Towards a Regionalization of Industrial Relations*

zyxw Towards a Regionalization of Industrial Relations* PAOLO PERULLI In this paper ‘industrial relations’ are considered in a much wider sense than is usual among industrial relations scholars. Formal collective bargaining is just one aspect of the complex relationships among social actors which take place at national, regional and local levels. These include, among other things, the sharing of norms and practices regarding human resources training and management, labour market and economic development policies. These relationships often involve state agencies, public sector officers and various interorganizational networks which give different contexts and meanings to industrial relations. Our hypothesis is that the regional dimension is particularly appropriate to observe how these relationships are currently evolving into different contexts. The region as a unit of analysis has recently been relaunched, after a long historical prevalence of the national (and/or the enterprise) level. As neo-industrial systems tend to decentralize and territorially agglomerate their structure, the main competitive advantages (skills, R & D, flexibility, technological innovation diffusion) are likely to be found at the regional level. However, the new regional dimension is the result of economic organizational processes and institutional frameworks which are largely shaped by national habits and constraints. This is why apparently rather similar regional dynamics - currently defined as ‘network organizations’ - remain different with respect to institutional and, in this wider sense, industrial relations contexts. zy The role of the region in industrial decentralization The role played by the region in capital/labour relations, be they conflictual or cooperative, has become increasingly importance since the crisis of the 1970s and the subsequent restructuring and reorganization throughout the western world. During the previous 30 years of postwar stability, the envisaged convergence towards a single model of industrial development (the so-called taylorist/fordist model) produced, among other things, a lack of interest in the various aspects of regional differences in capitalAabour relations. By the 1960s, the region had become at most a derivative category of analysis and a secondary locus of economic activity. Despite continuing differences in national industrial structure, there was widespread agreement that the most effective production unit was the giant corporation, which at the limit integrated in one physical structure the activities of independent firms in the industrial districts. (Sabel, 1988) zy *I want to thank Angelo Pichierri, Jean Saglio and the IJURR referees for their comments on an earlier version of this paper. Support from the Commission of the European Communities/SPES and the University of Warwick project on ‘Centralisation and Decentralisation in the Euro-Company’ is also acknowledged. 0 Joint Editors and Basil Blackwell Ltd 1993. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 IJF, UK and 238 Main Street, Cambridge, MA 02142. USA zyxw zy zyxwvu Towards a regionalization of industrial relations 99 The prevailing economic theories described/prescribed the processes of industrial localization according to the model of the product life-cycle. There were appropriate locations for the enterprise for each stage of the cycle: start up, maturity and decline (for a critical analysis see Pichierri, 1986). The sociological aspects, in particular regarding conflict and cooperation in capital/labour relations, were practically out of the picture. The crisis of the dominant industrial model, which actually exploded in capital/labour relations, later gave rise to a new interest in the various possible local forms of social and labour regulation, different from the dominant model. During this phase of rediscovery of a ‘plurality of industrial forms’ (Bonazzi, 1989), awareness of the regional implications of different production regimes increased. These developments fall within three interrelated research programmes. A first area of study is that of local economies. The focus here is on the discovery of a persistence of local and regional societies where mass production never dominated and peculiar social phenomena took place (small and medium-size entrepreneurship, networks of trust relations linking social actors as firm owners and workers, local government, employers and trade unions). These local and regional economies performed very well during the crisis of the 1970s as their social climate favoured industrial adjustment. In this period, capital/labour relations reached new heights of conflict in almost all the industrialized countries, but the conflict was greatest in those areas dominated by large firms and mass production. The search for more favourable, less conflictual production areas led many firms to opt for decentralization and delocalization, whether or not this was a conscious choice. In this case, it was large ‘central’ firms looking for less conflictual places with a larger, flexible low-cost labour force available in ‘peripheral’ regions. The implementation of new production units did not always mean the decentralization of low-tech phases. Generally speaking, ‘the “decentralized” departments work at the same technological level as the central factory. Obviously, those phases of the production process that are not as strategic or technologically sophisticated are usually transferred to the outside’ (Bagnasco, 1988). In other cases, there was an increase of endogenous firms, born in regions characterized by small-firm local economies and destined in some cases to become a market leader in their sector. These same phenomena were also studied on an international scale; this was mainly the subject of a second research programme. In particular, the French school of ‘regulation’ (Boyer, Mistral, Lipietz etc.) stressed the ‘planetary’regional aspects of the process. Lipietz (1984) described it as a ‘geographic separation of three levels in the division of labour. Level 1 : planning; level 2: skilled production; level 3: unskilled assembly. ’ Whereas the first two tended to remain for the most part in the centre, the third level could be moved to new areas of expansion (in industrialized or newly industrialized countries) according to a model of ‘peripheral fordism’. In this case, the crisis of the fordist paradigm was seen as giving place to a variety of different mixed patterns: ‘Future may be a more fuzzy situation, with a coexistence of many types of capital-labour relations, even in a single country: a configuration of complementary patterns of industrial relations’ (Lipietz, 1990). A third area of study which recognized the importance of the regional variable is that of ‘flexible specialization’. Piore and Sabel (1984) distinguish between areas dominated by mass production and flexible specialization areas, characterizing the former by the predominance of fordist and neo-fordist production methods and the latter by the prevalence of flexible relations among enterprises (non-vertically-integrated industrial organization) and between management and workers. Subsequent studies, guided by these new examples, looked for common characteristics in industrial organization even in areas very different from one another - from the high-tech American regions to the French technopoles, from Italian and German ‘industrial districts’ to the Japanese cities of automobile and electronic production (Scott, 1989; Esser and Hirsch, 1989). If there is a common link among different areas of research (in disciplines ranging zyxwvutsrqp 0 Joint Editors and Basil Blackwell Ltd 1993. z 100 zyxwvutsrq zy Paolo Perulli from sociology to economics to industrial organization to geography), it may be the recognition that capital/labour relations do not depend solely on technological and organizational factors within the firm (an obvious consideration nowadays, but one that dominated in the social sciences in the past). External factors play a significant role in determining whether capital/labour relations are cooperative or conflictual. These factors are institutional and cultural in nature, and vary greatly according to the locality. The role of the region becomes an important topic of research in this context. We can summarize the ‘rediscovery’ of the region along these lines of argument: ( 1 ) The persistence and the economic success of local and regional economies based on the ‘industrial district’ principle. The industrial district as the privileged unit of analysis, instead of the industry or the corporation, means that population density, infrastructures and ‘industrial atmosphere’ are referred not to the industry but to the region. It is a complex network of external economies, cost connections, historical and cultural heritages, which cover both interfirm and interpersonal relations (for the ‘rediscovery’ of Marshallian industrial districts, see Becattini, 1987). zy zyxwvuts zy (2) The decentralization pattern emerging in large firms, particularly multinationals. This means that both industrial organization and capital/labour relations are reorganized in decentralized units with more or less autonomy. In neo-fordist firms, decentralization does not give rise to fully autonomous business units, whereas this is the case in postfordist firms. Industrial relations are normally devoluted to plant-level organizations. ( 3 ) As small-firm industrial districts mature and large multinational firms decentralize, a pattern of convergence through mutual adjustment can take place. Alliances and subcontracting relations characterize at least some industrial regions. Small firms look for coordination structures at local level, and large firms encapsulate their local business units in the regional economy. (4) Partly as a consequence of the new industrial developments at local level, local and regional governments are induced to develop intervention in the local economy through assistance to small firms, new infrastructures and support agencies, vocational training programmes etc. This process has taken place in most regions during the 1980s, also in part as a consequence of the breakdown of ‘nation-state’ policies in the field of welfare and industrial policies. ( 5 ) Social actors, such as employers and trade unions, have developed new pragmatic approaches to industrial adjustment at local and regional levels. Whereas at national level these actors often tend to maintain rather conflictual roles, at local level a practice of (often informal) cooperation to solve employment problems prevails. Conditions external to the firm which favour cooperation In the pluralist theories of industrial relations, ‘while each partner pursues its own interests, both admit their mutual dependence’ (Bakke, 1945, cited in Fox, 1973). This philosophy of mutual behaviour was put to the test during the explosion of industrial conflict in the 1970s. In fact, according to the pluralists, mutual dependence meant that there was a sharing not only of rules for handling conflict, but also of values and even ideologies of mutual survival (Kerr, 1954), whereas in the 1970s the extreme division of labour which characterizes industrialized societies produced ‘low trust’ relations, which in turn produced social instability (Clegg, 1975). On the other hand, high trust relations seem possibly only if there is a change in the division of labour upon which industrial economies are based. The problem, unresolvable as far as enterprise-based industrial relations are concerned, is easier to handle if one considers that the partners refer to trust 0 Joint Editors and Basil Blackwell Ltd 1993. z zyxw zyxwv Towards a regionalization of industrial relations 101 relations and develop mutually dependent relationships, not only at enterprise level (the only one considered by pluralists), but also in the local society and economy where firms are located. In his book on citizenship, T.H. Marshall (1976) pointed out that cooperation in labour relations depends on how much the worker trusts the state (how it handles industrial conflict, guarantees social security etc.) and the managers of the firm he works for (the microcosm of the citizenship). In another part of his essay, Marshall stated that ‘the national community is too big and remote to create this type of trust and make it a constant motivating force. This is why many think that the solution to the problem lies in the development of more limited trust in the local community and in the work group in particular’. In this last respect, Marshall was one of the first to idedify four levels (information, consultation, delegation, participation in control) as stages during which modem industrial democracy actually developed. But the local community as a place where industrial citizenship develops is not given as much weight. This underestimation is probably justified by the fact that in sociology the community is identified with productive relations and forms of division of labour that are backward and premodern, hence insignificant in explaining the dynamics of modern capitalist productive relations. The literature on industrial relations has grossly ignored the local and regional dimensions, limiting itself to the workplace and national institutions. Another significant factor which may explain this neglect is that industrial relations as a discipline was born and developed in Anglo-Saxon countries with a majority of large firms and highly decentralized labour relations at plant level, whereas local and regional labour regulation systems were weak or non-existent. In this extremely fragmentary context, the firm (if not the plant) was the real level upon which capital and labour were evaluated, particularly as the discipline of postwar industrial relations grew stronger. Equally important is the fact that during this period the union organizations consolidated into large national unions with bargaining power and political influence. ‘In this era of large firms and unions, most unions are large organizations. But this was not always the case. The first unions were small, local organizations, and for a certain length of time they stayed that way’ (Olson, 1983). This observation is related to his theory of collective action, whereby small groups can procure collective goods more easily than big groups. Following this line of thought, one could maintain that after a period of initial development of local unionism (small unions able to procure collective goods more easily) and a traditional phase of general unionism (based on standard rules in effect in the large firm and on internal labour markets), opportunities opened up again for local regulation, in which cooperation is the product of participation in (and identification with) local institutions and regulation systems. This idea, among others, was recently set out by Streeck (1985) in his paper on the management of uncertainty. Managers are said to feel that external labour markets can work properly, i.e. provide for flexible job placement and labour mobility among different production units, only if there is a set of shared rules that standardizes the merchandise being handled, regulates the basic terms of the exchange, limits the size and area of competition both among buyers and sellers of the labour force, and allows participants in the market to develop stable, long-term expectations. The question of cooperation can also be examined using the tools of game theory. According to Elster (1989), some forms of cooperation are based on externalities created by individual action, for example, participation in group action, a production cut in an individual firm if it belongs to a cartel, a solidaristic wage policy. These behaviours are often, though not necessarily, tied to some sort of mutual behaviour. Every act of cooperation zyxwvu 1 . An additional consideration concerns Japan, the country of enterprise unionism. In this case, enterprise cooperation seems to depend on the late nature of the country’s industrialization, which led to the precocious formation of internal labour markets in large firms. In addition, Japanese industrial relations were literally ‘written’ according to the American model. 0 Joint Editors and Basil Blackwell Ltd 1993 102 zyxwvutsrq zy Paolo Perulli brings some small benefit to everyone, including the cooperator. Although the direct benefits for the cooperator are too small to motivate him to act, given that there are also the costs of cooperation, it is better for all if everyone (or at least some) cooperate rather than no-one. Under conditions of generalized cooperation, everyone reaps the many small benefits of the cooperation of others, which all told outweigh the cost of this cooperation. The ‘regionalization’ of cooperation is concerned with the presence of externalities. External factors are those which cannot be produced by the single enterprise nor, within the single enterprise, by means of the cooperation between capital and labour. We will consider here three main policy fields where cooperation can be improved at the regional level: (a) manpower policies and vocational training institutions; (b)wage policies in regional labour markets; and (c) industrial relations policies. We need to remember that in all three cases formal policies (and formal industrial relations regulations) are only one side of the story. On the other side, we find also informal conventions and practices which are of the utmost importance. As an example, retraining and relocation of the workforce which is involved in a redundancy process often take place as a result of informal cooperation among employers and trade unions at local level. Solidaristic wage policy at local level means that both employers and workers’ representatives will respect unwritten guidelines which limit wage differentials. The involvement of the workforce in decision-making concerning industrial organization (i.e. subcontracting, organizational and technological innovation, training) is often the result of a largely informal local culture which is diffused among and shared by the different social actors. We turn now to a systematic treatment of the three fields of cooperation. zyxwvu zy Regional systems of cooperation Three types of cooperation are considered here: capital-capital, that is between different enterprises belonging to the same area; labour-labour,when groups of workers or individuals cooperate (or not); capital-labour,with reference to industrial relations in the strict sense. We are looking here at cooperation as influenced (i.e. favoured or inhibited) by normative systems rooted in local culture and institutions. In his recent book on collective bargaining, Elster (1989) gives some examples of this kind of norms. Norm of equality means equal pay for everyone, regardless of type of work. Norms of equity range from norms of proportionality of the form ‘to each according to his X’ to the norm of ‘equal pay for equal work’. Reference-level norms give a privileged position to the status quo. Norms of fair division are rules for sharing the surplus between workers and the firm (ibid.: 215-16). A first area to consider concerns training in general and/or specific professional skills, which industrial development and technological innovation have made increasingly important Table 1 Types of cooperation Collective Capital-capital Labour-labour Forms of cooperation Individual Training paid for and carried out jointly by enterprises and the state Solidaristic wage Noncooperation In-house training Poaching Efficiency wages Job evaluation Human relations Neo-taylorism policy Capital-labour Mutual recognition z 0 Joint Editors and Basil Blackwell Ltd 1993. zyxwvu zyxw zy Towards a regionalization of industrial relations 103 for enterprises. Each firm can resort to the market, poaching those professional figures most in demand by paying them more. The phenomenon of ‘poaching’ is widespread, especially in the USA and Britain. The ‘inflationary’ effects of these policies are well known, as well as the spread of a culture of opportunism and low-trust relations among partners. Alternatively, the firm can train ‘in-house’ and not consider the skills as a collective good. But in this case, the firm, because it is concerned about making an investment to train workers it is destined to lose (which often happens when the worker moves to another firm or goes freelance), will often underinvest in training. Lastly, when there is cooperation among firms, unions and the state, which will normally occur at local level, training is considered to all intents and purposes a collective good for which enterprises cooperate, and for which unions and the state assume responsibility. This can be seen as a case of fair division of costs to enhance general productivity. This has been the case through legislation in Sweden and in the German Lander, and more informally (but also formally through collective bargaining) in the industrial districts of the Third Italy. zyx Germany In Germany, the system of vocational training in small business (Handwerk)is a particularly good example. It is the product of cooperation among employers’ voluntary associations (guilds) at county level, regional chambers of commerce (which are public law institutions with compulsory membership) and trade unions (through collective bargaining). The result is that in Germany the training contracts for apprentices, particularly in small firms, increased up to 700,000 (1984 figures), with a yearly number of 200,000 artisanal apprentices entering the labour force at a high level of qualification. ‘To the extent that the ability of the owners and operators to discover market niches, use advanced technology, ensure high product quality, and manage an enterprise efficiently has contributed to the sector’s employment performance, the latter is at least in part explained by the Handwerk training system’ (Streeck, 1986). Since many of the apprentices later move to large firms, the artisanal training system benefits not only the small firms but also the economy at large. It should be noted that the incentives for small firms to train people are not only monetary, i.e. cheap labour. In fact, approximately 66% of the total costs of the dual vocational training system is assumed by business, with the federal government, Lander, chambers of commerce and labour authorities providing the rest. In this context, why should firms train more people than they directly need, as actually happens? The answer is that ‘in many sectors of the crafts, it is necessary for a single firm to train more workers than it needs simply because it must replace not only older workers who retire but also skilled workers who leave for industry, and because some of the apprentices switch to unrelated occupations or better paid unskilled and semiskilled activities in industry’ (Maier et a l . , 1986). Note that the German training system is voluntary in nature, as government does not interfere directly in firms’ choices; however, the credible threat of government intervention in the field (as it was envisaged in the 1970s and 1980s to avoid skilled labour bottlenecks) can be enough to move the private economy and the parties in the industrial relations system to provide such a collective good as training. Sometimes, apportionment of the costs of training through collective bargaining (having all firms in one sector pay contributions into a common fund for supporting firms which train apprentices) has taken place. Moreover, technological innovation diffusion and the trend among small firms to customize their products and services seem to be major reasons for sustaining the training effort (Maier et al., 1986). The apprentices have full status in the firms: they are considered to be employees, they are members of the work council and they participate in collective bargaining. 0 Joint Editors and Basil Blackwell Ltd 1993. z 104 zyxwvutsr Paolo Perulli Italy In Italy, training policies have been devolved to the regions since 1971. This process gave rise to a huge variation in regional performances in the field. In 1985, a national law transferred small firm training to the regions. The evolution of the vocational training system during the 1980s is the product of two distinct channels: the traditional apprenticeship and the new workhraining contracts. Apprenticeship decreased during the 1970s and 1980s, but still accounted for 554,000 people (1984 figures), 66% of them in craft firms. It should be noted that the regional distribution of apprenticeship is changing, for although most apprentices are employed in northern regions, the relative percentage has declined (from 63.7% in 1980 to 56.6% in 1984),while the percentage of apprentices in southern regions has increased (from 18.9% in 1980 to 23.5% in 1984). To understand these data, it is useful to remember that the law gives the firm the possibility of employing apprentices without increasing the size of the firm itself this is a strong incentive for a firm to hire apprentices while remaining an artisan firm, i.e. maintaining strong economic and legal (particularly in the field of labour law) incentives. This is why unions tried to regulate the phenomenon through collective bargaining. Recent legislation (1990) introduced for the first time a regulation of individual layoffs in artisan firms. Collective bargaining agreements to apply the new norms were signed in 1991 only in Emilia-Romagna, Piedmont and Lombardy . However, the status of the apprentice is still weaker than that of the firm’s employees. On the other hand, a new workhraining contracts system was introduced in 1984, extending to these new forms of hiring system the economic incentives of apprenticeship. Workhraining contracts have to be approved by a regional tripartite body including employers, unions and public officers (Commissioni Regionali per I ’Impiego).The number of employees involved, largely in small firms, grew to 152,000 (1985 figures); 54% of them were in northern regions, 35 % in central regions and only 11% in southern regions. After that there was a spectacular increase until 1990, with a peak of 400,000 new contracts, and a decrease in 1991 to 238,000, probably due to economic recession. The geographical distribution still favours the northern and central regions (80% of the total contracts). Small firms with less than 50 employees account for 71% of the contracts. This distribution is possibly due to the fact that these contracts are better suited to meet a strong labour demand, and require some organization capability, which small firms don’t have but their associations do, particularly in northern and central regions. The consequence, according to some observers, is that they have been used to stabilize ‘illegal’ or ‘black’ forms of employment existing in the small-firm industrial districts (Brusco and Villa, 1987). Workhaining contracts have permitted small firms to hire cheap labour and so adjust flexibly to economic cycles, while the real training effort sustained by the firms is a debatable question. It is possible to conclude that in Italy incentives to train people are particularly strong in terms of financial, fiscal and legal incentives. Employing cheap labour permitted through legislation is the major incentive, but regional patterns vary greatly and much scope is left to regional actors. As no uniform national system of vocational training (such as the German one) exists, there are dramatic regional differences in the quality, scope and sponsoring institutions. zyxw France In France, the training system is still highly centralized and fragmented among different public bodies, with a very weak relationship with enterprises and social actors. Particularly weak are the roles of local and regional actors, which are normally delegated by central employment agencies. This is why the creation of autonomous regional actors responsible for employment and training policies has been strongly recommended (Commissariat Gknkral du Plan, 1989). Moreover, small and medium enterprises are not the focus of a training z zyx 0 Joint Editors and Basil Blackwell Ltd 1993. zyxwvu zyxw zy zyx zyxwvu zyxw Towards a regionalization of industrial relations 105 system which is currently much more oriented towards the implementation of short-term employment contracts. Though they recently increased, apprentices number only 150,000, which means that only one-fifth of all skilled workers receive their training through apprenticeship. One of the reasons for the weak regional system of training in France may be the structure and context of collective bargaining. In France, the traditional structure of collective bargaining is based on the branche and takes place at national (40%),regional (20%) or departmental (40%) level. However, the regionaUdepartmenta1negotiations involve only one-third of the employees covered by collective bargaining and, in most cases, they are ‘frasred’ into national contracts. More often local or regional before the second world war, the branche negotiations have been more and more centralized since then. Moreover, the branche negotiation itself has been weakened over the last few years and a trend towards enterprise negotiations has taken place (Ministbre du Travail, 1989). This trend seems to exacerbate the large firmhmall firm gap. In fact, only 5 % of enterprise contracts are signed in small firms (i.e. with less than 50 employees) and the employees in small firms account for only 0.3%of total French employees covered by enterprise contracts. If we consider that almost 80% of the employees covered by enterprise contracts are in a firm with more than 500 employees, it is obvious that the current trend towards enterprise bargaining in France is a process which excludes the SMEs (Ministbre du Travail, 1989). It is interesting to note that this is also the complaint of the ‘modernist’ section of French employers: ‘It is possible that a contradictory evolution between large firms and SMEs will occur: large firms will maintain their structures of personnel collective representation, whereas the SMEs will be lacking in any collective representation’ (Enterprise et Progrks, 1988). Furthermore, training negotiation increase at enterprise level has been negligible: from 0.9%of enterprise contracts in 1984 to 3.9% in 1988. This weakness, which is a major obstacle to industrial modernization and labour mobility, gave rise to the strong suggestion made by the French Commission on Social Relations and Employment for the Plan 1989-92 of developing collective bargaining at the local level (nkggociation locale de branche). In 1988, a national agreement between employers and unions on ‘industrial modernization’ decided that branche negotiations should be concerned with training needs in SMEs, due to the introduction of new technologies through the development of solidarity among firms. Moving in the same direction, French modernist employers ask for cooperation at the local level: ‘The privileged terrain for this kind of action is the employment district. It is actually at local level that networking among enterprises of every size can be naturally developed without excessive administrative obstacles, as well as solidarity with any voluntary partner’ (Enterprise et Progrbs, 1988). From these different sides it is possible to conclude that, in the French case, the lack of a local culture of cooperation seems to be a major issue. Wage policy We turn now to the area of wage policy. A first alternative here is to fix a salary for each job position by means of rigid job descriptions. Besides being part of the traditional job evaluation policy of the taylorist/fordist era, this wage solution has recently been relaunched in line with the philosophies of individual bargaining in many countries: the USA in particular, but also the UK and France. This situation was helped by the breakdown of indexation systems or automatic increases and by concession bargaining by the unions. A consequence of these policies is certainly a strong social and labour polarization within the firm and at local level as well. Among the undesired effects is the reduction in the degree of internal job flexibility. Firms of this type resort to external flexibility, i.e. numerical, based on the freedom to change the number of employees through hiring and firing (but see also the remarks made by Piore, 1986). As an alternative solution, efficiency wage policies have been experimented with; that means salary levels higher than those on the market and based on the efficiency of 0 Joint Editors and Basil Blackwell Ltd 1993. z 106 zyxwvuts zyxwvuts Paolo Perulli the firm. The aim of these firms is to make labour productivity ‘endogenous’ and raise it above market levels, hence not transferable to other firms. The human relations systems that prevail in these firms tend to use incentives, not just economic but social and psychological (group work, autonomy etc.), which motivate work and link it to that particular enterprise. In this case, there is still that ‘paradox of the democracy of work’ described by Aoki (1988) with reference to Japan: that is, the surplus of productivity resulting from the workers’ performance concerns the firm alone, and the workers and firm management join together against competitors and outsiders. In some cases, these firms develop models of ‘full employment’ that include ‘a complex package of psychological contracts, training, mechanisms of participatory management, career incentives and production bonuses in which the expectation of a long-term commitment between workers and management is of primary importance’ (Doeringer, 1989). The third model, based on a solidaristic wage policy, tends to introduce criteria of equity in the treatment of labour, instead of divisions among different skill levels (as in the first model) and commitment to a single enterprise or ‘firm egoism’ (as in the second model). Although a solidaristic policy has been fully implemented only in Sweden, norms of equity came to the fore in countries such as Germany and in some Italian regions. The unit of analysis here is not the single firm but the local and regional economy as a whole. As an example related to the small firm sector, both in Germany and in the Third Italy, artisanal and small industrial firms are involved in a system of collective bargaining at regional level. In fact, given the smallness of these firms, collective bargaining at firm level cannot take place. The result is that small firm wages in these regions are considerably higher than in regions and systems where no contractual relations exist at regional level, and the small findlarge firm wage differentials are less dramatic than, for example, between large and small firms in the United Kingdom (Doran, 1984). This ensures that wage competition cannot be a disruptive practice among small firms and that large firms cannot easily exploit small firms as low cost subcontractors. In this model, the advantages of the workers being part of an overall regional economy are stressed, with positive effects, for instance on horizontal and ‘external’ (among different firms) labour mobility: a considerable source of flexibility when the demand is not steady and industrial readjustment is ongoing (Regini and Sabel, 1989). If we recall the points previously made about the French collective bargaining system, we can reach the same conclusion concerning the lack of local regulation at branchhegional level. Industrial relations As far as the area of industrial relations is concerned, the literature has developed three models which largely correspond to our typology presented in Table 1 . The first model is identified in neo-taylorist/fordist industrial systems (‘external flexibility associated with a direct hierarchical control’, according to Lipietz, 1990) or neo-liberalist (‘neo-liberalism permits the firms to resort to the external labour market, minimizing the workers’ status as “members” of the firm and relying on the contract as the main tool for defining the terms of the exchange between management and worker’, according to Streeck, 1985). The second model is based on the individual worker’s involvement (sometimes as a member of a group) in the performance of the firm. Worker’s status is basically permanent and the labour relationship is handled not only on the basis of the contract, but according to a set of policies for human resource management (which can be implemented with or without the unions, but in all cases on a strictly enterprise basis). Lastly, the third model, with its mutual recognition and joint participation, implies cooperative relations between enterprises and unions at local economy level. The step beyond taylorism and fordism here implies that ‘the involvement is bargained at the level of the entire society, with workers and entrepreneurs bargaining social guidelines and production organization at the regional or national level’ (Lipietz, 1990). zy 0 Joint Editors and Basil Blackwell Ltd 1993 Towards a regionalization of industrial relations zyxw zy 107 The extreme variability and segmentation of industrial relations and labour regulation regimes make any scheme or typology a temporary and in many respects an inadequate tool. The intention here is to offer an interpretation of processes of capital/labour relations as greatly influenced by institutional, social and cultural variables in the broad sense. The development of these variables has been studied more often at the single enterprise or national level and seldom at the regional level. This is not to say that we do really find ‘neo-taylorist/fordist’ regions, ‘individual worker’s involvement’ regions, ‘postfordist’ regions. We can rather expect to find some correlation between the industrial organization and the industrial relations which prevail in each regional economy. As Leborgne and Lipietz have pointed out, there are regions where forms of horizontal ‘quasi-integration’ among firms is correlated with workers’ involvement in the firms, and regions where vertical disintegration is developed in conjunction with neo-taylorism. However, we must expect to find rather mixed situations. As an example of the neo-taylorist/fordist pattern we can consider old industrial areas, where large firms based on mass production predominate. Vertical integration and hierarchical relations among firms are the rule here. The workforce is largely semiskilled or deskilled, given the mass production organization. Often decentralization takes place as a matter of production decentralization towards low labour cost regions. Strict contractual relations with the workforce into the firms are developed. As an example of the ‘individual worker’s involvement’ pattern we can consider some high-tech areas in the United States (the so-called Californian model). They are predominantly characterized by recent industrial development, small and medium-size high-tech firms, non-union environment and union-avoidance employment policies based on individualized practices. High-tech enterprises have often been thought of as a type of enterprise with very similar worker and, even more so, management behaviour, i.e. as non-union firms based on individualized personnel policies. On the other hand, the same type of high-tech enterprises behaves rather differently in the USA, where employment policies are developed ‘at will’ (high salaries in exchange for high performance but with no expectations of enterprise loyalty), or in a region like Baden-Wurttemberg, where policies of management by objectives, once limited to managers or white-collar workers, are extended to workers. Another contrasting phenomenon, outsourcing, has often been presented as an example of cooperative relations among enterprises on a regional scale, but it can have different meanings according to its context. Whereas in the Italian or German industrial districts (Sabel, 1988) or in some French regions (Lorenz, 1989) outsourcing means dense community and trust relations, in the USA it is used mainly by enterprises with no union affiliation (Harrison and Kelley, 1990). Finally, as examples of mutual recognition, the postfordist pattern is usually considered as a region where small and medium-size firms predominate, with dense networks of productive and social relations and a mostly qualified labour force. Large firms in these regions are interested in developing cooperative relations with small firms, as a source of skills and incremental innovations, although opportunistic behaviour is not excluded. Three case studies: Baden- Wiirttemberg, Emilia-Romagna and RhBne-Alpes The regions considered here, Baden-Wurttemberg, Emilia-Romagna and RhBne-Alpes, are among the European ‘success regions’ in terms of economic performance indicators. They are often considered in the literature on regional economies as examples of areas of ‘flexible specialization’. They also correspond to a current stereotype which underlines the ‘southern European’ regional dynamism as opposed to the deindustrialization of ‘northern European’ old industrial regions (Reclus/DATAR, 1989). However, comparative research would make clear that many differences exist, as well as much confusion among various forms of ‘old local systems’, industrial districts and new forms of flexibly specialized industrial organization (Ganne, 1991). zyxwv zyxwvu 0 Joint Editors and Basil Blackwell Ltd 1993. 108 zyxwvutsrq zyxwvut zyxwv zyxwvu Paolo Perulli Table 2 Three European ‘success regions’ Baden-W urttemberg Flexible specialization SMEs, particularly medium size + Emilia-Romagna SMEs, particularly small size Rh6ne-Alpes Some SME districts Large national groups Important large national and multinational groups Interest organization Strong at regional/ sector level Strong at regionall district level Weak at regional level Industrial relations Strong at regionall sector level Regional collective bargaining for SMEs Weak at regional branche level Some SMEs enterprise bargaining No SMEs bargaining Strong regional structure Weak regional structure Local systems of concertation Strategic public planning Human capital policies Dual system + Strong regional structure Key actors Network of public/ private bodies In particular, industrial structure varies among these regions more than the ‘flexible specialization’ literature has admitted, and much more so the organization structure and institutional relationships. This is particularly true for industrial relations. The EmiliaRomagna case seems to be characterized by a regional system of industrial relations which is unknown in the RhGne-Alpes case, whereas in Baden-Wiirttemberg the regional variations in industrial relations are less pronounced if compared to the national German system. The three regions have strong industrial structures and an important share of total employment is in manufacturing. The relative share is 36.7% in RhGne-Alpes, 39.1 % in Emilia-Romagna and 47.9% in Baden-Wurttemberg (1985 figures). In other terms a process of deindustrialization has not characterized these regions during the 1980s. It should be noted, however, that this is particularly true for Baden-Wurttemberg, and at a lesser extent for RhGne-Alpes, with Emilia-Romagna in an intermediate position. As far as the dimensional structure of the industrial enterprises is considered, small and medium-size firms are highly represented, particularly in Baden-Wurttemberg and Emilia-Romagna. In Baden- Wiirttemberg, 50.6%of the workforce is employed in firms with less than 500 employees. Note that this percentage is 47.4 in Germany and only 33.2 and 40.6 in the northern old industrial Lander of Bremen and Hamburg. The mediumsize firms are particularly strong in Baden-Wurttemberg. Looking at firms with over 20 employees, only 8.9% of total employment is in firms with 20-49 employees; 41.7% is in firms with 50-499 employees; and 49.4%is in firms with more than 500 employees. Medium-size firms are not only subcontractors but also autonomous market agents. In mechanics and optics, medium-size firms (100-499 employees) account for 43% of the total export of their sectors. In chemicals, electro-technics and plastics, the export percentage of medium-size firms is between 20 and 30%of the total export of their sectors. However, the Land, and especially the Stuttgart region, also contains some of the most important German and multinational enterprises such as Daimler Benz, Bosch, IBM, Porsche, SEA, Kodak and Hewlett-Packard. zyxw zyxwvuts zyxw 0 Joint Editors and Basil Blackwell Ltd 1993. zyxwvu zyxw zy zy Towards a regionalization of industrial relations 109 The technological excellence of Baden-Wurttemberg is sustained by a strong R & D base: the relative expenditure is 3.5% of the Gross Regional Product, a percentage which is the highest among the industrial nations. R & D infrastructure is particularly decentralized through a regional system of innovation and technology diffusion centres. The efficiency of this system for SME is confirmed by data concerning the technology transfer: over 18,000contacts were made with firms in 1988,90% of which referred to firms of less than 200 employees. Vocational training policies are particularly developed in the Land. Out of a total $00 are trained in the vocational school system and 89,000 employment of 1,700,000,271 of those are in small and artisan firms (1989figures). As a consequence, the percentage of skilled labour force in manufacturing industry grew from 34% in 1970 to 48% in 1990. Among the components of the regional structure is the collective bargaining and industrial relations system. Baden-Wurttemberg is characterized by pilot agreements in the collective bargaining field, particularly in skill and training matters. Unpublished research conducted by C. Deutschmann of Tubingen University has shown that ‘soft’ factors in labour relations, such as personal networks, fringe benefits and trust relations, are highly developed amongst the Baden-Wiirttemberg firms. In Emiliu-Romagnu the small firms’ role is even higher, given that a large-firm sector is very limited. Not only is a substantial share of the total workforce employed in small and medium-size firms in traditional sectors, but the small high-tech entrepreneurship is also diffused. Out of a total working population of 1.5 million, 55,000are entrepreneurs and 376,000 are independent or self-employed. In terms of productive specialization, not only is EmiliaRomagna a leader in some consumer goods sectors, but it also ranks fourth (after Lombardy, Piedmont and Lazio) in high-tech sectors such as machine tools, chemical-pharmaceutics and industrial automation. Emilia-Romagna also ranks fourth among the Italian regions as far as R & D expenditures as a percentage of Gross National Product are concerned (Malerba and Onida, 1990). In other terms this region seems to be the first, among the regions of the Third Italy, to have successfully advanced towards a more innovative neo-industrial pattern. As far as vocational training is concerned, 35,000are trainees in the provincial schools (1987 figures) and 17,000 are on-the-job trainees through workkraining contracts. Apprentices in Emilia-Romagna number 52,000,10% of the total Italian apprenticeship; 50% of the male apprentices in the region are in engineering, 60%of the female apprentices are in the textile and tertiary sectors. Collective bargaining plays a major role in skill and on-the-job training promotion. Research conducted by ISFOL (the Italian Ministry of Labour Agency) during the mid-1980s showed that the propensity to negotiate vocational training at enterprise level was much higher in Emilia-Romagna provinces (in Modena 47% and in Ferrara 40% of the sampled enterprises had negotiated vocational training) than in any other Italian region (average propensity 17%). The role of regional collective bargaining is particularly important for small and artisan firms. Emilia-Romagna is the only Italian region where regional contracts for small-firm sectors have been applied since the 1970s. In 1992, a regional contract involving the small-firm sectors was signed for the fourth time, whereas other Italian smallfirm regions like Veneto have signed it only for the first time. This has led to a certain wage differential among small firms in different regions, but more generally to a ‘normative practice’ concerning wage equality and work rules in small firms, in which Emilia-Romagna seems typical of other Italian small-firm regions. In the RhBne-Alpes case, the regional economic structure is traditionally more characterized by large firms (40% of the regional employment). Chemicals, electric plants, electrometallurgy and similar sectors were created during the second industrial revolution and still form the base of the industrial economy. Moreover, the role of national and multinational groups in the Lyon economy is high: 70% of total employees work in firms with their z zyxwvu 0 Joint Editors and Basil Blackwell Ltd 1993 110 zyxwvutsrq zy zy zy Paolo Perulli headquarters in Paris, and foreign firms located in the Lyon region number 103, with 28,000 employees - 15% of total industrial employment. However, the crisis of the 1970s has provoked a new interest in small-firm districts more decentralized in the RhBne-Alpes (Oyonnax in plastics, Val d’Arve in metal turning), which have a long historical tradition. In the meantime, attempts have been made to turn the Lyon economy towards new high-tech development (as in biotechnology) and four new technopoles in the Lyon area have been designed. Partly due to the recent dynamism of local government and the economic elite in promoting Lyon as an ‘international city’, a trend towards the location of high-tech firms in the Lyon region has been observed. Almost 1500 firms classified as ‘innovative’ (in sectors such as industrial machinery, pharmaceutics, machine tools, software etc.) employ 40,000 employees, that is, 27% of total industrial employment in the Lyon region. However, the small firm development of the region is not its central feature. Another difference between RhBne-Alpes and regions like Emilia-Romagna and BadenWurttemberg is the weakness of regional actors. The collective economic actors, such as employers and union organizations, are very weak, particularly at department and regional level. Only recently an attempt to develop a social dialogue, for example at the Economic and Social Regional Council level, has been developed. A regional policy to develop human resources and training is not established, although the RhBne-Alpes region has the reputation of having a well-developed system of general education and a diversified vocational training system. In ‘postfordist’ regions, local and regional institutions are supposed to play a significant role in strengthening their regional and local economies in fields like training, industrial infrastructures, financial support, real services etc. This is a very important aspect of the typology. In fact, both in the neo-taylorist/fordist and in the ‘individual worker’s involvement’ patterns, local and regional institutions play a very limited role, in the first case because of the predominance of mass-production large firms which tend to internalize these functions, and in the second because local institutions are much weaker than market forces and each firm is interested in its internal labour market. In other words, in the first case ‘hierarchy’ predominates, in the second ‘market’ is the ruling institution. In both cases, little space is left to intermediate forms of regulation in which local and regional actors can flourish. This is the case in the third pattern developed here. Regions such as Baden-Wurttemberg, Emilia-Romagna and, to a lesser extent, Rh8ne-Alpes, have developed a set of industrial policies at regional or local level to promote their industrial structure. These policies are often part of a larger cooperation among local government, industrialists’ associations and unions, and they are at least partly related to a good industrial relations climate. This is particularly true in the German and Italian cases, where the social actors have strengthened their links and cooperate in a number of local institutions. This is less true in the French case, where the interests’ representative system is weaker at regional level and still corresponds to the previous industrial phase, with each local interest responding to its national centralized structure. Among other factors, these regions seem better suited to utilize the trend taking place in Europe towards the weakening of the nation-states compared to the increased role, through alliances and new decision-making networks, of regional actors (Schmitter, 1990). This trend (at least functionally) corresponds either to a demand for autonomy on the part of local actors or to an interest on the part of the European Community in forming alliances with local actors in order to avoid the rigidities of national institutions. Some regions, particularly if their productive structure is internationally competitive and their social actors tend to cooperate, can try to develop a regional economic policy which will increase their competitive success. zyxwvu 0 Joint Editors and Basil Blackwell Ltd 1993. zyxw zy zyxwvutsrq zyxw Towards a regionalization of industrial relations 111 Conclusions: towards new conflicts? A trend towards the regionalization of industrial relations has been observed in different national contexts during the last two decades. However strong are the unifying forces at work, like the new network organization of firms - in which ‘local’ and ‘global’ are no longer in conflict - and local actors are motivated to play a more direct regulatory role, substantial differences emerge even among the so-called ‘postfordist’ regions. Looking at sorriz common features of the postfordist models of flexible accumulation, it must be emphasized that :hey are differently shaped by institutional context and structural constraints. First of all, consider industrial organization as it concerns large firm-SME relationships, and contracting relationships in particular. Clusters of firms where spatial proximity is a condition for participation in changing product design and manufacturing are a common example here. However, contractual relationships may be different if (a) there are large firms with their headquarters and/or other production networks external to the region; (b) there are large firms strongly linked to the regional networks; (c) there are mostly non-hierarchical relationships among SMEs; (d) there are various combinations of the three. It should be noted that as far as a regional production apparatus is considered, it is normally a complex ‘superset’ of production systems (Storper and Harrison, 1991). We have found such a difference in contrasting RhBne-Alpes, Baden-Wurttemberg and Emilia-Romagna. A second issue concerns the labour skill base of the region. The literature has often underlined the richness of informal sources of skills through on-the-job training as a common feature of flexible specialization regions (Piore and Sabel, 1984: 274ff.). However, these sources are more and more weakened by social factors (mobility, changes in family- and firm-based ties, cultural propensity towards industrial jobs etc.) and technological/ organizational factors as well (the knowledge base required for handling new technological systems, designing or marketing new products, is somewhat different from that needed for mastering craft tasks). The traditional sources need to be strengthened by vocational and school training, which is provided in different ways by regionalhational training systems. Small-firm districts need to develop new organizational forms for providing such goods (Trigilia, 1991). Thirdly, work attitude and commitment to ‘sustaining product quality are often considered as part of postfordist models. However, they vary greatly as far as local institutions, traditions and practices of labour management relations go. Fourthly, wage-setting has to combine incentives that reward individual initiative and norms of equity which preserve collective identity and cooperation. But to define this compromise, legal protections, welfare and collective bargaining structures are relevant. If such structures and institutions are well developed, as in Germany and to a different extent in the Third Italy, the regional economy will be different than in the USA, UK, France and Japan, where a strong dualism between large and small firms is the rule. Finally, we have considered trends towards the creation of local/regional levels of industrial relations regulation. This same process may also present another aspect, that of deregulation. A significant example is that of social dumping in Europe, defined as ‘recourse to working conditions and social norms which are substandard with respect to the productivity levels that the economy could normally justify’ (Observatoire social europten, 1990). In other words, free capital and labour mobility in Europe could enhance regional differences in labour costs, social security programmes and collective bargaining. In their decisions to localize their investments, enterprises (especially the multinationals) would be attracted by the existence of different regional regulation regimes, not only for capital but for labour too. In fact, certain firms might be attracted by regions where cooperation between capital and labour has produced the accumulation of skills, outsourcing networks, specialized resources, social consensus - though most probably this argument 0 Joint Editors and Basil Blackwell Ltd 1993 z 112 zyxwvutsrq Paolo Perulli zyxwvu concerns only those investments which use qualified labour, technicians, researchers and managerial and professional elites sensitive to certain social and environmental conditions. It is interesting to note that both in Baden-Wurttemberg and in Rh8ne-Alpes a substantial amount of new industrial employment has been created in the last years by foreign companies (often multinationals), particularly in high-tech sectors. These companies have localized their European headquarters and their R & D plants in these regions. However, other firms (or the same firms with regard to their production plant location) consider it more worth while to settle in areas with low labour costs, because there is flexible legislation o r substandard union agreements. Considering the growing importance for firms of indirect labour costs (insurance, welfare etc.), what has been defined as ‘regime shopping’ could prove to be an occasion for a widening of regional differences, especially in southern Europe. The threat of delocalization o r further deindustrialization of areas considered ‘socially rigid’, the creation of ‘free enterprise zones’ o r enclaves (at the periphery of Europe o r even outside the borders), in which immigrant workers carry on activities under the same conditions as in their country of origin, could be the consequence. zyxw zyxwvutsr zyxwvu zyx Paolo Perulli, DAEST, Istituto universitario di archittettura di Venezia, Ca’tron, S. Croce 1957, 30125 Venezia, Italy References Aoki, M. (1988) A new paradigm of work organization and coordination: lessonsffom the Japanese experience. UNU/WIDER Working Papers. Bagnasco, A. (1988) La costruzione sociale del mercato. I1 Mulino, Bologna. Bakke, E.W. (1946) Mutual survival: the goal of unions and management. Harper and Row, New York. Becattini, G. (1987) Mercato e forze locali: il distretto industriale. I1 Mulino, Bologna. Bonazzi, G. (1989) Storia del pensiero organiuativo. F. Angeli, Milano. Brusco, S. and P. Villa (1987) Flessibilith del lavoro, segmentazione e livelli di occupazione. Politica Economica 2 . Clegg, H . (1975) Pluralism in industrial relations. Italian translation in G.P. Cella (ed.), La contesa industriale, Edizioni Lavoro, Roma, 1980. Commissariat GCnCral du Plan (1989) Rapport de la commission relations sociales et emploi. La Documentation FranGaise, Paris. Doeringer, P. (1989) US macroeconomic pelformance and labor eficiency: a structural managerial interpretation. Working Paper, Department of Economics, Boston University, Boston, MA. Doran, A. (1984) Craft enterprises in Britain and Germany: a sectoral study. Anglo-German Foundation for the Study of Industrial Society, London. Elster, J. (1989) R e cement of society. Cambridge University Press, Cambridge. Entreprise et Progrks (1988) Enjeux d’une politique sociale. Liaison sociales, Documents, 74. Esser, J. and J. Hirsch (1989) The crisis of fordism and the dimensions of ‘postfordist’ regional and urban structure. International Journal of Urban and Regional Research 13.3, 417-37. Fox, A. (1973) A social critique of pluralist ideology. Italian translation in G.P. Cella (ed.), La contesa industriale, Edizioni Lavoro, Roma, 1980. Ganne, B. (1991) Sistemi industriali locali: che cosa insegna una comparazione tra Francia e Italia. Stato e Mercato 3 1. Harrison, B. and M. Kelly (1990) Outsourcing and the search forflexibiliry. Paper prepared for the conference on Pathways to Industrialization and Regional Development in the 1990s, University of California at Los Angeles, 14-18 March. ISFOL (1988) La formazione nella contrattazione aziendale. F. Angeli, Milano. Kerr, C. (1954) Industrial relations and the liberal pluralist. Proceedings of the Seventh Annual Meeting of the IRRA, 28-30 December. Lipietz, A. (1984) I1 monetarism0 centrale contro il fordismo periferico: banche e nuovi paesi in via di industrializzazione. Stato e Mercato 11. zyxwv 0 Joint Editors and Basil Blackwell Ltd 1993. zyxw zy zyxwvu zyxwv zyxwvutsr zyxwv Towards a regionalization of industrial relations 113 Lipietz, A. (1990) Capital/labour relations at the dawn of the XXIst century. Working Paper, CEPREMAP, Paris. Lorenz, E.H. (1989) Nk amici nk estranei: reti informali di subappalto nell’industria francese. In D. Gambetta (ed.), Le strategie della jiducia, Einaudi, Torino. Maier, H.E. et al. (1986) The role of smalljirrns and crafr businesses in the professional integration of youth: a comparison of the Federal Republic of Germany, France, Italy and Ireland. WZB, IIMJLMP discussion paper. Malerba, F. and F. Onida (1990) La ricerca scientijica. Sipi Confindustria, Roma. Marshall, T.H. (1976) Cittudinanza e classe sociale. UTET, Torino. Ministkre du Travail (1989) Bilan 1988de la ntgociation collective. La Documentation FranGaise, Paris. Observatoire social europCen (1990) Dumping social: un concept opkrationnel? 56, Bruxelles. Olson, M. (1983) La logica dell’azione collettiva. Feltrinelli, Milano. Pichierri, A. (1986) I1 declino industriale. Rosenberg & Sellier, Torino. Piore, M.J. (1986) Perspectives on labor market flexibility. Industrial Relations 25.2. __ and C.F. Sabel (1984) The second industrial divide. Basic Books, New York. RecludDATAR (1989) Les villes ‘europtennes ’. La Documentation FranGaise, Paris. Regini, M. and C.F. Sabel (1989) Strategie di riaggiustamento industriale. I1 Mulino, Bologna. Sabel, C.F. (1988) The reemergence of regional economies. Paper, Massachusetts Institute of Technology, March. Schmitter, P. (1990) The European Community as an emergent and novel form of political domination. Paper, Stanford University, April. Scott, A.J. (1989) Flexible production systems and regional development. Journal of Urban and Regional Research 12.2, 171-86. Shimada, H. (1980) The Japanese employment system. The Japan Institute of Labour, Tokyo. Storper, M. and B. Harrison (1991) Flexibility, hierarchy and regional development: the changing structure of industrial production systems and their forms of governance in the 1990s. Research Policy 20. Streeck, W. (1985) The management of uncertainty and the uncertainty of management: employers, labour relations and industrial adjustment in the crisis. Paper prepared for the conference ‘Stato, politica economica e relazioni industriali in Europa: quale destino?’, 19-2 1 September, Brescia. __ (1986) f i e territorial organization of interests and the logic of associative action: the case of artisanal interest organizations in West Germany. WZB discussion paper, Berlin, October. Trigilia, C. (1991) Tre equivoci della piccola impresa. In Regione Emilia-RomagnaICNEL, La piccola impresa, Jovene, Napoli. 0 Joint Editors and Basil Blackwell Ltd 1993.