212).
This third objection is especially telling given the pivotal role of centralized
bargaining in a corporatist system. Consider wage bargaining for example.
A cursory look at the historical record shows that the arbitration system has
coexisted with a wide variety of wage-fixing systems. Table 13.1 summarizes
the history of Australian wage fixation. It is clear that early on the Arbitration
Commission developed into a unique centralized forum which based its
14
The first official survey of award coverage in 1954 showed that 89.5% of the work-force
were covered by awards: 42.5% were covered by federal awards while the remaining 47% were
shared out between the six states.
15
The Hancock Committee commissioned a poll in 1984 which showed that 56% of people
thought that the Arbitration Commission did a good job at national wage fixing while 16%
thought it did a poor job (Hancock Committee 1985: iii. 212). In another poll that year, people
were asked whether the Commission should continue in its major role of determining wages
and working conditions: 76% answered yes; 14% answered no; and 10% didn’t know (Hancock
Committee 1985: ii. 227).
16
The Hancock Committee rejects this view (1985: ii. 156 and n. 10). For a recent review
of some of the literature see Gill (1987a).
TABLE 13.1 Key events in national wage fixation in Australia, 1904–1983

1904 Conciliation and Arbitratioon Act passed.
1907 Harvester judgement based on cost of living (COL).

Basic wage Margins
1912 Higgins applies ‘needs 1912 Increases based on ‘work value’

principle’ through COL or ‘comparative wage justice’
adjustments. In 1922 annual (employees doing same work in
changes replaced by automatic different firms should get same
quarterly adjustments (AQA). pay). Margins set
independently of the basic
wage.
1931 Emergency 10% reduction.
1934 The economy’s ‘capacity to 1935 Metal increases flow-on.

pay’ principle is affirmed as the
appropriate basis for increases
but AQA continue.
1939 War loadings give most

unskilled workers a margin.
1947 Commission fails to stop

another metals flow-on.
1953 AQA ended. Replaced in 1956 1954 Flow-ons formalized into metal
by annual reviews. From 1961 trades ‘test cases’ which set a
awards based on a mix of formula for all other margins.
national productivity and COL.
Total wage rule
1967 Separate hearings replaced by a single national wage case. But unions
rely on market strength to get increases outside of national wage cases.
1974 Reintroduction of quarterly wage indexation (6–monthly after 1978).

Partial indexation and increases outside the system lead to its collapse.
1981 New round of metals-led free collective bargaining.
1983 Wage pause followed by the Accord.

390 ROBIN ARCHER
awards on national criteria such as the needs principle, the economy’s

capacity to pay, and the doctrine of comparative wage justice. But the
important point to note is that the Commission later showed itself to be quite
compatible with a highly decentralized system closely resembling British-
style free collective bargaining. In particular in the late 1960s and early
1970s the Commission’s national wage decisions had become little more
than a guaranteed minimum for weakly organized workers and the most
important wage rises were achieved in individual award changes negotiated
outside the arbitration system. Evidence for this can be seen in the sharp rise
in consent awards (which are simply ratified by the Commission) and
overaward payments (which are paid completely independently of the Com-
mission), and in increased strike activity.17 By the financial year 1974/5 the
contribution of national wage cases to increases in average male award
wages had fallen to just 21.2 per cent (Howard 1977: 271). After a period
of centralized indexation there was a further period of decentralized free
collective bargaining in the early 1980s.
This is very damaging for the tradition-boundness thesis. In its strongest
form, the thesis claims that certain tradition-carrying ideologies or institu-
tions exist which in themselves are a sufficient condition for the development
of corporatism. We have seen that amongst the ideologies there are no
candidates for this role, and that amongst labour, employer, and state
institutions, only the Arbitration Commission is a viable candidate. But at
times, and indeed immediately prior to the Accord, the Arbitration Com-
mission has coexisted with liberal wage-fixing arrangements, so it can hardly
be a sufficient condition for corporatism. However, finally to lay the tradi-
tion-boundness thesis to rest we also have to tackle the weaker claim that the
Commission was a necessary condition for corporatism. To test this weaker
claim we need to identify what were the main conditions which led to the
emergence of the Accord. If the Commission is not among them then the
claim is false.
A comprehensive analysis of the conditions which led to the emergence of
the Accord would be a major task. But briefly we can say that the Accord
came about because the labour movement adopted it as a policy at a time
when it had the power to implement it.
The policy developed out of new ideas which began to emerge during the
late 1970s and early 1980s in both the political and industrial wings of the
labour movement. A National Committee of Inquiry began a systematic
reappraisal of Labor Party policies in 1978, and in the same year the metal
workers’ union began publishing a series of booklets on alternative ways of
dealing with Australia’s growing economic crisis. In particular they
17
On consent awards see Gill (19876: 8). On overaward payments and strike activity see
Hancock Committee (1985: ii. 193 ff. and 54).
identified the social wage and industry development as industrial issues that
required unions to negotiate not only with employers but also with govern-
ment. It was these changing attitudes within the unions—especially powerful
left-wing unions like the metal workers—that were crucial.
A peculiar combination of boom and bust ensured both that these unions
committed themselves to the Accord, and that they had a powerful voice in
setting the terms of the agreement. The ‘boom’ was a phoney resources
boom in 1980–1 which strengthened the negotiating position of the unions
who, led by the metal workers, won huge wage rises in the market. This also
strengthened the unions’ hand in negotiating the terms of the Accord with
the Labor Party. Labor urgently required a coherent alternative to the
Liberal government’s monetarist-inspired ‘fight inflation first’ strategy if it
was to avoid a fourth successive defeat at the impending elections. And this
required an agreement with the unions on wage restraint. The left-wing
unions used their bargaining strength to widen the Accord to include social
wage issues and an interventionist industry development policy. The ‘bust’
was the 1982 recession which soon made it clear that the unions’ underlying
strength was being seriously eroded. Massive lay-offs and soaring inflation
reinforced the unions’ commitment to the Accord. In the twelve months to
May 1983, employment in the metal and engineering industries fell by 17
percent.18
The Accord was an agreement between Labor and the unions. The
Arbitration Commission played no role in formulating it. Of course, because
of the constitutional division of power, the newly elected Labor government
had to turn to the Commission to implement the Accord’s income policy.
But this hardly suggests that the Commission was a necessary condition. On
the contrary if it were not for Australia’s peculiar constitutional division of
powers, the government could have simply implemented its policy directly.
In a sense the Commission was an impediment. It was an independent party
outside the original Accord agreement which had to be convinced that it was
worthy of approbation. This is not to suggest that there were no advantages
to operating through the Commission. It did, for example, provide a ready-
made forum for national wage fixation, and it released the unions and the
government from some of the responsibility for policing the system. But this
is a long way from being tradition-bound. Australia’s newly emerged cor-
poratism is a product, not of tradition, but of the choices made by the labour
movement in response to the circumstances of the 1980s. The prima-facie
18
The crucial importance of the recession in cementing the Accord supports Maier’s
contention (1984: 50) that it is their ‘sense of potential economic vulnerability’ that drives small
highly developed countries towards corporatism, and, more generally, highlights Katzenstein’s
thesis (1985) that small open economies tend to adopt corporatism because of their extreme
vulnerability to external shocks. During the early years of the Accord, Bob Hawke liked to
dramatize the crisis confronting Australia by comparing his task with that of the wartime Labor
Prime Minister John Curtain.
392 ROBIN ARCHER
observation was right: Australia does provide a counter-example to the

tradition-boundness thesis.
13.3 DEALING WITH STAGFLATION
Recently a number of economists have taken an interest in the effects of

corporatist industrial relations systems on macro-economic performance. In
particular, attention has focused on claims that corporatist countries have
dealt better with stagflation than countries with other industrial relations
systems. For these economists corporatism is largely equated with central-
ized wage-bargaining structures and macro-economic performance is typic-
ally measured as a function of unemployment.19 However the empirical
evidence for this claim is inconclusive. Indeed an OECD study (1988: 35)
finds that: ‘On a country-by-country basis there appears to be no direct
relationship between macro-economic performance and bargaining struc-
ture in either the 1970s or 1980s.’ Matti Pohjola draws a similar conclusion
on the basis of his survey of both empirical and theoretical studies. Pohjola
suggests that ‘more could be learnt about corporatism by examining the
experiences of individual countries—alternating corporatist and non-
corporatist episodes—instead of making direct cross country comparisons’
(Chapter 3 in this volume).
Australia has experienced a number of more or less centralized wage-
fixing ‘episodes’ since the onset of stagflation in 1972. We have already seen
that decentralized free collective bargaining was the order of the day in the
early 1970s. In 1975 this was replaced by centralized quarterly wage indexa-
tion which gradually broke down and was followed in 1981 by a renewed
bout of decentralized bargaining. To complete the picture we need to
consider the wage-fixing arrangements that have prevailed since 1983 under
the auspices of the Accord.
The Accord was explicitly designed to deal with stagflation. It identified
‘simultaneously high levels of unemployment and inflation’ as ‘the funda-
mental feature’ of the crisis afflicting Australia in the early 1980s (ALP/ACTU
1983: 1). The Accord rejected the monetarist solutions adopted by the
Liberal government and noted that the OECD countries which have dealt
best with stagflation are those that ‘have eschewed monetarism and have
instead placed substantial importance on developing prices and incomes
policies by consultation’ (ALP/ACTU 1983: 2). An incomes policy is vital
according to the Accord because, by restraining wage increases, it will allow
the government to pursue expansionary policies designed to promote
economic and employment growth, while retaining control of inflation.
19
See Bruno and Sachs (1985), Rowthorn and Glyn (1990), and Calmfors and Driffill
(1988).
Without wage restraint, ‘economic recovery will soon lead to increased

inflation, thus forcing the government to adopt contractionary anti-inflation
policies which will truncate the recovery and prevent any restoration of full,
or near full employment’ (ALP/ACTU 1983: 1).
Thus wage restraint is a fundamental prerequisite for ‘the parties’ prime
objective of full employment’ (ALP/ACTU 1983: 1). The Accord attempts
to combine this with a second key objective, namely ‘to ensure that the living
standards of wage and salary earners and non-income earning sectors of the
population requiring protection are maintained and through time increased
with movements in national productivity’ (ALP/ACTU 1983:4). To achieve
both these objectives—employment growth through wage restraint and the
maintenance of living standards—the Accord advocated a return to a central-
ized wage-fixing system based on full cost of living adjustments. Increases
could also be sought on the basis of increased national productivity, and
reduction to the emerging 38–hour-week norm could continue, but other-
wise unions must make no extra claims. The Accord partners also reached
agreement on price surveillance, taxation, and a number of ‘supportive
policies’ covering industry development and the social wage (Archer 1988:
216–18). Here however I will concentrate on the incomes policy commit-
ments which are at the heart of the Accord.
The final years of the Fraser Liberal government were marked by escalat-
ing social conflict. This was forcefully brought home to the public in 1982
when a demonstration by recently retrenched steel workers forced their way
past police and broke down the doors of Parliament House. During the 1983
election campaign Labor leader Hawke contrasted this conflict with his call
for consensus and national reconciliation. Having won the election, Hawke
immediately convened a National Economic Summit of state and federal
governments, unions, and business leaders, and won from it endorsement
for a return to centralized wage fixing. Against this background the ACTU,
supported by the government, applied to the Arbitration Commission for the
establishment of a centralized wage-fixing system based on the provisions of
the Accord. The Commission agreed to a system of twice-yearly hearings at
which full cost of living increases would be granted unless extraordinary
circumstances persuaded it to the contrary, and so long as unions gave a
commitment to make no extra claims.
This system remained in place until September 1985, when, in an agree-
ment with the government known as the ‘Accord Mark IF, the unions agreed
to trade off wage rises due to cost of living and productivity increases for tax
cuts and the introduction of employer-funded occupational superannuation.
But a much bigger change was on its way.
In May 1986 the Treasurer, Paul Keating, publicly suggested that
Australia was threatening to turn into a ‘banana republic’.20 The Treasurer’s
20
Interview broadcast on Sydney radio station 2GB, 14 May 1986.
394 ROBIN ARCHER
comment inaugurated a huge shift in public perceptions and conveniently

marks the beginning of a new period in which the government pursued
policies that were fundamentally at odds with the original Accord agree-
ment. Stagflation was replaced by the balance-of-payments deficit as the
main focus of economic strategy.
Figs. 13.3, 13.4, and 13.5 show the different components of the balance-
of-payments crisis confronting Australia by mid-1986. The terms of trade
were deteriorating rapidly as key export commodity prices fell by about 10
per cent (Fig. 13.3). In a country where primary produce accounts for approx-
imately 80 per cent of export earnings this had a dramatic effect. Foreign
indebtedness rose sharply (Fig. 13.4). So did import penetration (Treasury
1988a: 38–41). All this contributed to a large and worsening current account
deficit (Fig. 13.5).
In June 1986 Prime Minister Hawke responded to the balance-of-pay-
ments crisis with a National Economic Statement that called for cuts in real
wages and deep cuts in public spending. The contractionary budget that
followed made it clear that even the Accord’s basic commitment to reduce
unemployment had been dropped. Hence both the key objectives of the
original Accord—reduced unemployment and the maintenance of living
standards—were no longer being pursued by the government. This was a
crucial juncture for the Accord. The ACTU could easily have returned with
righteous indignation to the old-style politics of opposition. Instead it chose
FIG. 13.3 Terms of trade, Australia

Note: Terms of trade as measured by the ratio of the implicit price deflator for exports of
goods and services to the implicit price deflator for imports of goods and services.
FIG. 13.4 Net external debt, Australia

Note: Net external debt is defined as total gross debt less official reserve assets and
Australian lending overseas.
FIG. 13.5 Balance on current account, Australia
to push ahead with the ‘Accord process’. Two days after the 1986 budget,
when it became clear that the old wage-fixing system would not survive, the
ACTU began formulating a workable alternative: the Accord Mark III. The
ACTU proposed a new ‘two-tier’ system which was accepted by the Arbitra-
tion Commission in March 1987. Under the first tier the Commission
awarded a flat-rate $10 increase to all workers. Under the second tier
workers with industrial muscle could win increases up to a maximum of a
further 4 per cent by making enterprise-level productivity agreements. This
partial decentralization was reinforced by the August 1988 national wage
case which awarded increases that were conditional on unions agreeing to a
396 ROBIN ARCHER
‘structural efficiency’ principle which foreshadows industry- and enterprise-

level negotiations with employers aimed at radically revising and modern-
izing all awards. This was widely viewed as a decisive shift away from
centralized wage fixation.
In summary then, there have been five wage-fixing episodes since the
onset of stagflation. These are shown in Table 13.2. So has the Accord
improved Australia’s macro-economic performance? Figs. 13.6 and 13.7
and Tables 13.3 and 13.4 show movements in some of the main indicators.
A cursory examination seems to support the public perception that the
Accord brought some significant improvements. Clearly, in judging whether
the Accord was successful in tackling stagflation, its first three years are of
special significance. For it was only during this period that stagflation was
the focus of the government’s policy and that the Accord’s stagflation-
orientated incomes policy was being pursued.
TABLE 13.2 Wage-fixing systems under stagflation in Australia

1972–4 Decentralized Rapid wage increases
1975–81 Centralized Wage indexation: undermined by discounting
CPI adjustments and lack of government
support
1981–2 Decentralized Further rapid wage increases
1983–7 Centralized Wage pause followed by CPI indexation:
some discounting in exchange for social wage
benefits
1987–9 Partially Two-tier system and structural efficiency
decentralized reviews
Both unemployment and inflation improved markedly in the early years

of the Accord. In two years inflation was halved from its 1982–3 high of 11.5
per cent although it then began to rise again (Fig. 13.6). This was accom-
panied by strong employment growth. The government met its promise to
create 500,000 jobs in its first three years, and by the end of 1987–8, over 1
million jobs had been created. Increased participation (notably by women)
reduced the impact of employment growth on the rate of unemployment.
Nevertheless by June 1986 unemployment had fallen from a peak of 10.2
per cent to 7.8 per cent (Fig. 13.7). The new jobs were fuelled by strong
growth in the Gross Domestic Product. Negative growth in 1982–3 was
followed by three years of continuous GDP growth at almost 5 per cent per
annum (Table 13.3). The contractionary policies announced in the 1986
budget can be seen in reduced GDP growth and a slight increase in
unemployment during 1986–7.
All in all the first three years of the Accord seem to provide strong
prima-facie evidence of the advantages of corporatism.
FIG. 13.6 Inflation rates, Australia: quarterly data, percentage change

on a year earlier
Note: D = December; M = March; J = June; S = September.
However international comparisons shed some doubt on this. While there

is no doubt that Australian GDP and employment growth were remarkably
strong by international standards (Table 13.3 and Table 13.4), neither
inflation nor unemployment compare especially favourably with OECD
norms. As Fig. 13.6 shows, the fall in inflation that occurs during the
Accord’s first two years parallels a similar fall in the inflation rates of
Australia’s major OECD trading partners (G7 plus New Zealand) which
occurred more than a year earlier. Moreover in 1985–6 Australia’s inflation
rate rises sharply while that of its major trading partners is stable and falling.
Table 13.4 shows that since receiving a boost from the 1982 recession,
Australia’s unemployment has never managed to fall below the OECD
average.
Given the ambiguity of international comparisons I will turn to Matti
Pohjola’s suggestion and review the literature that has compared the per-
formance of the different wage-fixing systems that have operated in Australia
under stagflationary conditions. The Accord’s strategy for tackling stag-
flation is built on the assumption that wage indexation will restrain the
aggregate growth of real wages. Hence whether indexation does in fact
reduce aggregate real wage results is an important test of its success.
Lewis and Kirby conduct such a test by developing a theoretical model of
the wage determination process which attempts to estimate the equilibrium
real wage which could be expected on the basis of labour supply and demand.
They conclude that over the first two years of the Accord ‘equilibrium real
398 ROBIN ARCHER
FIG. 13.7 Labour market aggregates, Australia

Note: Old series linked to new series at Feb. 1978 and Apr. 1986.
TABLE 13.3 Real GNP in Australia

Year Australia3 Sum of seven
countries’3
1978–9 4.8 4.3
1979–80 1.9 2.4
1980–1 3.3 0.9
1981–2 1.8 0.8
1982–3 –1.2 0.5
1983–4 4.9 4.7
1984–5 4.9 3.8
1985–6 4.6 3.0
1986–7 2.7 2.6
1987–8 3.6 n/y/a
a
Data refer to GDP at 1984–5 price levels.
b
Financial year data based on sum of
seasonally adjusted quarterly data. This
column comprises the United States, Japan,
West Germany, France, Italy, Canada, and
the United Kingdom using 1980 exchange
rates and price levels.
Source: National sources and OECD Quarterly
National Accounts.
wages were over 5 per cent below what would have been expected from
previous experience’ (Lewis and Kirby 1987).21
A number of studies draw similar conclusions about the earlier 1975–81
indexation period. In particular Gregory (1986) measures the difference
between actual average weekly earning outcomes and those that would be
expected if the Arbitration Commission were simply a veil for market forces.
Two wage equations are developed to estimate expected market outcomes.
One is based on unemployment and the other is based on overtime data.
Both suggest that ‘overall, the 1975–81 indexation experiments kept the rate
of growth of nominal wages below levels that might have been expected’.
Furthermore, ‘these gains were not lost in a subsequent wage rebound’
(S70). Dornbusch and Fisher (1984) reach a similar conclusion about the
rate of growth of real wages. They estimate that indexation reduced real
wages by a little more than 2 per cent in the period 1975–80. This is similar
21
Simes and Richardson (1987) and Pissarides (1987) have produced alternative wage
equations which challenge these results. According to Simes, Lewis and Kirby have failed
adequately to specify labour market pressures. However Simes himself has been criticized. John
Beggs, an economist at the Australian National University, believes that Simes’s own work
would show that the Accord had been effective if he based his results on a non-parametric test.
TABLE 13.4 International employment and unemployment
Year3 Civilian employment (% change on previous perioid) Unemployment rate (OECD standarized) (% of labour force)
United Japan West France United Italy Canada Aus- United Japan West France United Italy Canada Aus- Total
States Germany Kingdom tralia0 States Germany Kingdom tralia OECDd
1978–9 3.7 1.1 1.0 n/a 2.9 0.9 4.0 0.1 5.8 2.2 3.3 5.6 5.4 7.3 7.9 6.3 5.1
1979–80 1.8 1.2 1.5 n/a 0.6 1.4 3.8 2.3 6.2 2.0 2.9 6.1 5.2 7.5 7.3 6.1 5.3
1980–1 0.5 1.0 0.2 n/a –3.0 1.1 3.0 2.7 7.4 2.2 3.6 6.7 8.3 7.5 7.2 5.8 6.2
1981–2 0.1 0.8 –1.2 n/a –2.6 –0.4 –0.1 1.2 8.3 2.2 5.3 7.8 10.7 8.3 8.7 6.1 7.4
1982–3 –0.6 1.4 –2.0 –0.1 –1.7 –0.3 –2.9 –1.7 10.1 2.5 7.3 8.2 12.0 8.9 12.3 8.8 8.7
1983–4 3.8 0.9 –0.6 –1.4 1.1 0.4 2.8 0.9 8.2 2.7 7.5 8.9 12.2 9.7 11.3 9.6 8.3
1984–5 2.9 1.0 0.5 –1.2 1.7 0.5 2.5 2.8 7.2 2.6 7.2 10.1 11.4 9.9 10.9 8.5 7.9
1985–6 2.0 0.5 0.9 –0.5 0.9 0.6 3.3 3.8 7.0 2.7 6.9 10.2 11.1 10.6 9.9 7.9 7.8
1986–7 2.5 0.9 1.0 0.1 1.2 0.1 2.0 2.3 6.6 2.9 6.4 10.6 11.0 n/y/a 9.4 8.2 7.6e
1987–8 2.6 1.6 n/y/a n/y/a n/y/a n/y/a 3.7 3.0 5.7 2.7 n/y/a 10.4 9.3 n/y/a 8.1 7.7 n/y/a
a
Seasonally adjusted.
b
Excludes agricultural and non-market sector employees.
c
Data based on ABS Cat. Nos. 6202.0 and 6203.0.
d
Total OECD comprises Canada, the United States, Japan, Australia, Austria, Belgium, Finland, France, Germany, Italy, The Netherlands, Norway,
Spain, Sweden, Switzerland, and the United Kingdom.
e
Preliminary.
Source: OECD Quarterly Labour Force Statistics and OECD Main Economic Indicators (national sources are also used where necessary to update civilian
employment data).
to Lewis and Kirby’s (1987) finding that equilibrium real wages were reduced
by 2.6 per cent during 1975–81 with little evidence of a subsequent ‘catch-
up’ effect.
If these studies hold up for the 1975–81 indexation period they provide
important support for the claim that the 1983–7 indexation period was also
effective in reducing real wages. For during the 1975–81 period, wages drift
outside the system reduced the downward effect of indexation on real wages
and ultimately undermined it. Partial discounting of full CPI indexation
which began just twelve months after the system was introduced put pres-
sure on unions to make up the difference outside the system. During the
1983–7 period the Commission has explicitly sought to minimize wages
drift by rigorously enforcing its ‘no extra claims’ ruling. And when discount-
ing of full CPI indexation has taken place it has generally been offset by
social wage improvements22 thereby taking pressure off unions to move
outside the system. Furthermore the 1983–7 indexation period did not end
in a decentralized wages explosion like its predecessor but rather in a
managed partial decentralization which maintained a cap on aggregate wage
outcomes. It seems reasonable to suggest, therefore, that the Accord will be
at least as effective as the 1975–81 system in reducing real wages. Given the
assumption that employment growth is, at least in part, a function of real
wage restraint,23 these findings imply that the Accord has been responsible
for at least some of the 1 million jobs that have been created under its
auspices. A comparison of Australian employment growth with that of the
major OECD economies (Table 13.4) suggests a similar conclusion. In 1988
Treasurer Keating could claim that ‘our rate of job growth is unmatched
in the Western world’ (1988: 2). Since the economic claims that are made
on behalf of corporatist systems generally rest first and foremost on their
employment performance, it is fair to conclude that the Australian Accord
should join the ranks of the corporatist ‘star performers’.24
Of course, in the Accord’s own terms, job creation was only one of two
key objectives. With falling real wages, the second key objective—the mainr-
tenance of living standards—seems to be out of reach. But this ignores the
importance of the social wage trade-offs mentioned above. In fact the
government is keen to point to the statistics in Table 13.5 which show that
while both award rates and average earnings have fallen in real terms during
22
Tax cuts, pensions, family allowances, and the introduction of occupational superannua-
tion and universal health insurance have been prominent examples of social wage improvements
which have been traded off for reduced real wages.
25
In the late 1970s this assumption was hotly debated in Australia, but it is now widely
accepted. See the Australian Financial Review, 26 Sept. 1988.
24
Further supportive evidence for the efficacy of the Accord comes from two studies by
Beggs and Chapman (1987a and 1987fc) who argue that it has significantly reduced strike
activity. Through both an international comparison and a model of previous Australian
experience they show that changing macro-economic conditions account for only 40% of the
decline in strike activity during the first three years of the Accord.
402 ROBIN ARCHER
TABLE 13.5 Wages and household income in Australia

(% change on preceding year)
1982–3 1983–4 1984–5 1985–6 1986–7 1987–8 1988–93
Employment –1.9 0.9 3.2 3.7 2.9 3.1 2.8

At constant (1984–5)
prices0 0.2 –1.9 –0.6 –3.3 –3.6 –3.0a –1.0
Average award
rates of pay
Average earnings 2.0 –2.4 0.8 –1.9 –1.9 –0.6 1.3
(national accounts
basis)b
Household dis-
posable income –0.2 3.9 2.3 2.8 0.2 2.1 3.0
Household dis-
posable income –1.8 2.6 1.1 1.4 –1.3 0.6 1.5
(per capita)
a
Forecast.
b
Non-farm sector.
c
Deflated by the private consumption deflator.
the life of the Accord, average real household disposable income and (except
for one year) average real household disposable income per capita have
shown reasonable growth.
However even measuring the performance of the Accord with respect to
both employment growth and the maintenance of average living standards
is somewhat restrictive. Bob Rowthorn (Chapter 4 above), for example, has
suggested that wage dispersion should also be taken into account. He rightly
argues that people do not only want work. They want well-paid work.
There has been a wide-ranging and inconclusive debate about whether the
Australian arbitration system has led to more egalitarian wage dispersion.25
Current official wisdom holds that there is nothing especially unique about
the structure of Australian wage relativities.26 And it is certainly not clear
that periods of centralized wage fixation have compressed wage relativities
vis-d-vis other periods. For example there is evidence that relativities
widened in 1953–67 (when the basic wage plus margins system might have
been expected to compress them), that they were compressed in 1967–74
(when the relatively free reign of market forces might have been expected to
widen them), and that they widened again in 1975–82.27
25
For recent reviews of the literature see Gill (1987a) and Hancock Committee (1985: iii,
appendix II).
26
See Hancock Committee (1985: iii. 11–15) and National Labour Consultative Council
(1987: 14).
27
For the first two periods see Gill (1987fc). Presumably part of the compression in 1967–74
is due to the Arbitration Commission’s implementation of equal pay for women in the early
1970s. For the third period see Peetz (1985: 16–23 and appendix C). Unfortunately Peetz does
However it does seem that the Accord has halted the expansion of
relativities that preceded it. Peetz (1985: 20) has shown that during its first
two years the Accord has maintained existing relativities. The flat dollar
increases awarded in the 1987 and the 1988 national wage cases may slightly
compress wage relativities. But even if they do not, it is not necessarily
appropriate to criticize the Accord for failing to generate a more egalitarian
wage dispersion. Because what really matters is the dispersion of total
income. Given that die willingness of the Accord partners to trade off real
wage reductions for social wage improvements has been an important
feature of the Accord’s success, this is of fundamental importance. Indeed
the original Accord sees fiscal rather than wages policy as the main vehicle
for achieving redistributive measures. In light of this, wage dispersion is not
really an appropriate measure of the Accord’s success.28 Unfortunately the
work needed to establish whether the Accord has compressed or widened
total income dispersion has yet to be done. Prima facie the government’s
record is ambiguous. Some initiatives, such as the introduction of Medicare,
where a flat-rate premium for health insurance was replaced by a fixed-rate
percentage of income, have had a progressive effect on income dispersion.
Others, such as income tax cuts where the top marginal rate was reduced
from 60 per cent to 49 per cent and the bottom rate from 25 per cent to 24
per cent, have been regressive.
13.4 INDUSTRIAL RESTRUCTURING
Since the mid-1980s stagflation has started to fade in the OECD countries
as both unemployment and inflation slowly decline. Instead, attention has
begun to focus on another, perhaps more fundamental, economic trans-
formation: far-reaching industrial restructuring in the advanced capitalist
countries. There have been a number of heralds of this transformation.
Michael Piore and Charles Sabel have been especially influential. Piore and
Sabel (1984) and Sabel (1982) argue that a combination of international
pressures from the newly industrialized nations and domestic consumer
demand is forcing manufacturers in the advanced capitalist countries to
abandon the old ‘Fordist’ mass production methods in favour of ‘flexible
specialization’.
not distinguish between the contributions of the 1975–81 and the 1981–2 periods to the
widening of relativities, although evidence of a regressive change in income distribution from
1978–9 to 1981–2 suggests that the latter period made an important contribution, as might be
expected.
28
This also casts doubt on Freeman’s (1988) attempt to link corporatism, or even just
centralized wage bargaining, with low wage dispersion.
404 ROBIN ARCHER
In Australia this general imperative has been given special urgency by the
balance-of-payments deficit: whatTreasurer Keating (1988:4) calls ‘Australia’s
number one economic problem’. And few on either the left or the right
would disagree. Even though commodity prices have now recovered to their
1984–5 level, the long-term downward trend means that Australia cannot
continue to rely on its abundant natural resources to support a high standard
of living (Fig. 13.8). Rather the traditionally complacent and highly
protected manufacturing sector will have to become more internationally
competitive in order both to find new export markets and to fend off growing
import penetration. The government has recognized that this will require a
wholesale micro-economic reform programme of industrial restructuring
that aims to improve the productivity of individual firms. And, in order to
allow firms to make these adjustments, the government has placed a high
priority on achieving greater labour market flexibility (Department of In-
dustrial Relations 1988).
FIG 13.8 Real commodity prices, Australia

a
Nominal US$ non-fuel commodity prices weighted by Australian export, value deflated
by the United Nations US$ index of manufactured export unit values
But surely the greatest enemy of labour market flexibility is corporatism.

Whatever the merits of corporatism in dealing with the macro-economic
implications of stagflation, surely industrial restructuring calls for a decen-
tralized, liberal industrial relations system. But the extraordinary thing about
Australia is that Accord-style corporatism has proved compatible with the
micro-economic goals of industrial restructuring. In fact the Accord, or
rather the ‘Accord process’, has proved more than compatible with these
goals. It has become the driving force in bringing them about.
To understand how this has been possible requires an understanding of
the crucial role being played by the unions. And this in turn requires an
understanding of the fundamental transformation of the union movement
that the Accord has set in place. Less than ten years ago the Australian union
movement looked quite similar to its British counterpart. This provides a
benchmark against which the extent of the transformation of Australian
unionism can be measured. Whereas in Britain the Thatcher government
thought industry could only be modernized by undermining the unions, the
Hawke government is showing that modernization can be achieved with the
unions in the driver’s seat.
The Accord has made the unions more confident and forward-looking.
The old defensive unionism has been left behind and in its place is a new
agenda-setting unionism which aims, in the words of one of its architects, to
provide ‘a fundamental challenge for leadership to the nation’ (Carmichael
198(5). The decision to shift to the two-tier wage-fixing system of the Accord
Mark III was a key expression of this transformation. In particular it was an
expression of both the importance which the ACTU attaches to its role as a
national agenda-setter, and its willingness to adapt to the continually chan-
ging conditions imposed by the world economy in order to maintain this role.
But the shift to the two-tier system was not the only expression of this
transformation. More important, in a way, was the ACTU’s 1986 Mission
to Western Europe. Senior unionists visited Sweden, Norway, Austria,
Germany, and Britain. In a comprehensive report entitled Australia Recon-
structed they highlighted the advantages of the Swedish and Norwegian
industrial relations systems and emphasized the importance of ‘strategic
unionism’. Strategic unionism requires the development of an integrated
long-term strategy. To this end Australia Reconstructed recommended that
‘Australian unions should continue to develop the co-ordinated national
approach to policy formulation and implementation under the umbrella of
the ACTU’ (ACTUATDC 1987: 189). At its 1987 biennial conference the
ACTU adopted the recommendations of Australia Reconstructed as its policy.
Australia Reconstructed also served to re-emphasize union concern with
industry development. Industry policy has been an abiding theme of the
unions in the Accord period (Ewer 1988). In the early years of the Accord,
special importance was attached to the formulation of tripartite industry
plans in which employer and union commitments were given in return for
government assistance. Now, however, the unions are shifting their focus
from negotiations with government to direct negotiations with employers
which aim to reach agreement on plans to modernize production at both an
industry and an enterprise level.
For its part, the government believes that this sort of employer-union
co-operation is essential to the success of restructuring. Confrontation
between workers and employers would not only slow restructuring and
increase its costs. More importantly for the government, it would inhibit the
406 ROBIN ARCHER
development of a ‘productive consciousness’ in which ongoing changes to

improve productivity are widely accepted.
In September 1986 an agreement in principle was reached between the
ACTU and the employer peak councils (the CAI and the BCA) on the need
to improve efficiency and productivity at an industry and enterprise level.
This became incorporated into the two-tier wage system the following year.
In order to win the second tier increase of 4 per cent, unions had to offer
productivity improvements of an equivalent value. These typically involved
the giving up of so-called ‘restrictive work practices’. Two further agree-
ments have since been reached between peak union and employer organiza-
tions. The tripartite National Labour Consultative Council (1987) has
sponsored a report on Labour Market Flexibility in the Australian Setting, and
the CAI and the ACTU (1988) have issued a Joint Statement on Participative
Practices.
The report by the National Labour Consultative Council (NLCC) ex-
amines the implications for Australia of the OECD’s (1986) Dahrendorf
report on labour market flexibility. ‘Labour market flexibility’ has become
something of a catchphrase in the second half of the 1980s. The trouble is
that it actually refers to a great many different things. The NLCC report
identifies six distinct elements of labour market flexibility which are of special
relevance to Australia: the general level of wages, relative wages, non-wage
labour costs, labour mobility, work practices and work patterns, and educa-
tion and training.29 The difference between these elements is crucial. For
while some, such as more flexible relative wages, may be incompatible with
corporatism, others, such as a more flexible general level of wages, may
depend on it. The NLCC report argues that: ‘In comparison with other
countries Australia’s experience has been above average in terms of both
[external] labour mobility and the responsiveness of wages generally to
unemployment and inflation, and on a par with many countries in relative
wage flexibility and developments in labour on-costs. . . . However . . .
significant deficiencies remain in Australia’s education and training systems
and in the area of work organisation and working patterns’ (NLCC 1987:
1). This assessment is confirmed by OECD (1988b).
The government believes that greater flexibility in these two deficient
areas does not require wholesale deregulation (Department of Industrial
Relations 1988: pp. ii, 8). On the contrary, it argues that an at least partially
centralized industrial relations system is needed to maintain aggregate wage
restraint, systematic restructuring in all industries, and union co-operation.
Moreover a centralized system is needed to develop national training stand-
ards. Indeed in a number of countries, studies which are sympathetic to
corporatism have increasingly focused on training issues in particular as one
29
See NLCC (1987: 13). The Dahrendorf report also discusses conditions of labour such
as employment protection, and regulations relating to the labour market. OECD (1986: 9 ff.).
of the only areas where unions can secure an input in an increasingly hostile
and deregulated economic environment.30
To rectify these deficiencies the government has launched a labour market
reform programme which has two main elements: reform of post-school
education and training, and award restructuring. The first is principally
being pursued by the government with the support of the unions. The
second is principally being pursued by the unions with the support of the
government. Both take as their starting-point the importance of encouraging
the growth of a highly skilled work-force which is seen as a prerequisite for
maintaining a high-productivity, high-wage economy. But again, it is the
unions that have been setting the overall agenda. Indeed, according to the
Australian Financial Review, it is a ‘seasoned communist’, ACTU Assistant
Secretary Laurie Carmichael, who ‘has been responsible over the past year
for driving the debate on training and skilling in industry to the center of the
micro-economic reform agenda’ (6 Oct. 1988: 48).
Post-school education and training reform is focusing on the higher
education system. Under new funding arrangements the government is
seeking to make the higher education system more responsive to national
economic priorities. Institutions will be required to submit ‘educational
profiles’ outlining their goals and how they relate to national economic
priorities (Department of Employment, Education and Training 1988a: 29).
‘Performance indicators’ are being developed to measure the achievement
of these goals (p. 85). The government is also planning a large expansion of
the higher education system. Forty thousand new student places are to be
created in the 1989–91 triennium. This will mean that, by the end of the
triennium, 110,000 new places in higher education (an increase of 43 per
cent) will have been created since the government came to power in 1983
(Department of Employment, Education and Training 19886: 3). To offset
the cost of this growth a special graduate tax is being introduced. A
tripartitely controlled training fund levied on employers is also being con-
sidered.
Awards are legally enforceable documents which contain sets of job
classifications, pay rates, and conditions of employment for all individuals
employed in an occupation or industry. Many awards contain obsolete
classifications and conditions, some dating back to the beginning of the
century, which inhibit flexible working patterns and the efficient use of new
technology. But the goal of the unions in promoting award restructuring is
not merely to update these classifications and conditions. More importantly,
the unions are seeking to establish a career structure within each award so
that the lowest-paid worker can, by obtaining the right qualifications,
30
See Finegold and Soskice (1988) for a discussion of this in the British context, and Crouch
(1988), who takes a more general Europe-wide view.
408 ROBIN ARCHER
progress to the highest-paid job. Progression along this career path will be
linked to the acquisition of clearly established skill levels and pay rates will
be organized so as to create an incentive for each worker to acquire superior
skills.
Of course by allowing long-established conditions and working practices
to be altered the unions are giving up an important source of power. Job
classifications have typically been the final trenches into which unions retreat
in order to defend their members. For this reason the unions, or rather the
metal unions which are taking the lead, have set two preconditions which
employers must agree to before award restructuring can proceed. The first
is a guarantee of job security. The second is a commitment to the estab-
lishment of in-plant consultative committees which will be ‘the prime imple-
mentation instrument for the changes proposed under the restructuring’
(Harrison 1988). These committees will have an ongoing existence. The
unions envisage that they will subsequently become responsible for training
programmes. They mark the first systematic attempt to establish plant level
co-determination in the private sector in Australia. The Metal Trades
Federation of Unions see them as ‘the basic building block of industrial
democracy’(1988: 4).
Although the metal unions were the first to begin implementing award
restructuring, the ‘structural efficiency’ principle which the Arbitration
Commission laid down as a prerequisite for wage rises in the 1988 national
wage case has ensured that there will be a systematic restructuring of all
awards by 1990.
The structural efficiency principle requires all unions and employers to
commit themselves to industry- (and in some cases enterprise- ) level
negotiations to restructure their awards. There is no denying that this, along
with the attempt to link pay rates to the skill needs of particular industries.,
implies a more decentralized wage-fixing system. But does it imply a less
corporatist system? I think not. For one thing wage bargaining has only been
partially decentralized. Wages will still be subject to a maximum aggregate
outcome which will be a product of centralized negotiations. In the 1988
wage case this aggregate figure was largely determined by negotiations be-
tween the government and the ACTU prior to the Commission’s hearings.
In any case, wages are not the only thing that matters. Centrally negotiated
framework agreements on productivity, labour market flexibility, and par-
ticipative practices are also important manifestations of corporatism. Indeed
the growing willingness of employers to negotiate these sorts of agreements
at both an economy-wide and an industry level suggests that, for the first
time, an explicit class co-operation component is emerging from the Accord
process, marking the beginning of a fully tripartite Australian corporatism.
13.5 FUTURE PROSPECTS
Finally, I want to assess whether the conditions underpinning the Accord are
strengthening or weakening. That requires an assessment of the state of the
key parties in the Australian industrial relations system: the unions, the
employers, the Arbitration Commission, and the Labor government. Having
done that I want briefly to speculate about where an ongoing Australian
corporatism may lead.
The last section highlighted the major transformation which has taken
place within the Australian union movement. The shift from a defensive
sectional unionism to a strategic unionism in which the ACTU is the key
corporatist bargaining partner is dependent on the consolidation of a strong
centralized organization. We saw in section 13.2 that strong centralized
organization has three components. Two of them—organizational unity and
concentration of power—have been historically weak in Australian unions.
Both are now strong.
Following the consolidation of a single peak council in the early 1980s,
organizational unity is now being further strengthened by a concerted
ACTU-led effort to amalgamate as many as possible of the small craft unions
into about twenty large industrial unions. The government has backed this
objective by changing the industrial relations legislation to make amalgama-
tions easier and by asking the smallest unions (with less than 1,000 mem-
bers) to explain why they should not have to amalgamate.
Whether concentration of power will be diminished by the partial
decentralization of wage bargaining that is taking place is open to debate.
There is no doubt that wage claims continue to be the most important
bargaining chip in the hands of the unions. Although the ACTU remains a
major player in determining the aggregate outcome of wage rounds, it is clear
that power in this area is being somewhat decentralized. But it seems likely
that the ACTU’s growing influence on the overall macro-economic agenda
more than compensates for this. A recent survey of the most influential
people in Australia lists ACTU Secretary Bill Kelty as ‘one of the two most
important decision-makers in Australian politics’.31 Either way, throughout
the Accord period, concentration of power has been much greater than at
any previous time.
Ironically it is the one component of centralization which was historically
strong that is now causing concern at die ACTU. According to survey data
compiled by the Australian Bureau of Statistics, membership density has
fallen from 49 per cent in 1982 to 46 per cent in 1986.32 And it is falling in
31
‘The Australian Financial Review, 6 Oct. 1988: 48: ‘The ACTU Secretary must be counted
as one of the two most important decision-makers in Australian politics, sharing the mantle with
his close friend and ally, the Treasurer.’
32
An alternative official series based on union returns shows membership steady at 55%. The
410 ROBIN ARCHER
a period of strong employment growth during the tenure of a Labor govern-

ment. Worse still, union membership in growth industries and amongst
young workers is especially low (ACTU 1987: 15). It may be that the unions
are losing a generation of workers. Future Strategies for the Trade Union
Movement outlines the ACTU’s (1987) response to these problems. The
ACTU’s strategy focuses on amalgamations to bring about larger better-
resourced unions, improved union services including new financial services
such as the ‘ACTU Privilege Card’ which offers credit facilities and cheap
home loans, and an expensive publicity campaign to improve the image of
the union movement.
Overall then, for the immediate future, the preconditions for strong
centralized union organization have been enhanced, although, in the longer
term, much will depend on the ACTU’s success in reversing falling mem-
bership density.
But there is another threat to union unity which cannot be ignored. Six
years of corporatism have led to growing discontent amongst sections of the
rank and file. In particular, divisions have emerged in the all-important metal
workers’ union where leaders who have begun to articulate this discontent
have come into conflict with others who are leading the industrial restruc-
turing process. So far these developments have not been too damaging.
However, the possibility of union leaders losing touch with their members is
an ever present danger in a corporatist system. At its worst this can result in
union leaders mobilizing their central power to discipline, rather than
represent, their members. The ACTU recognizes this. Australia Recon-
structed emphasizes that the most successful corporatist countries combine
both central co-ordination and strong local organization and work-place
involvement (ACTU/TDC 1987: 176–8). But the point does not get the
attention it deserves.
Furthermore, with a tight labour market and growing dissatisfaction after
six years of real wage reductions, it would clearly be possible for a strong
union to lead a ‘break-out’ from the centralized wage-fixing system which
would result in wage ‘explosions’ similar to those that occurred in 1973–4
and in 1981–2. When the domestic airline pilots went on strike in August
1989 there was a widespread fear that this was about to happen. Rejecting a
national wage case decision that granted all workers a 6 per cent wage rise
with extra increases for the low paid, the already highly paid pilots demanded
a 30 per cent increase. Heavy-handed opposition to their claim came not
only from the airlines, but also from the government33 and the ACTU. This
ACTU (1987: 11) believes that ‘the lower figure is the more accurate of the two, although it
does slightly understate the true figure’.
33
The government provided military aircraft to help break the strike and backed a common
law action in which the court ruled that the pilots’ industrial action was illegal.
effectively marginalized the pilots, and after three months—a very long
dispute by Australian standards—it was clear that they had been completely
defeated. Despite fears about the precedents being set by government action
against the strikers, other unions remained united in their opposition to the
pilots’ demands.
In contrast to the unions the employers are as divided as ever. These
divisions were exacerbated by militant ‘New Right’ employers in the mid-
1980s who set up yet another peak employers’ organization. National wages
cases demonstrate the extent of the divisions. At the 1988 national wage case,
ten separate employer submissions were made to the Arbitration Commis-
sion (1988: 24–35) ranging from opposition to the very existence of the
arbitration system to support for the ACTU’s wage claim. However, the fact
that some employer organizations were prepared to support wage increases
at all was seen as evidence of a growing preparedness to play the corporatist
game. In fact the ACTU had been keen to encourage direct negotiations with
the employers prior to the hearings.34 A growing acceptance of corporatism
on the part of some employers seems to be part of a trend. It is assisted by
the fact that the New Right push seems to be running out of steam. The
trend can be seen in the willingness of the CAI to negotiate agreements with
the ACTU on issues such as participative practices, and is especially clear in
the metal industry where the Metal Trades Industry Association (1986) has
proposed a ‘Compact’ with the metal unions.
Although it was neither a sufficient nor a necessary condition for the
emergence of corporatism, the Arbitration Commission has provided a
useful ce.ntral forum through which to channel corporatist wage bargaining.
The new Industrial Relations Act is the first comprehensive attempt to
rewrite the legislation in this area since the original Conciliation and Arbitra-
tion Act was passed in 1904. In the main it merely updates and streamlines
the much amended older Act. However it does make one significant change.
The new Act allows for ‘certified agreements’ between unions and employers
which will allow individual employers to make special deals on wages and
conditions so long as they meet a Commission-administered test of the
public interest. The Commission has always been able to certify consent
agreements. But by formally authorizing a procedure which allows parties to
opt out of generally applicable wage-fixing principles, the new procedure
tends to undermine the Commission’s ability to provide institutional support
for a centralized wage-fixing system.
Finally, what about the Labor government? Prime Minister Hawke’s
July 1987 election victory, which brought the Labor Party an unpreced-
ented third consecutive term of government, has guaranteed that the
54
‘The ACTU . . . purposefully tried to bring employers into the accord bargaining process
. . . broadening the incomes strategy into a three-way accord of unions, government and
employers.’ Australian Financial Review, 5 Sept. 1988: 5.
412 ROBIN ARCHER
transformations wrought by the Accord will be more than a passing phase.

But there is a growing threat to the party’s electoral strength which is at least
in part a function of the Accord’s success. Falling real wages and an
increased profit share have led to a widespread perception that the govern-
ment is more concerned about its big business ‘mates’ than it is about the
woes of workers. As a result some of Labor’s core working-class supporters
are deserting the party. The July 1987 election was indicative. The govern-
ment won only by increasing its vote in a few carefully targeted marginal
seats. In some of its traditional strongholds its vote fell sharply, although not
by enough to lose the seat. This was widely interpreted as a protest vote by
the party faithful. However voting studies show that this sort of protest vote
can quickly become an entrenched habit. The loss of working-class support
was also apparent in a series of by-elections, and in the 1988 defeat of the
twelve-year-old NSW Labor government. The federal government has
responded to these developments by emphasizing the social justice com-
ponents of its policy. But further deterioration in the balance of payments
throughout 1989 compounded the problem and weakened the government’s
reputation for sound economic management.
What would happen if Labor lost power? The public involvement aspects
of corporatism would almost certainly come to an end.35 But if the labour
market parties continue to grow into corporatism, a more attenuated version
based on class co-operation may continue to exist. Even this is doubtful,
however. The Liberal Party’s industrial relations policy was formulated at
the height of New Right influence and it commits the opposition to a far-
reaching deregulation of the entire industrial relations system. The unions
are trying to guard against this by developing a stronger relationship with
employer organizations. But nascent corporatist sympathies amongst the
employers could quickly dissipate in a new electoral environment. In the
end, the only guarantee of Australian corporatism is the continued commit-
ment of the two original Accord partners. If either union support waned or
the government fell, corporatism would be seriously weakened. The second
problem is more likely to strike first.
But suppose Labor keeps winning at the polls, and the current industrial
restructuring process is successfully completed, what then is the future for
Australian corporatism? The restructuring process is focused on the tradi-
tional welfare goals of the labour movement: the creation and equitable
distribution of wealth. Yet the labour movement has always identified with
a second set of goals as well: participatory goals. And the Accord process
clearly carries the potential for major advances on this second front. Even in
the midst of a balance-of-payments crisis, the unions have been able to use
award restructuring to entrench important rights for workers and their
35
Although the ACTU (1987: 37) thinks that ‘it should still be possible to “trade” [with
non-Labor governments], albeit on less advantageous terms’.
representatives at the work-place. Union ambitions extend well beyond this,

however. In Australia Reconstructed the ACTU (ACTU7TDC 1987: 21–3)
recommends the establishment of a tripartitely administered National
Development Fund. This is an early attempt to deal with a general problem
which will sooner or later confront unions operating in a corporatist system.
Wage restraint in the present is predicated on the expectation of returns in
the future. But capital investment is a major determinant of that future. And
so unless unions can exercise some control over investment, it may not be
rational for them to exercise restraint. The expansion of union-controlled
superannuation funds is providing another vehicle through which invest-
ment decisions can be influenced. Treasurer Keating (1988) has suggested
that: unions should see the funds as a source of ‘institutional muscle’ which
could be used to protect their power in a more ‘hostile political envir-
onment’.
The National Development Fund proposal is clearly inspired by Swedish
experiments in economic democracy. There are certain advantages to having
second bite at the cherry. The Accord was a successful adaptation by the
Australian labour movement of failed British efforts to institutionalize a
social contract. In time, maybe the same can be done with failed Swedish
efforts to inaugurate economic democracy.
REFERENCES
ACTU (1987), ‘Future strategies for the trade union movement’, D164–87, Sept.
ACTU/TDC Mission to Western Europe (1987), Australia Reconstructed, Canberra:
AGPS.
ALP/ACTU (1983), Statement of Accord by the Australian Labor Party and the
Australian Council of Trade Unions Regarding Economic Policy, Melbourne: ACTU.
APPLE, N. (1988), The Economy and Industry Development: Report to AMWU National
Conference, July, AMWU.
ARCHER, R. (1988), ‘The Australian Accord’, International Review of Applied Eco-
nomics,!: 213–32.
Australian Conciliation and Arbitration Commission (1988), National Wage Case,
Aug., Dec. 640/88 M Print H4000.
BEGGS, J. J., and CHAPMAN, B. J. (1987a), ‘An empirical analysis of Australian strike
activity: estimating the industrial relations effect of the first three years of the prices
and incomes Accord’, Economic Record, 63: 46–60.
(19876), ‘Declining strike activity in Australia 1983–85: an international phe-
nomenon?’, Economic Record, 63: 330–9.
BERGER, S. (1981), Organising Interests in Western Europe, Cambridge: Cambridge
University Press.
BRUNO, M., and SACHS, J. D. (1985), The Economics of Worldwide Stagflation, Oxford:
Basil Blackwell.
414 ROBIN ARCHER
CAI/ACTU (1988), Joint Statement on Participative Practices, Melbourne: CAI and

ACTU.
CAMERON, D. R. (1984), ‘Social democracy, corporatism, labour quiescence and the
representation of economic interest in advanced capitalist society’, in Goldthorpe
(1984a).
CARMICHAEL, LAURIE (1986), ‘Seminar and Conference Discussion Paper’, AMWU.
CHAPMAN, B. J., ISAAC, J. E., and NILAND, J. R. (1984) (eds.), Australian Labour
Economics: Readings, 3rd edn., Melbourne: Macmillan.
CLARK, C. M. H. (1955) (ed.), Selected Documents in Australian History: 1851–1900,
Sydney: Angus & Robertson.
CLARK, V. S. (1906), The Labour Movement in Australasia, Westminster.
Committee of Review into Australian Industrial Relations Law and Systems (1985),
Australian Industrial Relations Law and Systems (Hancock Report), Canberra: AGPS.
CROUCH, C. (1979), The Politics of Industrial Relations, London: Fontana.
----- (1985), ‘Conditions for trade union wage restraint’, in L. N. Lindberg and
Charles S. Maier (eds.), The Politics of Inflation and Economic Stagnation,
-----(1986), ‘Sharing public space: states and organised interests in Western
Europe’, in John A. Hall (ed.), States in History, Oxford: Basil Blackwell.
----- (1988), ‘The future of industrial relations from a trade union point of view’,
mimeo, Trinity College, Oxford.
CROWLEY, F. (1980) (ed.), A Documentary History of Australia, iii: Colonial Australia
1875–1900, Melbourne: Nelson.
DABSCHECK, BRAHAM, and NILAND, JOHN (1981), Industrial Relations in Australia,
Sydney: Allen & Unwin.
DEERY, S., and PLOWMAN, D. (1985), Australian Industrial Relations, 2nd edn.,
Sydney: McGraw-Hill.
DE LEPERVANCHE, M. (1975), ‘Australian immigration 1788–1940: desired and
unwanted’, in Wheelwright and Buckley (1975).
Department of Employment, Education and Training (1987), Skills for Australia,
Canberra: AGPS.
------(1988a), Higher Education: A Policy Statement, Canberra: AGPS.
------(1988b), A New Commitment to Higher Education in Australia, Canberra: AGPS.
Department of Industrial Relations (1988), Labour Market Reform: The Industrial
Relations Agenda, Canberra: AGPS.
DORNBUSCH, R., and FISHER, S. (1984), ‘The Australian macroeconomy’, in R. E.
Caves and L. B. Krause (eds.), The Australian Economy: A View from the North,
EWER, P. (1988), ‘Industry policy in Australia: debate and practice’, International
Review of Applied Economics, 2: 189–212.
------, HiGGiNS, W., and STEVENS, A. (1987), Unions and the Future of Australian
Manufacturing, Sydney: Allen & Unwin.
FINEGOLD, D., and SOSKICE, D. (1988), ‘The failure of training in Britain: analysis
and prescription’, Oxford Review of Economic Policy, 4: 21–53.
FORD, B., and PLOWMAN, D. (1983) (eds.), Australian Unions: An Industrial Relations
Perspective, Melbourne: Macmillan.
FORD, G. W., HEARN, J. M., and LANSBURY, R. D. (1987) (eds.), Australian Labour
Relations: Readings, 4th edn., Melbourne: Macmillan.
FRENKEL, S. J. (1988), ‘Australian employers in the shadow of the Labor Accords’,
Industrial Relations, 27: 166–79.
GILL, F. (1987a), ‘Inequality and the arbitration of wages in Australia: an historical
perspective’, Australian Quarterly, 59: 210–21.
-----(19876), ‘Determination of wage relativities under the Federal Tribunal 1953–
74’, Working Papers in Industrial Relations, No. 2, Sydney: Department of
Industrial Relations, University of Sydney.
GOLDTHORPE, J. H. (1984a) (ed.), Order and Conflict in Contemporary Capitalism,
Oxford: Clarendon Press.
------(19846), ‘The end of convergence: corporatist and dualist tendencies in modern
Western societies’, in Goldthorpe (1984a).
GOLLAN, R. (1960), Radical and Working Class Politics: A Study of Eastern Australia
1850–1910, Melbourne: Melbourne University Press.
------(1975), ‘The ideology of the labour movement’, in Wheelwright and Buckley
(1975).
GREENWOOD, G. (1955) (ed.), Australia: A Social and Political History, Sydney: Angus
& Robertson.
GREGORY, R. G. (1986), ‘Wages policy and unemployment in Australia’, Economica,
53: S53–S74.
GRIFFIN, GERARD, and GIUCA, VINCENT (1986), ‘One union peak council: the
merger of ACSPA and CAGEO with ACTU’, Journal of Industrial Relations, 28:
483–503.
HAGAN, J. (1981), The History of the A.C. T. U., Melbourne: Longman Cheshire.
Hancock Committee (1985), see Committee of Review into Industrial Relations Law
and Systems.
HARRISON, G. (1988), ‘Initial AMWU proposals for restructuring of trades classifica-
tions in the steel industry’, AMWU.
HiGGlNS, WINTON (1987), ‘Unions as the bearers of industrial regeneration: reflec-
tions on the Australian case’, Economic and Industrial Democracy, 8: 213–236.
HOWARD, W. A. (1977), ‘Australian trade unions in the context of union theory’,
Journal of Industrial Relations, 19: 255–273.
H. V. Evatt Research Centre (1988), The Capital Funding of Public Enterprise in
Australia, Sydney: H. V. Evatt Foundation.
KATZENSTEIN, P. J. (1985), Small States in World Markets, Ithaca, NY: Cornell
University Press.
KEATING, P. J. (1988), Budget Speech 1988–89, Canberra: AGPS.
KYLOH, ROBERT (1985), ‘Overaward payments and wage flexibility’, Wages and
Incomes Policy Research Paper No. 5, Canberra: Department of Employment and
Industrial Relations.
416 ROBIN ARCHER
LEHMBRUCH, GERHARD (1979), ‘Consociational democracy, class conflict and the

new corporatism’, in Schmitter and Lehmbruch (1979).
----- (1984), ‘Concertation and the structure of corporatist networks’, in Goldthorpe
(1984a).
LEWIS, P. E. T., and KIRBY, M. G. (1987), ‘The impact of incomes policy on
aggregate wage determination in Australia’, Economic Record, 63: 156–61.
MCLEAN, I., and RICHARDSON, S. (1986), ‘More or less equal? Australian income
distribution in 1933 and 1980’, Economic Record 62: 67–81.
MAIER, C. S. (1981), ‘The fictitious bounds of wealth and law: on the theory and
practice of interest representation’, in Berger (1981).
----- (1984), ‘Preconditions for corporatism’, in Goldthorpe (1984a).
MARTIN, R. M. (1980), Trade Unions in Australia, 2nd edn., Melbourne: Penguin.
MATHEWS, J. (1988), A Culture of Power: Rethinking Labour Movement Goals for the
1990s, Sydney: Pluto.
Metal Trades Federation of Unions (1988), The Metal Unions’ Handbook for Delegates,
Sydney: MTFU.
Metal Trades Industry Association of Australia (1986), ‘MTIA proposals for a
compact with the metal unions’, Dec.
National Institute of Economic and Industry Research (1988), ‘Restructuring
Australia’, National Economic Review, Mar., No. 8, Melbourne.
National Labour Consultative Council (1987), Labour Market Flexibility in the
Australian Setting, Canberra: AGPS.
OECD (1986), Labour Market Flexibility, Report by a high-level group of experts to
the Secretary-General, Paris.
----- (1988a), Economic Outlook, 43, June, Paris.
----- (19886), Economic Survey on Australia 1987/1988, Paris.
OFFE, C. (19 81), ‘The attribution of public status to interest groups: observations on
the West German case’, in Berger (1981).
PANITCH, L. (1986), ‘The development of corporatism in liberal democracies’, in L.
Panitch, Working Class Politics in Crisis, London: Verso.
PEETZ, D. (1985), ‘The Accord and low income earners’, Wages and Incomes Policy
Research Paper No. 7, Canberra: Department of Employment and Industrial
Relations.
PlORE, M. J., and SABEL, C. (1984), The Second Industrial Divide: Possibilities for
Prosperity, New York: Basic Books.
PlSSARlDES, C. A. (1987), ‘Real wages and unemployment in Australia’, Center for
Labour Economics London School of Economics, Discussion Paper No. 286,
June.
PLOWMAN, D. H. (1986), ‘Employers and compulsory arbitration: the Higgins era
1907–1920’’, Journal of Industrial Relations, 28: 588–609.
----- (1988), A Long Haul: The Origins of the Confederation of Australian Industry,
Industrial Relations Research Center Monograph, May, University of New South
Wales.
PONS, X. (1984), Out of Eden, Sydney: Sirius.
RAWSON, D. W. (1986), Unions and Unionists, 2nd edn., Sydney: Allen & Unwin.
REGINI, M. (1984), ‘The conditions for political exchange: how concertation

emerged and collapsed in Italy and Great Britain’, in Goldthorpe (1984a).
RIMMER,M. (1985), ‘Incomes policy in Australia, 1975–85’, Working Paper85–031,
Nov., Australian Graduate School of Management, University of New South
Wales.
ROWTHORN, BOB, and GLYN, ANDREW (1990), ‘The diversity of unemployment
experience since 1973’, in S. A. Marglin and J. B. Schor (eds.), The Golden Age of
Capitalism, Oxford: Oxford University Press.
SABEL, C. (1982), Work and Politics: The Division of Labour in Industry, Cambridge:
SCHMITTER, PHILLIPE C. (1981), ‘Interest intermediation and regime governability
in contemporary Western Europe and North America’, in Berger (1981).
----- and LEHMBRUCH, G. (1979) (eds.), Trends toward Corporatist Interest Intermedia-
tion, London: Sage.
SCHOTT, K. (1984), ‘Consensual macroeconomic policies: a cross country study of
economic performance and political arrangements in the 1970s’, Wages and
Incomes Policy Research Paper No. 2, Canberra: Department of Employment and
Industrial Relations.
SIMES, R. M., and HORN, P. M. (1988), ‘The role of wages in the Australian
macro-economy’, paper prepared for the 1988 Australian Economics Congress, 24
A.ug.–2 Sept.
---- and RICHARDSON, C. J. (1987), ‘Wage determination in Australia’, Economic
Record, 63: 144–55.
STILWELL, F. (1986), The Accord and Beyond, Sydney: Pluto.
STRAUSS, G. (1988), ‘Australian labour relations through American eyes’, Industrial
Relations, 27: 131–48.
Treasury (1988a), Budget Statements 1988–89, Canberra: AGPS.
----- (19886), Economic Round-Up, Sept., Canberra: AGPS.
TREGILLIS, S. (1987), ‘Uprooted, ripped-off, reconstructed’, Australian Society,
Sept.: 17–19.
TURNER, I. (1983), In Union is Strength, 3rd edn., Melbourne: Nelson.
WARD, R. (1965), The Australian Legend, Melbourne: Oxford University Press.
WHEELWRIGHT, E. L., and BUCKLEY, K. (1975), Essays in the Political Economy of
Australian Capitalism, \, Sydney: Australia and New Zealand Book Co.
WHITFIELD, K. (1988), ‘The Australian wage system and its labor market effects’,
Industrial Relations, 27: 149–65.
INDEX OF NAMES
Notes: Entries in bold type indicate contributions to this volume.
Adam, G. 30 bargaining 49–56 passim, 85–6,

Ahlen, K. 117 n. 90–1, 96–7, 102, 123
Aiginger, K. 252 relationship of employment and
Akerholm, J. 293 bargaining 133,144,218–19
Alestalo, M. 189, 194 study of corporatist economies
Alexander, J. 26 45–56 passim, 59, 64–8, 211
Allen, T. 122 n., 191–2, 194, 205 Cameron, D. R. 383 n.
Andersen, T. M. 310 n. Carlsson, B. 201, 263
Andrews, M. 62 Carmichael, L. 385–6 n., 405, 407
Aoki, M. 46, 78, 116 n. Cawson, A. 38
Archer, R. 7, 20, 31, 35, 46, 301 Chapman, B. J. 403 n.
The Unexpected Emergence of Clark, C. M. H. 381–2 n.
Australian Corporatism Cooper, R. N. 335 n.
377–413 Cremer, J. 77
Armingeon, K. 33, 34 Crouch, C. 377 n., 379, 380, 409 n.
Arrow, K. J. 221 Crowley, F. 382
Barro, R.J. 57, 185
Bean, C. R. 56, 65 Dahrendorf, R. 406
Beggs, J. 399 n., 401 n. Danthine, J.-P. 364, 366, 367, 372
Bell, L. 93 n., 94 n., 110 Davidson, C. 66 n.
Benn, A. W. 248 de Lepervanche, M. 386
Benoit, J. P. V. 203 Dewatripont, C. Ill
Bertolt, O. 33 Dooley, M. 180, 186
Biffl, G. 350 n., 370 n. Dornbusch, R. 399
Blaas, W. 6–7, 11, 37, 152 Dreze, J. 133
The Swiss Model: Corporatism Driffill, J. (with Calmfors) 25, 33,
or Liberal Capitalism 363–75 53, 380 n.
Bluestone, B. 93 n. indices of centralized wage
Boltho, A. 215 bargaining 49–56 passim, 85–6,
Bosworth, B. 139 90–1, 96–7, 102, 123
Braun, A. R. 84 n. relationship of employment and
Breuss, F. 366, 367 bargaining 133, 144,218–19
Bruno, M. 2, 26, 38, 56, 132 study of corporatist economies
Bunel, J. 32, 34 45–56 passim, 59, 64–8, 211
Bunn, R. 36
Burda, M. 120 n., 133 Elvander, N. 33, 34, 35, 36, 315 n.
Burtless, G. 120 n. Erixon, L. 310 n.
Butschek, F. 349 n. Ersson, S. 57
Esping-Andersen, G. 120,151,192,
Calmfors, L. 67 193–4,215, 303
Calmfors, L. (with Driffill) 25, 33, Ewer, P. 405
53, 67, 380 n.
indices of centralized wage Fagerberg, J. 319 n.
Farrell, J. 67, 77
420 INDEX OF NAMES
Faxen, K. O. 117, 118 n. Heidenheimer, A. J. 188

Feldstein, M. 180, 186, 192 Higgins, Justice 387–8,391
Finegold, D. 407 n. Hoel, M. 67
Fisher, S. 399 Horioka, Ch. 180, 186
Flanagan, R. J. 35, 118, 346 n. Horn, H. 67
Flora, P. 188 Horvei, T. 201, 206
Frank, R. 349 n. Howard, W. A. 390
Franz, H.-J. 39 Hyman, R. 215
Fraser, M. 393
Freeman, R. B. 47, 53, 56, 211 Jackman, R. 47, 53, 54, 66–7, 68
and wage dispersion 89 n., 93 n., Jallade, J.-P. 188, 189, 194
94 n., 110, 403 n. Johansen, L. 67, 287
Friedman, J. W. 69 Johnson, C. 245 n.
Frisch, H. 346 n. Jonsson, L. 118
Giddens, A. 26 Kaitala, V. 74, 75, 219

Gill, F. 388 n., 390 n. Kaldor, N. 203–4,215
Gintis, H. 180, 186 Kalecki, M. 203, 326, 340
Gladstone, A. 31, 34, 36 Katzenstein, P. J. 26, 30
Glyn, A. 7–8, 14, 40, 202, 299, 309 nature of corporatism 12, 178,
Corporatism, Patterns of 214, 298
Employment, and Access to and small open economies 16,
Consumption 132–61 211, 212, 302, 391 n.
and taxation and corporatism and Switzerland 31, 35, 85, 368 n.,
9–10, 65, 204, 311, 314 369 n.
Goldthorpe, J. H. 26, 242, 385 Kaufmann, F. X. 26
Gollan, R. 381, 383 Keating, P. J. 393–4, 403, 404, 413
Gorbachev, M. 20, 22 Kelty, B. 409
Grant, W. 38 Kemp, M. C. 65, 71, 72, 74
Green, E. J. 70 Keynes, J. M. 2, 339 n., 345 n.
Gregory, R. G. 399 Kindleberger, C. 215, 234
Grout, P. A. 217 Kirby, M. G. 397, 399^01
Guger, A. 3, 6, 10, 39, 255 Klau, F. 110
Corporatism: Success or Korkman, S. 11–12, 16, 258
Failure? Austrian Exchange Rate Policy and
Experiences 338–59 Employment in Small Open
Gustavsen, B. 46 Economies 280–97
Korpi, W. 12, 35, 192, 193–4, 303
Haase, K. 275 Kosonen, K. 9, 221, 228, 262, 299,
Hagemann, R. P. 197 303
Hamada, K. 71, 72, 74 Saving and Economic Growth
Hanf, K. 38 from a Nordic Perspective
Hansen, B. 67 178–207
Hardin, R. 26 Kotlikoff, L. J. 192
Harrison, B. 93 n. Kreisky, Bruno 342, 346
Harrison, G. 408 Kurihara, K. K. 203
Haurie, A. 74
Hawke, Bob 378, 388, 391 n., 393, Lambelet, J.-Ch. 364, 366, 367, 372
394,405,411–12 Lancaster, K. 11,75,217,219,287
Heckscher, E. 262 Landesmann, M. 6, 10–11, 139,
Hedstrom, P. 39 210, 313, 328
INDEX OF NAMES 421
Industrial Policies and Social Pekkarinen, J.
Corporatism 242–77 Corporatism and Economic
(with Vartiainen), Social Performance in Sweden
Corporatism and Long-Term Norway and Finland 298–336
Economic Performance and employment dispersion
210–40 134–5, 138, 178
Lane, J.-E. 57 and External trade 200, 258
Layard, P. R. G. 56, 62, 65 and Finland 214, 258, 273
Lefranc, G. 34 and state intervention 200, 203, 273
Lehmbruch, G. 12, 24, 25–6, 37, (with Pohjola and Rowthorn),
242, 377 n. Social Corporatism and
Leonard, D. 65 Economic Performance 1–23
Leppanen, S. 55, 60 Pestoff, V. 27
Lewis, P. E. T. 397, 399–401 Peterson, M. 38
Lmdbeck, A. 53 Phelps-Brown, H. 30
Lmdgren, H. 262 Piore, M. J. 46, 403
Long, N. V. 65, 71, 72, 74 Pissarides, C. A. 399 n.
Lundberg, E. 160, 262, 263 n. Ploeg, F. van der 71, 210, 217
Pohjola, M. 178, 210, 288
McCallum, J. 26, 56 and bargaining structures 37, 217,
McDonald, I. M. 62 219, 392, 397
Maier, C. S. 377 n., 379, 385, Corporatism and Wage
391 n. Bargaining 44–78
Martinelli, A. 32 and employment dispersion
Meade, J. E. 62 134–5, 144
Mehrling, P. G. 75 (with Pekkarinen and Rowthorn),
Meidner, R. 114, 333 Social Corporatism and
Mierlo, T. van 30, 31, 32 Economic Performance 1–23
Mittelstadt, A. 110 Pontusson, J. 117 n., 248 n.
Mitterand, Francois 32, 146, 216 Porter, R. 70
Mjoset, L. 200, 202, 203, 224 Poterba, J. M. 185
Modigliani, F. 182 Pryor, F. L. 1, 86 n.
Mulder, C. B. 65
Regini, M. 216, 378 n.
Newell, A. 56, 58, 69 Rehn, G. 35, 38, 114, 262–3, 333
Nickell, S. J. 56, 57, 62, 65 Rhodes, M. 29
Nielson, N. C. N. 195, 205 Richardson, C. J. 399 n.
Niemala, H. 193 Righart, H. 29
Nowomy, E. 252 n. Risager, O. 284 n.
Rivlin, A. B. 139, 312 n.
Offe, C. 30, 377 n. Robinson, Joan 339 n.
Okun, A. M. 100 Rowthorn, R. E.:
Olson, M. 26, 27, 45, 57, 355 and Austria 39, 138, 141, 356
O’Shea, C. 387 n. Corporatism and Labour
Oswald, A. J. 62 n., 64 Market Performance 82–125
and employment 132, 133, 134–5,
Paloheimo, H. 122 142, 144
Palokangas, T. 72 and Sweden 65, 309
Panitch, L. 14, 385 wage dispersion 39, 53, 178, 196,
Parri, L. 33 299, 300, 402
Peetz, D. 402–3 n., 403 (with Pekkarinen and Pohjola),
422 INDEX OF NAMES
Rowthorn, R. E. (cont.) Strath, B. 213, 267 n.

Social Corporatism and Summers, L. H. 180, 185, 186
Economic Performance 1–23 Swedberg, R. 39
Symons, J. S. V. 56, 58, 69
Sabel, C. F. 46, 403
Sachs, J. 2, 26, 38, 56, 132, 146 Tarantelli, E. 47, 58–61, 70
Saglio, J. 32, 34 Taylor, L. 203
Sautter, C. 216 Thatcher, Margaret 34, 216
Sawyer, M. 188, 193, 194, 195, 197 Therborn, G. 1–3, 85, 125, 212, 304
Scharpf, F. W. 38, 40, 217, 263 n., Lessons from ‘Corporatist’
311 n. Theorizations 24–41
Schmidt, M. G. 26, 38, 364, 366, 367 Tichy, G. 355
Schmitter, Ph. C. 12, 24, 25–6, 37, Tocqueville, A. de 27
242, 385 Treu, T. 32
Schneider, V. 38 Turner, J. 26
Schott, K. 75
Seidel, H. 347, 355 Uusitalo, H. 193, 195
Sen, A. K. 73, 211, 217, 219, 221
Shalev, M. 28, 35 Vartiainen, J. 10, 11, 226, 328
Sharpe, L. 38 (with Landesmann), Social
Shimomura, K. 74 Corporatism and Long-Term
Simes, R. M. 399 n. Economic Performance
Sisson, K. 31, 34 210–40
Siven, C.-H. 118 Vartia, P. 187, 200, 202
Skinner, J. 192 Viotti, S. 67
Skjaeveland, A. 197–8 Voorden, W. van 32
Slomp, H. 30, 31, 32 Vuori, S. 187, 200, 202
Soderstrom, H. T. 67, 182
Solow, R. E. 62 Walterskirchen, E. 355
Soskice, D. 407 n. Ward, R. 381, 382 n.
Spence, W. G. 382 Wells, J. 142
Steindl, J. 346 n. Wibe, S. 57
Stokstad, P. E. 197–8 Williamson, P. J. 1, 85 n.
Stolz, P. 367 Windmuller, J. 31, 34, 36
Strand, J. 66, 68 Wyplosz, C. 146
INDEX OF SUBJECTS
age, in bargaining theory 218–19 industrial restructuring 140,

agriculture 38 246–7, 255–7, 274–5, 342, 355,
Finnish 258, 261, 325, 326, 328 358
Norwegian 271 investment 228, 249–54, 340,
associative action 2, 25–7 350, 352, 355
Australia 378–413 National Bank 223, 226, 352, 354
the Accord process 387, 390–1, social welfare 39, 120–2, 124–5,
392–403,404–8,411–13 189, 190–1, 353
Arbitration Commission 387–90, state enterprises 226–7, 353
411 subsidies 274, 358
award restructuring 407–8, 412–13 taxation rates 197, 198, 352
corporatism: choice of 7, 20, 46, technology gap 220, 228, 237,
378, 379–92; future of 409–13; 252–6, 355
success of 392–408 trade unions 28–30, 223–4, 225,
demarketization 138, 143 346–9, 350
and employers associations 31, unemployment 104, 124–5,
386,411 140–1, 342–3, 354
industrial restructuring 404–6, wage dispersion 39, 83–8, 91–9,
412 102–11 passim, 115–16, 356–8
inflation 39, 392–4 wage flexibility 350–2
‘mateship’ concept 381–2 Austro-Keynesianism 274, 311,
National Development Fund 413 345–6, 355
trade unions 382–6, 390–2,
405–8, 409–13; and Labor banks:
Party 383, 384–6, 390–1, central 215–16, 223, 225, 280–2,
411–13; and unionization 85, 284, 296, 303
383, 386,409–10 in corporatist state 226
training 406–8 bargaining, centralized
wage bargaining 388–90, 393, in Australia 388–90, 393, 395–6,
395–6, 405–6 405–6
wage indexation 397, 399, 401 changing 17–19, 306, 315
Austria 249–57, 274–5, 301, 338–59 comparative national indices 50,
consumption and employment 59, 126
spreading 138–41, 143, 151, or decentralized 44–7, 63–8, 76,
155–7 328, 331, 334–5, 368
corporatism 31–6, 49–55, 59–60, and European integration 239–40
222, 346–9, 359 and inflation 53, 339
economic performance 249–57, in Nordic countries 305
274–5, 353 solidaristic 108–10,113–23,315
employment 47–53, 104, 119, and trade unions 212, 218
249–51, 256, 342–5, 366; full bargaining structure
252, 255, 274, 341–2, 353, 355; and age effect 218–19
of immigrants 340–3,356, Calmfors/Driffill index 50
370–1; of women 99, 105, and economic growth 54–5, 70–5,
124–5, 140–1, 343–4 76
exchange rate policy 292, 296, and employment 49–53 passim,
342, 352, 354, 358 75–6, 95–9, 111–13, 217–18
424 INDEX OF SUBJECTS
bargaining structure (cont.) and democracy 13–14, 213, 221

short term analyses 61–8 and employment 47–58, 60–8,
Tarantelli’s index 59 95–113, 132–61
and wage dispersion 94–5 Finnish 323–31
bargaining theory 77, 78, 222–3 future of 15–20, 221–3, 239^10,
Belgium 47, 138, 152–7 passim, 297, 409–13
161, 370 and industrial restructuring
bargaining structure 29–4 211–14, 216, 242, 245–6
social welfare 39, 190–1 and inefficiencies of capitalism
benefits, welfare see social welfare 217–20
as institutionalized conflict 12–13,
Calmfors and Driffill index of 36, 125, 300, 303–4
centralization in bargaining 50 long term considerations 68–70,
Calvinism 29 210–40,244–8
Canada 35, 49–53, 59–60, 138, 143 national ranking 14–15, 50, 59, 126
wage dispersion 86, 88, 93, 103–6 in Nordic countries 298–336
passim, 356 Norwegian 317–21
capital: stability of 300–1, 333
allocation 199–203, 213, 226 Swedish 306–17
dynamic inefficiencies of 217–20, Swiss 363–75
287 Tarantelli’s index 59
mobility of 67, 72 technocratic 247
capital accumulation, Nordic model and technological change 213,
179–82, 199–203, 204–7, 228 216,220,222,237
capital flow 282, 288, 295, 322 credit:
Catholicism 29, 344 consumer 198–9, 206
Chartism 28 rationing 226, 281–2
chemical industry 252, 253, 266, 267 state control of 201–2
clothing industry 263, 264, 266, 267 Czechoslovakia 22
competition, in labour market 44–5,
66, 109 de-industrialization 40, 249, 268, 276
conflict: decommodification 151–2, 155–6
class 35, 75, 303–4 demarketization 135–47, 158–9, 161
and evolution of corporatism 221–3 Denmark 180, 196–9, 281, 370
institutionalized 36, 125, 300, 303–4 bargaining structure 31, 33, 35
consensus, and corporatism 26, 33, consumption and employment
36, 68–70, 125, 300, 304, 330–1 spreading 152–7, 160
consumption: employment 49–53, 82, 85, 119,
from earnings 147–61 138–9, 141
from property 149, 154 and European Monetary System
consumption spreading 133–4, 293, 295
147–61,311 exchange rate policy 282–3, 293, 294
corporatism: saving 185, 197–9, 204
Australian 378–413 social welfare 189–94, 204
Austrian 338–59 wage dispersion 91–5 passim,
concepts of 1–3, 12–15, 24–7, 101–2, 107
58–9, 82, 178, 298–300, 377–8 deregulation see financial markets,
and consensus 26, 33, 36, 125, deregulation
300, 303–4, 329 devaluation 283–91
in decline 322, 359, 377 in Finland 215, 223, 258, 273,
and ‘demarketization’ 135—11 289–91, 325–32, 334
INDEX OF SUBJECTS 425
in Norway 292, 319, 321, 332 exports 148, 152–4, 212, 215
in Sweden 275, 311–15, 332 Australian 394, 404
see also exchange rate policies Austrian 253, 254
discrimination, sex or ethnic 108, 109 Finnish 258–60, 263, 273, 274,
325,327–8
economies, small open 211–216, Norwegian 272, 276, 319–20
280–97, 302, 303, 354 Swedish 262, 265, 266, 275, 310,
Eire 7, 34, 35, 189 334
employers associations 30–3, 386, 411
employment 104, 119,249–51, Fascism 37
343–5, 372–3, 374 financial markets
in Australia 396 deregulation 185, 205–6, 226,
in Austria 104, 256, 342–5, 366 255, 281–2, 288, 295, 322
commitment to full 119, 286, and growth 225–6
291, 295, 300; in Austria 252, Finland 257–61, 273–4, 323–31
255, 274, 341–2, 353, 355; in agriculture 258, 261, 325, 326, 328
Norway 332; in Sweden 262, bargaining structure 31, 33–6,
275, 310, 313–16, 332 222, 225
comparative national indices 102, and central bank 289, 296, 303
121, 126–9, 167–9, 235–6, 250, consumption and employment
400 spreading 138–41, 143–5,
and corporatism 47–58, 60–8, 154–5, 157–8, 160
95–113, 119, 132–61 corporatism 49–55, 59–60, 214,
in Denmark 138–9, 141 222,323–31
in Finland 257–8 devaluation 215, 223, 258, 273,
in Norway 268 289–91, 325–32, 334
and productivity growth 135–61 economic performance 301,
spreading 8–9, 132–61 302–6,323–31
state 135–42, 144, 146, 150–5, employment 47–56, 249, 257–8
158, 194, 238, 263 exchange rate policy 282–3, 295,
in Sweden 307, 308–9 296–7, 329
in Switzerland 364–8, 370–3, 374–5 forestry 325, 328
and wage dispersion 107–13 growth 200–2, 249, 257–61,
and wages policy 243–4 273–4, 323
of women 99, 105, 124–5, 140–1, industrial restructuring 257–61,
159, 275, 343–4, 373 273–4, 326, 330
engineering 263, 266, 267, 273 industrialization 187, 259
entry barriers 355 inflation 325–8, 329
equality, principle of 107–10, 133, investment 249, 257–61, 289,
188–9, 192, 194–6 325, 326
ethnic minorities, discrimination labour supply 214–15, 329
against 108, 109 saving 179–87, 192, 197–9, 204
Europe, Eastern 20–3 social welfare 120–3, 187–96, 204
European Community 20 state enterprises 226–7, 260, 273,
European integration 239–^10, 276 274, 303
European Monetary System 16, taxation 196–9, 205, 331
157,295–6,297 technology 220, 228, 253–4, 257,
European Trade Union Institute 33 259–61
exchange rate policies 11–12, 269, trade unions 225, 304–5, 306,
280–97, 329, 370, 375 323,326–7, 330–1
expenditure, public sector 227–8, 238 unemployment 323, 325–30
Finland (cont.) industrial order 25, 33–7

wage dispersion 85, 91–2, 96, 98, industrial policy, and corporatism
101–7 passim 10–12,244–76
wage rigidity 56 industrial restructuring see
Fordist production methods 18, 46, restructuring, industrial
116, 334–5,403 industrialization 199, 215, 259
forestry 252, 264–7 passim, 295 industry, nationalized see state
in Finland 258, 261, 289, 296, enterprises
325, 328 inflation 53, 243–4, 284, 285, 287
France 28–35, 146, 157, 189–92, in Australia 392–4, 397
216, 247, 272 in Austria 352, 354
in Finland 325–8, 329
game theory, and bargaining 27, in Norway 332
61–74, 84–5,217–23 in Sweden 262, 308, 310–11, 317,
Germany, West 56, 59, 88, 189–90, 332
268 integration, international 16–17,
bargaining structure 28–30, 33–7, 239–40,276,281
215–16 interest rates 226, 281–2, 284, 287,
economic performance 25, 252, 288,354
340, 341 intervention, state see state
immigrant labour 370–1 intervention
growth: investment 179–80, 185–7, 199–203
and bargaining structure 54–5, Australian 413
70–5, 76 Austrian 254, 340, 352, 355
comparative national indices 180, comparative national indices 180,
228–9,251,341 232,251
and corporatism 210–40 and consumption spreading 148,
employment 234–7 152, 154, 160
finance for 179–80, 186–7, in corporatist states 182–7,
199–202, 225–6,231–4 199–203
gross domestic product 228–9 and exchange rate policies 286, 287
manufacturing 229–30 Finnish 260–1, 325, 326
productivity 230–1 and growth 179–80, 186–7,
and trade unions 57–8, 223–5 199–202,225–6,231–4
Norwegian 270
hours, working 136–44, 146, and pension funds 192–3, 413
149–55, 161 Swedish 262, 263, 275, 314
housing 198–9, 206, 320 Swiss 368–9, 370
‘hysteresis’ effects 284–6 investment allowances 228, 254,
260
immigration, and labour supply 141, Italy 157, 216
214–15, 340–3, 366, 368–72 bargaining structure 31–5, 37, 47
passim wage dispersion 39, 91, 105
income, distribution of 58, 74–5, welfare state 122–3, 189, 190–1
193, 194–6,412
incomes policy 35, 84, 224–5, 322, Japan 28, 31, 34–5, 247, 268
326–31 employment 25, 47–53, 59–60,
see also wages policy 143, 145
INDEED index (Index of wage dispersion 91–5 passim,
Employment and Earnings 103–4, 113,356
Dispersion) 100–7 wage rigidity 109–10
Labor Party, Australian 378, 383, consumption and employment

384–6, 390–1, 393,411–13 spreading 139, 145, 154–5,
labour market 157, 159
changes 105–7 corporatism 49, 50, 59, 301,
competition 44–5, 66, 109, 115 302–6,317–22
flexibility of 46 credit control 226, 320
INDEED index 100–7 de-industrialization 249, 268, 276
segmentation of 108–10 devaluation 282–3, 319, 321, 332
solidarity 119–25 economic performance 268–73,
supply 214–16, 233, 329, 343–4, 276, 301, 302–6, 317–22
366,372–4 employment 119,249,268,
Walrasian ideal in 5–6, 14, 109, 271–2, 302–6, 317–22; full 332
115, 220 exchange rate policy 269, 292–3
and women see women external trade 272, 276, 319–20
labour, mobility of 114–15, 118, industrial restructuring 246–7,
213, 218–19, 262, 316 272–3, 276, 320, 322
investment 187, 200–2, 270
market liberalization 44–5 multinationals in 200
Marxism, and corporatism 385 and North Sea oil 139, 141, 155,
metal industries 252, 253, 263–7, 157, 249, 268–76 passim,
272, 274, 327–8 319–21,322
mining 260, 263, 264, 266, 267 saving 179–87, 192, 197–9, 204
‘misery index’ 53, 61, 100 Social Democrats 303, 305,
mobility of labour see labour, 319–20
mobility of social welfare 120–3,187–96,
modernization see restructuring, 204–5
industrial state enterprises 226–7, 270–1,
monetary policy, in small open 303
countries 280–3, 284, 296 state intervention 225, 320–2
motivation, under solidaristic subsidies 270,271,319–20
policies 116–19 taxation 196–9, 205, 320
multinationals 16–17,200,220, technology 253, 254
252–3, 268, 275, 375 trade unions 225, 304, 305, 306,
322
nationalized enterprises see state unemployment 271, 272, 321
enterprises wage dispersion 84–8, 91–8
Netherlands 25, 138–40, 146, 151, passim, 101–7 passim
370
bargaining structure 29–30, 31–2, oil, North Sea
33–6, 47 and Norway see Norway
consumption spreading 154, and UK 272
156–7, 160
corporatism index 50, 59 payroll tax 262
welfare state 39, 190–1 pension funds 192–3, 290–1, 303,
New Zealand 34, 35, 47–53, 59 413
niche industries 212, 239, 261, 268, petrochemicals 263, 266, 267
274 Poland 21, 22
Norway 268–73, 276, 292–3, 301, ‘prisoners dilemma’ 58
317–22 privatization 255, 353
bargaining structure 31, 33–5, productivity 8–9, 135–61, 230–1,
224–5 252, 260–1, 270, 323
productivity (cont.) comparative national indices 121

comparative national indices in Denmark 189–94, 204
167–9, 171–3, 230, 236, 251 and employment spreading 133
and wages policy 111, 114, 243^1 expenditure 227–8
profit sharing 117 in Finland 120–3, 187–96, 204,
public goods, and centralized 323, 326, 329
bargaining 58, 68 in Norway 120–3, 187–96, 204
punishment strategies, game theory and solidaristic bargaining 119–23
68–9, 219, 222 in Sweden 120–3, 187–96^ 204,
311
radical liberalism 381–2 in Switzerland 189, 367, 370, 372
‘raiding’, and consumption and taxation 64, 65, 134, 148,
spreading 148, 152, 154 192–4, 195, 306
regional policies 270 Soviet Union 20–3, 328, 333
research and development 220, 245, specialization, in small open
260, 261, 267, 268, 276 economies 212, 258, 273. 274
restructuring, industrial 113–16, state, corporatist, and investment
140, 242, 245–7, 262–8, 275 182–7, 199–203,226–8
Australian 404–6,412 state employment 17–18, 194, 227,
Austrian 252–7, 274–5 238
corporatist response to 211, and employment spreading
212–14,216 135–42, 144, 146, 150–5, 158
Finnish 257–61, 273–4, 326, 330 state enterprises 226–7
Norwegian 272–3, 276 Austrian 226–7, 254–5, 256,
Swedish 262, 267, 275, 311–15 274–5, 277
Swiss 374 Finnish 227, 260, 273, 274
see also technology, new Nordic 200–1, 303
retirement, and labour supply Norwegian 226–7, 270–1, 303
343–4, 366 Swedish 227, 266–7
Ricardian Equivalence Principle state intervention 84, 199–203,
185, 186 214–15, 302–3, 305–6, 334
in Australia 386–7, 411
saving 73, 179–87, 192, 197–9, 203, in Finland 305–6, 323
262, 335 in Norway 225, 320–2
comparative national indices 180 state and public sector expenditure
and corporatism 221, 228 227–8, 238
see also capital accumulation ‘strategic unionism’ 405, 409
Sen’s isolation paradox 73, 211, strike activity 317, 368, 378, 390
217, 219, 220, 221 structural change see restructuring,
shipbuilding 261, 263–5, 270–3, 274 industrial
Social Charter 20, 239^0 subsidies, industrial 201
Social Democracy 20, 29, 303, 311, in Austria 274
316–17,385 in Finland 261
social steering, and corporatism in Germany 270
25–6, 37–40 in Norway 270, 271, 319–20
‘social wage’ 87, 120 in Sweden 263–5, 267, 270,
social welfare 39, 119–23, 189–93 275–6,310
in Austria 120–2, 124–5, 189–91, USA 270
353 Sweden 249, 262–8, 275–6, 301,
Bismarckian tradition 189, 190, 306–17
193–5 bargaining structure 31, 33, 35, 36
consumption and employment investment in 368–9, 370

spreading in 138–41,151–61, social welfare 189, 367, 370, 372
311 trade unions in 28–30, 368
corporatism 49–53, 59, 214, 301, wage dispersion 101, 373
302–17
devaluation 275, 311–15, 332 Tarantelli’s index of corporatism 59
economic performance 199–202, taxation 74–5, 192–9, 205, 306,
249, 262–8, 275–6 320, 331, 335, 352
employment 47–53, 249, 275–6, national comparisons 196–7
307, 308–9; full 119,262,275, and saving 186, 197–9, 202–3, 262
291, 310, 313–16, 332; state and social welfare 64–5, 134, 148,
263,311–13 192–5
exchange rate policy 282–3, 291, Taylorist work organization 46
296 technology, new 213, 216, 219–22,
external trade 253, 262, 265, 266, 228, 237, 252–6
275, 310, 334 in Austria 355
industrial restructuring 246–7, in Finland 257, 259–61
262–8,275, 311–15 in Sweden 262, 265, 267–8, 276
inflation 262, 308, 310–11, 317, see also restructuring, industrial
332 textiles 263, 264, 266, 267, 272
investment 249, 262, 263, 275, 314 trade unions 57–8, 222–5, 246–7,
mobility of labour 219, 262, 316 287–90, 333, 335–6
saving 179–87, 192, 197–9, 204, in Australia 382–6, 405–8, 409–11
262 in Austria 28–30, 223–4, 225,
Social Democrats 303, 305, 311, 346–9, 350
316–17 development of 27–30
social welfare 39, 120–3, 187–96, in Finland 304, 326–7, 330–1
204,311 future of 44–5, 46
solidaristic policies 114–19, in Norway 225, 304, 305, 306,
224–5, 262 322
state enterprises 227, 266–7, 303 in Sweden 311, 314, 316–17
subsidies 263–5, 267, 270, 275–6, unionization rates 304–5, 306,
310 323, 326, 331
taxation 196–9, 205, 262 training 213, 216, 220–1, 227, 245
technological progress 220, 228, in Australia 406–8
262, 265, 267–8, 276 in Austria 255, 274
trade unions 224–5, 246, 304–6, in Finland 260, 261, 273
311, 314–17 transfers see consumption spreading
wage dispersion 39, 84–8, 91–6
passim, 101–9 passim, 113, 315 unemployment 272, 285–6
wage earners and pension funds in Australia 392, 397
conflicts 248, 303, 316–17, 335 in Austria 255, 275
wage restraint 311,314–16 comparative national indices 121,
Switzerland 363–75 126–9,251,400
bargaining structure 31, 33–7, 86, and corporatism 49–50, 52–3,
368 75–6
employment 47, 82, 138, 364–8, in Finland 323, 325–30
370–3, 374–5; of foreign labour hidden 82–3, 104, 124–5
366, 368, 370–2; of women in Norway 271, 321
366, 369, 373–4 in Sweden 311–13
exchange rate policy 370, 375 in UK 271
unionization rates 57, 62, 85, wage restraint 217–18, 311, 314–16,
409–10 320, 325–7, 335
in Finland 323, 326, 331 wages policy 133–4, 159, 160,
in Norway 305, 306, 322 243–4, 262
in Sweden 304–5, 306 solidaristic 219–20,224
United Kingdom 189, 216, 248, see also incomes policy
271, 295, 340 wages, real 55–7, 62–3, 224, 283–4,
bargaining structure 28, 30, 34, 287–8, 296
35, 59 in Australia 397,399–401
United States 39, 123, 268, 270 in Finland 323, 328–30
bargaining structure 28, 30, 31, in Sweden 311, 314
32–3, 34 Walrasian ideal labour market 5–6,
corporatism index 59, 86 14, 109, 115, 220
employment performance 48, 49, welfare state 187–96
51–3, 60 Bismarckian tradition 189,190,
employment spreading 138, 143, 193–5
151 and distribution of income 193,
wage dispersion 88, 91–5 passim, 194–6
103–9 passim, 113, 356 and equality 188–9, 192, 194–6
wage rigidity 56, 110–11 financing 64, 65, 192–6
see also social welfare
wage dispersion 5, 39, 86–96, women:
100–13, 120, 315, 373, employment of 40, 108, 109
402–3 employment rates: in Austria 99,
comparative national indices 91, 105, 124–5, 140–1, 343–4; in
92, 102, 121, 126–9 Sweden 275; in Switzerland
male/female 89, 90–4, 96–8, 366, 369, 373–4
101–2, 104 wage dispersion amongst 89,
occupational differentials 87–8 90–4, 96–8, 101–2, 104
wage flexibility 56, 109–11, 116–17, working hours see hours, working
350–2 ‘working poor’ 83, 120

Social Corporatism: A Superior Economic System (Full Book

Uploaded by

Copyright:

Available Formats

Social Corporatism: A Superior Economic System (Full Book

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Social Corporatism: A Superior Economic System (Full Book

Uploaded by

Copyright:

Available Formats

This is a rekeyed full text file of the volume

Social Corporatism: A Superior Economic System?

Originally published in 1992 by Oxford University Press

File produced by Macmillan Publishing Solutions, Chennai

Studies in Development Economics embody the output of the

Jukka Pekkarinen, Matti Pohjola,

CLARENDON PRESS • OXFORD

Oxford is a trade mark of Oxford University Press

Published in the United States

© The United Nations University 1992

World Institute for Development Economics Research (WIDER)—

All rights reserved. No part of this publication may be reproduced,

British Library Cataloguing in Publication Data

Library of Congress Cataloging in Publication Data

Typeset by Alliance Phototypesetters, Pondicherry

Post-war macro-economic management in OECD has long pursued what

employment exchanges where each staff member has a case load of 15

1. Social Corporatism and Economic Performance: Introduction

ROBIN ARCHER, Balliol College, Oxford, UK.

3.1 Average employment and unemployment in 17 OECD

7.1 Productivity in manufacturing 230

3.1 Classification of countries according to their degree of

8.3 Share of high-technology products in exports of manufacturing

1.1 THE CONCEPT OF SOCIAL CORPORATISM

The notion of corporatism as an economic system distinct from both

1.2 LABOUR MARKET PERFORMANCE

The received view in the economic literature is that perfect competition

is that the bargaining structure may matter only in turbulent circumstances.

1.3 DISTRIBUTION OF CONSUMPTION AND SAVING

In Chapter 5 Andrew Glyn approaches the question of labour market

One method for spreading employment is simply to hold down labour

widely in the population. In Belgium, by contrast, relatively fast productivity

1.4 LONG-RUN PERFORMANCE AND INDUSTRIAL POLICIES

It is implied by several essays in this volume that exclusive attention to the

The papers by Landesmann and Vartiainen and Landesmann and

draws attention to financial integration which tends to render these tradi-

1.5 FURTHER REFLECTIONS ON THE CONCEPT OF

As was indicated at the beginning of this introductory chapter, in the course

institutionalization of conflict and that of consensus as the basis for industrial

seems to be historically associated with a strong thrust by strongly organized

industrial orders based on institutionalized conflict conducive to a strategic,

1.6 THE FUTURE OF CORPORATISM

The corporatist economies are in the process of change. Their relationship

The integration of the corporatist economies into the capitalist world

contributes directly to national output, whilst an increase in its sales abroad

latter quite reasonably expect higher wages in compensation. However, their

intervention. Indeed, the state already performs this kind of role on an

or avoid competitive tax reductions. This raises more general questions

1.7 ON THE TRANSFERABILITY OF THE CORPORATIST MODEL

A final intriguing question concerns the transferability of corporatist institu-

countries along corporatist lines, with a vigorous market economy, an activist

2.1 THE ‘CORPORATIST’ PARADOX

find any ‘clear evidence in favour of a non-monotonic relation between cen-

2.2 THREE ‘CORPORATIST’ CONTRIBUTIONS TO SOCIAL THEORY

The hard core of contemporary ‘corporatism’ can be summed up in: ‘interest

of corporatism should be unbundled and sorted out from the perspective of

2.3 ASSOCIATIVE ACTION AND TYPES OF CLASS FORMATION

Corporatism, of whatever definition, presupposes, derives from, a particular

classes of industrial capitalism. Secondly, as class organizations labour

prohibition. From this common revolutionary heritage, modern socialist