The Corporatist Ideal-Type and Political Exchange

zyxwvut zyxwvuts zyxwvuts zyxwv Political Studies (1 W2),XL,255-272 The Corporatist Ideal-Type and Political Exchange zyxw zyxw MARTIN J. BULL* University of Salford zyxwvut The fundamental reason for corporatism’s persistence in political science debates is its failure to respond to the demands of political theory and present a convincing idealtype to capture the relationship between interest groups and the state. Corporatist writers have misused ideal-typea and the most refined example to date of the corporatist ideal-type (Cawson’s) is structurally flawed. There arc more profound problems than this, however, in the construction of a corporatist ideal-type because of the nature of the dynamics at the heart of the corporatist process: political exchange. Every change of paradigm begins with a new exaggeration.‘ In a recent book Schmitter and Streeck note that ‘“corporatism” seems unwilling to go away - either as a subject of inquiry or as a practice of governance’.’ The former is an irrefutable claim,3the latter, however, appears to be a startling one. That ‘corporatism’as ‘a practice of governance’ has been suffering from a general decline in Western Europe - if indeed it ever existed - is not only clear to the casual observer but has been documented by specialists in the field.’ The claim can only be understood, therefore, in relation to corporatism’s persistence as a subject of inquiry. This is an example of one of political science’s more curious paradoxes: rather than making a practice of governance a subject of inquiry, the zyxwv The author would like to thank Desmond King, Noel OSuHivan, George Tsebelis and one of the journal’s anonymous reviewers for their comments and suggestions on an earlier draft of this article. They bear no responsibility for the final text and its arguments. ’ K. von Beyme, ‘Neo-corporatism: a new nut in an old shell?, International Political Science Review, 4 (2), p. 191. * P. Schmitter and W. Streeck, ‘Foreword’, in I. Scholten (ed.). Political Stability and NeoCorporatism: Corporatist Integration and Societal Cleavages in Western Europe (London, Sage, 1987), p. vii. For a recent example of its persistence as a subject of inquiry, see the exchange between Cox and Cawson in this journal: A. Cox, ‘The old and new testaments ofcorporatism: is it a political form of the state or a method of policy-making?, Political Studies, XXXVI (1988), 294308; Alan Cawson, ‘In defence of the new testament: a reply to Andrew Cox, “The old and new testaments of corporatism”’, Political Studies, XXXVI (1988), 310-15. ‘ See, for example, M. Cameri and C. Donolo, ‘Oltre I’oriaonte neo-corporatista: Alcuni scenari sul futuro politico del sindacato’, Stato e Mercato, 9 (1983), 475-503 and M. Regini, ‘Political bargaining in Western Europe during the economic crisis of the 198Os’, in 0. Jacobi, B. Jessop, H. Kastendiek and M. Regini (eds), Economic Crisis, the Trade Unions and the State (London, Croom Helm, 1986), Ch.3. Moreover, Schmitter himselfhas more recently undermined his own claim in ‘Corporatism is dead! Long live corporatism!’ (The Andrew Shonfield lectures, IV), Government and Opposition, 1: 24 (Winter 1989), 54-73. ’ 0032-32 17/92/02/0255-18/$03.00 zyxwv 0 1992 Political Studies 256 zyxw zyxwvutsrq zyxw zyxw The Corporatirt I&al-Type and Political Exchange discipline at times seems to prefer creating - or at least maintaining - a ‘practice of governance’from a subject of inquiry. This means that one has to explain why it is that corporatism has been kept alive as a subject of inquiry. The explanation is in corporatism’s failure to respond to the demands of political theory and present a convincing ideal-type to capture the relationship between interest groups and the state. This failure is demonstrated by two lines of analysis. The first investigates many corporatist writers’ misuse of the idea of an ideal-type as a Weberian instrument of analysis. The second outlines and then investigates the most refined example, to date, of the corporatist ideal-type, that of Cawson. It is argued that this ideal-type is faulty in its structural determinants. It is also argued, however, that any corporatist ideal-type based on structural factors is likely to be limited in value because of the nature of the dynamics at the heart of the corporatist process, political exchange. zyxwv zyxw zyxwv Corporatism and the Methodology of Ideal-Types The 1988 debate between Cox and Cawson highlights the importance of a correct use of ideal-types. Cox argues that corporatist writers have confused their original corporatist ideal-type with a description of policy-making in modem states which does not require a different concept to pluralism. He argues that the ideal-type can nonetheless retain its value as a political form of the state if shorn of its state and societal variants; or, more accurately, if shorn of its societal variant (which Cox argues is indistinguishable from pluralism), leaving the state variant as the corporatist ideai-type.’ In his reply, Cawson, in addition to challenging the nature of Cox’s corporatist ideal-type (which will be dealt with later), argues that Cox fails to recognize the logic of analysis through ideal-types. Ideal-types are ‘purified’abstractions, often from historical cases, rarely found in pure form in empirical cases. What is important in comparative analysis, therefore, is the way in which different types are combined in any given empirical case, or the extent to which they approximate particular ideal-types! There is little doubt of the validity of this point. Ideal-types are ‘ideal’, in Scruton’s words, ‘in the sense of being construed entirely according to theoretical laws that explain them, and not according to the actual world. At the same time the models are chosen so as to correspond as nearly as possible to the actual world.” Yet it has been this need to strike a balance between ‘theoretical laws’ and the ‘actual world’ which has bred a persistent dilemma for political science:whether or not to revise an ideal-type to account for changing reality or how far to respond to changing reality in the search for a theoretically useful ideal-type. ‘Totalitarianism’ as an ideal-type, for example, originated in the dilemma of what to do with an expression formulated to describe the Soviet Union under Stalin but failing adequately to describe the Soviet Union in the post-Stalin period. This dilemma was first noted by Alex Inkeles and has been neatly summarized by Archie Brown: Either we keep revising the ‘model’ in accordance with changing reality within the Soviet Union, so that the Soviet Union remains by definition ‘totalitarian’ or we stick with a model (or, more precisely, an ideal-type) and ’ Cox, ‘The old and new testaments of corporatism’. Cawson, ‘In defence of the new testament’, p. 31 1. ’ R. Srmton. A Dictionory of Political Thoughr (London, Macmillan, 1982). p. 21 1. MARTINJ. BULL zyxw zy 257 try to ascertain how far the Soviet Union has deviated from that model (or ideal-type).’ The dilemma was resolved in both ways. Some authors persisted with totalitarian interpretations of the Soviet Union after Stalin, but the validity of these interpretations rested on the assumption that a country could be ‘totalitarian’ without displaying all of the six features outlined as essential by Friedrich and Brzezinski! Other authors argued that totalitarianism no longer provided an accurate characterization of the Soviet Union and other communist states. New interpretations of the Soviet system -including, for example, those of corporatism and ‘institutional pluralism’ - were developed, therefore, to demonstrate how far the Soviet Union had deviated from the totalitarian idealtype. Of the two options, Brown favours the latter. The former he describes as ‘more misleading than helpful’.’’ Revising the model to account for changing reality leads to ‘conceptual stretching’ or to the use of confusing labels which, in this example, attempt to maintain the totalitarian nature of the Soviet Union at the same time as taking out one of totalitarianism’s essential features, like ‘totalitarianism without terror’. Corporatism has followed a similar misleading course. The main cause of the ‘conceptual stretching’ from which it has suffered has been through adjustments ofthe concept by numerous writers to account for changing reality, the effect of which has recently been described as ‘leaving the rest of us anxiously wondering whether or not to plunge in, like holiday-makers facing a grey and uninviting sea’.’’Indeed, it is not so much that the debate has reached baffling proportions; rather that it has reached the point where it is difficult to deny the existence of putative ‘corporatist tendencies’ without retreating to the accusation of concept inflation or challenging the specific point of reference the author is using. Even the idea, apparently fundamental to corporatism, of the existence and intervention of the state has been ignored through the identification ofcorporatism as constituting an agreement between labour and business in the absence of the state.” To understand the ‘grey and uninviting sea’ and the misuse of ideal-types,it is necessary to identify the original cause of the confusion. This lies in the dispute zyxw zyx zy zyxwvutsrqpo zyxw A. Brown, ‘Pluralism, power and the Soviet political system: a comparative perspective’, in S. G. Solomon (ed.), Pluralism in the Soviet Union: Essays in Honour of H. Gordon Skilling (London, Macmillan, 1983). pp. 65-6. The original work by A. lnkeles is ‘Models and issues in the analysis of Soviet society’, Survey, 60 (July 1966). ’ C. J. Friedrich and 2. K. Brzczinski, Totalitarian Dictatorship and Aurocracy (Cambridge, MA,Harvard University Press, 1956). lo Brown, ‘Pluralism, power and the Soviet political system’, p. 66. ” A. Weale, ‘Monopoly groups’, Times Higher Education Supplement (24 Oct. 1986). l 2 See,for example, A. Martinelli, ‘Organised business and Italian politics: Confindustria and the Christian Democrats in the post-war pcriod‘, West European Politics, 2:3 (1979), p. 85 and A. Martinelli and T. Treu, ‘Employers’ associations in Italy’, in J. P. Windmuller and A. Gladstone (eds), Employers’ Associations and Industrial Relations: a Comparative Study (Oxford, Oxford University Press, 1984), p. 287. The authors are referring to the Italian case. J. Goetschy, ‘The neocorporatist issue in France’, in I. SchoIten (ed.), Political Stability and Neo-Corporarism: Corporatist Integration and Societal Cleavages in Western Europe (London, Sage, 1987). p. 185, follows a similarly confusing line in analysing France. The problems of concept infiation and terminological ambiguity are something of which corporatist writers themselves have become increasingly aware. See, for example, P. C. Schmitter, ‘Democratic theory and neocorporatist practice’, in W. Maihofer (ed.),Noi si mura: Selected Working Papers of the European University Institute (Florence, European University Institute, 1986), p. 165. 258 zyxwvutsrqp zyxw zy The Corporatist Ideal- Type and Political Exchange between those who viewed corporatism as a distinctive mode of interest intermediation and those who viewed corporatism as a distinctive mode of policy formation.” Schmitter, in 1982, attempted to resolve this confusion of the existence of ‘two corporatisms’ (which he labelled ‘corporatism-1’ and ‘corporatism-2’). He suggested that the expression corporatism be reserved for the first paradigm (interest intermediation), with its polar opposite being pluralism; while the second paradigm should be termed concertation, with its polar opposite being pressure.“ Cawson subsequently rejected this idea and argued - as will be developed later - that corporatism can only be viewed as a combination of ‘corporatism’ and ‘concertation’.’’ Despite the clear-cut demarcation of corporatism in both Schmitter’s and Cawson’s definitions, however, corporatist writers have not been averse to maintaining the term ‘corporatism’ when writing onfy about policy formation; that is, in the absence of the distinctive mode of interest representation deemed essential by both Schmitter and Cawson.16 Indeed, most corporatist writing has come to be exclusively concerned with this aspect, with a consequent tendency to search for a corporatist ideal-type which Jits current reality. A recent example is in the third volume of Schmitter’s edited series on neocorporatism where Goetschy, having adopted Schmitter’s definition as ‘an ideal-typical definition in the Weberian sense’, then suggests that zyx zyx zyxwvu zyxwvutsr on the basis of our analysis of French neo-corporatist trencfs . . . one should seek for a definition of a less institutional nature which could take into account to a greater extent more flexible, fluctuating and partial forms of neocorporatism.” This simply makes explicit an approach which has been implicit in much of the writing on corporatism since the mid-1970s. and which has given the concept an elastic quality. The original dispute, then, was resolved by everybody having their way. The confusion was completed with the concomitant expansion of the concept with respect to its possible location at different levels of the political system. What began as ‘corporatism’ became simply one possible level of analysis of the political system (‘macro-corporatism’) to which were added two further levels, ‘meso-corporatism’ and ‘rnicro-corporatism’.’’ In short, even if Cox’s I’ This is not to overlook two other approaches to corporatism associated with Winkler (a system of political economy different from capitalism and socialism) and Jessop (a form of state within capitalist society). While these have undeniably added to the diversity ofcorporatism, they are not crucial to explaining the misuse of ideal-types in what has been the central area of corporatist debate. For examples of the other approaches, see J. Winkler, ‘Corporatism’, European Journal of Sociology, 17 (l976), 100-36 and B.Jessop, ‘Corporatism, parliamentarism and social democracy’, in P.C. Schmitter and G. Lehmbruch (eds), Trends Towardr Corporatist Intermediation (London, Sage, 1979). “ P. Schmitter, ‘Reflections on where the theory of neo-corporatism has gone and where the praxis of neocorporatism may be going’, in G. Lehmbruch and P. Schmitter (eds), Patterns of Corporatist Policy-making (London, Sage, 1982). pp. 259-90. I’ A. Cawson, Corporatism and Political Theory (Oxford, Blackwell, 1986), Ch. 4. I* This has been done, moreover, even under the general editorship of Schrnitter himself. See, for example, I. Scholten (ed.),Political Stability and Neo-Corporatism: Corporatisr Inregration and Societal Cleavages in Western Europe (London, Sage, 1987), which is Volume 3 of Schmitter’s edited Sage Series in Neo-corporatism. I’ J. Goetschy, ‘The neo-corporatist issue in France’, p. 193 (emphasis added). I’ A. Cawson (ed.), Organised Interests and the State: Studies in Meso-Corporatism (London, Sage, 1985). Cawson later viewed these as having created a more sophisticated distinction between different types of corporatism (‘In defence of the new testament’, p. 314). This might have been the case had the original product not been built on shifting sands. A parallel development to the zy MARTINJ. BULL 259 own use of ideal-types is open to question, the writing on corporatism has hardly been more rigorous and the methodological thrust of his argument remains valid: the tendency to adjust the definition of corporatism with empirical observation, while ensuring the continued discovery of corporatist ‘tendencies’, has hindered the search for a consistent ideal-type. Why has this occurred? Two closely interrelated reasons can be given. First, disagreements about a nascent concept are inevitable. Cawson, for example, argues that disputes over concepts such as ‘democracy’ and ‘power’ have not detracted from their significance.” One might add that it would be unreasonable to expect all corporatist writers to reason together. They do not (and can hardly be expected to) constitute a unified ‘school of thought’. As indicated above, however, there is a difference between definitional disputes in the search for a consistent ideal-type and the harnessing of the search for an ideal-type to empirical analysis. The major characteristic of corporatist writing has been the latter. The second reason is that corporatism has been unduly affected by its chief competitor, pluralism, and its history. Indeed, it might even be suggested that corporatism has had little option but to emulate the history of pluralism in order to survive. It is perhaps surprising that Cawson did not choose ‘pluralism’ (instead of ‘democracy’ and ‘power’) as an example of how definitional disagreements need not reduce a concept’s significance, for this has been one of the most ‘under-explicit’, ‘evolving’, ‘mutating’ and ‘inconsistent’ of theories, to the point where it has recently been described as ‘no more than an anti-theory’.M This fluctuating and evolutionary nature has been pluralism’s greatest strength. It has allowed its supporters to draw on a great wealth of ‘pluralist’ literature to argue that all recent empirical developments can be accounted for in pluralist theory. But it could be argued that this has been the result of precisely the same process as that outlined for corporatism above: namely, that writers calling themselves pluralists have been analysing political systems and developing ‘pluralist’ theories while assuming that what they have been observing can still be described as pluralism. The pluralist ideal-type has always managed to ‘fit reality’ and it has never, therefore, been subject to the rigorous analysis of ideal-types. Corporatism, however, is apparently not to be granted such rights by pluralists: there is only room for one ‘anti-theory’ and pluralism got there first. As Sartori has pointed out, ‘there is no end to pluralism, for we are never told what is nonpluralism’.*’ Two wamng camps, then, continually shift the conceptual and terminological ground in an attempt to claim changes in the nature of policymaking as compatible with their theoretical constructs. Seen from this perspective, it is evident that two related but what should be clearly distinct debates have become confused: first, the viability of corporatism as a theoretical ideal-type; secondiy, the extent to which empirical evidence of changes in state-group relations can be accounted for in pluralist theory. Cawson’s claim that Cox’s definition of pluralism is little more than a description of existing practice in liberal democracies is, ironically, a perceptive illustration zy z zyx zy zyxwvu zyx expansion of corporatismwas a form of ‘analytical retreat’ from some of the more grandiose claims of the 1970s about corporatism, but this point is not crucial to the argument developed here. Cawson, ‘In defence of the new testament’, p. 313. G . Jordan, ‘The pluralism of pluralism: an anti-theory?’, Political Studies, XXXVIII (1990), p. 286. ’I G . Sartori, ‘Concept misinformation in comparative politics’, American Political Science Review, 4: LXIV (1970), pp. 1050-1. 260 zyxw zyx zyxwvutsr The Corporatist Ideal-Type and Political Exchange of the whole pluralist-corporatist debate because precisely the same charge might be laid at the door of much of the work on corporatism.” The pluralistcorporatist clash over empirical analysis has overshadowed and obstructed the progress of a fundamental and complex theoretical debate. Corporatism represents an attempt to walk a ‘conceptual tightrope’. The objective is to change the entry for ‘corporatism’ in the International Encyclopaedia of the Social Sciences, which currently reads ‘see fascism’, without describing something Of the many who have which can already be found under the entry ‘pl~ralism’?~ argued that this attempt fails, most have argued that corporatist writers fail to distinguish corporatism adequately from pluralism rather than fascism (or the fascist corporate state). This is because the new corporatist model has been extrapolated from existing practices which, until then, had been observed to be pluralist rather than fascist, and there has consequently been a rush to pluralism’s defence. This has resulted in a lack of attention to a key problem in constructing the corporatist ideal-type: distinguishing it from fascism. The chief consequence of this lack of attention is that corporatist writers have too readily and for too long accepted a corporatist ideal-type (Schmitter’s) which fails to do just that. This will now be investigated in the context of a search for the corporatist idealtype. Has the ‘grey and uninviting sea’ washed up a theoretically cohesive corporatist ideal-type? zyxwv zyxwv zyxw Tbe Corporatist IdeaLType: From Scbmitter to Cawson There are two closely interrelated requirements in the construction of an idealtype. The first is the definition of distinct institutional traits. The second, which is often overlooked, is that an ideal-type is, in Scruton’s words, ‘supposed to show why the actual events that mimic it develop as they do’.” Even if an ideal-type is unlikely ever to be validated empirically, the determinants of developments towardr the ideal-type are crucial to the coherence of the ideal-type itself. This gives the institutional traits a ‘dynamic’ rather than ‘static’ quality, in that the nature of their functioning is crucial. This also explains, however, why the most frequently used definition of corporatism as an ideal-type (Schmitter’s of 1974) has been erroneous.zsHis ideal-type is concerned solely with the political form of the state; that is, a distinctive mode of interest intermediation. With respect to the ‘conceptual tightrope’ the definition, as an ideal-type, is distinguishable from pluralism (which would not account for a systematic imbalance in all interest group access to the sate which the definition implies). But it is not - and is not intended to be - distinguishable from a fascist corporate state. Peter Self has noted that Cawson. ‘Indefence of the new testament’, p. 310. lnternaiional Encyclopedia of the Social Sciences (New York Macmillan, 1968), Vol. 3, p. 404, Vol. 5, pp. 334-41 and Vol. 12, pp. 164-8. The Encyclopaedia of the Social Sciences (New York, Macmillan, 1930), published over 30 years earlier,has no entry at all for corporatism, only entries for fascism (Vol. VI. pp, 133-9) and pluralism (Vol. X11,pp. 170-3). I‘ Scruton, A Dictionary of Political Thought, p. 212. 1J Schmitter viewed corporatism as ’. . . a system of interest intermediation in which the constituent units are organised into a limited number of singular, compulsory, non-competitive, hierarchically ordered and functionally differentiated categories, recognised or licensed (if not created) by the state and granted a deliberate representational monopoly within their respective categories in exchange for observing certain controls on their selection of leaders and articulation of demands and supports.’ P. C. Schmitter, ‘Still the century of corporatism?, Review oJ Politics, 36 (1974). pp. 93-4. I’ zyxwv zy zyxwv zyx MARTINJ. BULL 26 1 Schmitter’s definition describes a comprehensive incorporation of interest groups wilhin the f i m e w o r k of government. It does not say why this development is occumng nor how and in whose interest the system will work.26 As a result, Schmitter’s definition incorporates a wide range of regimes which he subsequently divides into those which have experienced ‘state-imposed’ corporatism and those which have experienced ‘societal corporatism’. This poses considerable problems for the analyst wishing to use Schmitter’s definition as an ideal-type which stands in contrast to pluralism because it is not clear whether a trend towards corporatism is the result of an increase in the power of the state or an autonomous development towards monopoly amongst interest groups. This overused definition, then, in limiting itself to a political form of the state, and remaining ambivalent with respect to its determinants, is an insufficient ideal-type from which to construct a pluralist+orporatist continuum. The fact that corporatist writing has depended so much on this definition has provided fertile ground for the pluralists’ theoretical onslaught. Cohen and Pavoncello, for example, have argued that, despite the ambiguities, Schmitter’s definition, in fact, contains implications about the power of the state.” If this is so, of course, the independent variable which distinguishes countries on a pluralist-corporatist continuum is ‘state power’, and the corporatist ideal-type is that which resembles the fascist corporate state. Thus Cox takes this route and concludes that ‘the state ultimately must dominate’, even under ‘societal corporatism’, and that therefore Schmitter’s definition b useful as an ideal-type but only as the state, as opposed to societal, variant of corporatism; that is, there is no such thing as ‘societal corporatism’.” Much corporatist writing has failed to respond adequately to this theoretical challenge. Rather than do so it has shifted its focus to analyse corporatism as a distinctive mode of policy-making, which has been subject to a further pluralist onslaught on the grounds that there is nothing distinctive about this mode of policy-making. Cawson is one of the few writers who have seen that to respond to the demands of political theory, corporatist writing needs to go beyond this divide. Cawson’s theoretical position is important because he not only responds to the pluralist challenge but, in doing so, rejects both the positions of Schmitter and those corporatists concerned only with policy-making. He overcomes the divide between corporatism as a mode of interest intermediation and corporatism as a mode of policy-making by combining the two and advancing an ideal-type which, in the identification of its determinant, precludes any association with state corporatism, thus apparently responding successfully to the pluralist challenge on both fronts. zyxwv zy zyxwvuts Corporatism is a specific socio-political process in which organisations representing monopolistic functional interests engage in political exchange l6 P. Self, Political theories of Modern Government: its Role and Reform (London, Allen & Unwin, 1985), p. 110. 2’ Y. Cohen and F. Pavoncello, ‘Corporatism and pluralism: a critique of %hitter’s typology’, British Journal of Political Science, 17 (1987), 117-22. A similar point is made by Diane Sainsbury, ‘Corporatism and pluralism: on the utility of the corporatist/pluralist dichotomy in political analysis’, in D. Sainsbury (ed.), Democracy, Store and Justice (Stockholm. Alunquist & Wiksell International, 1988). p. 94. Cox, ‘The old and new testaments of corporatism’, pp. 296-7 and 304-5. 262 zyx zyxwvutsr The Corporatist Ideal-Type and Political Exchange zy zyxw with stage agencies over public policy outputs which involves those organisations in a role which combines interest representation and policy implementationthrough delegated selfenforcement.29 This represents a rejection of the methodological basis of the distinction drawn both by Schmitter between ‘corporatism’ and ‘concertation’ and by Cox between ‘state corporatism’ and ‘pluralism’ on the same grounds: that they both make the error of drawing a distinction between the political form of the state and the nature of policy-making. Such a distinction is mistaken because it prevents the development of a comprehensive account of the state and political economy through corporatist theory. Corporatism involves the organisational link between intermediation and policy formation, not one or the other . . . What makes corporatism distinctive is the fusion of representation and intervention in the relationship between groups and the State.Io Pluralism and corporatism, then, are to be viewed as ‘distinctive processes coexistingin any given society, so that one can speak of a corporate and competitive sphere of politics’.” They represent end-points on a continuum with ‘corporate pluralism’ as a mid-point. What, for Cawson, is the determinant which locates countries at different points on the continuum? This is the crucial question for it is only through this that Cawson can distinguish his corporatist ideal-type from state corporatism. In his search for the ‘independent variable’ Cawson rejects those advocated by Martin (the degree of access and role in policy formation and implementation) and Crouch (the degree of membership discipline in the groups themselves) on the grounds that they are both zyx dependent variables which follow from changes in the independent variable of the degree ofconcentrationof political interests.They will not suffice for a more general theory which seeks to set corporatism within a historical proces~.’~ Martin’s variable is criticized for collapsing the distinction between state and societal corporatism. Crouch’s is rejected because, rather than concentrating on aspects of ‘structure’, it focuses more on the functions which groups perform and the strategies they adopt. ‘But what we really want to know’, Cawson writes, ‘is whether the choice between these strategies is voluntary, i.e. is it up to the group to determine its own position along the continuum? Cawson’s answer to this is unequivocal and lies in his identification of a structural variable: the concentration of political interests. Corporatism, then, is characterized by a limited number of groups, fixed interest domains, a hierarchical order and no competition; while pluralism is characterized by a large number of groups, overlapping interest domains, a fluid power structure and pure c~mpetition.~’ Cawson, CorporaIism and Political Theory, p. 38. yI I’ I’ ” Cawson, Corporafismand Political Theory, pp. 71 and 39. Cawson, Corporafirmand Political Theory, p. 39. Cawson, Corporatim and Polifical Theory, p. 42. Cawson, Corporafirmand Poliiical Theory, p. 42. zy zyxw MARTINJ. BULL 263 Cawson’s elaboration of corporatism appears to be the most concise so far advanced by corporatist writers, not only because of its comprehensiveness and ability to synthesize the contributions of other corporatist analysts, but also because it attempts to confine the concept to a structural trait: what determines the development of corporatism is the degree of concentration of political interests. This appears to free corporatism from association with claims about the power of the state implicit in Schmitter’s definition and thus prevents the collapseof the distinction between ‘statecorporatism’ and ‘societalcorporatism’, leaving the latter as the end-point of the continuum. At the same time, the ‘combination of interest representation and policy implementation through delegated self-enforcement’evidently differs from a situation of pluralism which focuses primarily on inputs (and not implementation of policy by interest groups) and fails to account for a systematic imbalance in the distribution of power which the definition implies. Cawson’s ideal-type, prima facie, appears successfully to walk the conceptual tight-rope. On closer analysis, however, it encounters fundamental problems. His quest for a corporatist ideal-type which is freed from authoritarian connotations and has certain clearly defined structural-institutional traits and determinants leads him to make a causal link which is open to question and which undermines the pluralistxorporatist continuum he has so carefully constructed. This will now be investigated. zyxwvu zyxwv Corporatism and Political Exchange: the Role of the State If Cawson’s ideal-type is successfully distinguishable from pluralism, what is the factor which makes it distinguishable from state corporatism or fascism? The answer is nor the concentration of political interests itself but the distinctive relationship those interests develop with state agencies, one characterized by, returning to Cawson’s definition, ‘political exchange’. This is the key phrase in Cawson’s definition because political exchange embodies the dynamics of the ‘specific socio-political process’ which is corporatism and which distinguishes it from state corporatism. Yet, considering that political exchange is at the heart of the corporatist ideal-type, the concept remains surprisingly undeveloped in the remainder of Cawson’s book.%It is essential, therefore, to explore the nature of this concept and its determinants. The concept of political exchange (scambiopolitico)has been developed largely within the field of Italian political science by Pizzorno and others.35It may be defined as a relationship entered into by the state and interest groups in which the state gives up part of its decision-makingauthority to interest groups in exchange for those groups guaranteeing their members’ adherence to the decisions reached. The very definition of political exchange makes an assumption about the state which is unaccounted for in pluralist theory: the state is an actor - or a ‘system of power’ - rather than a mere arena within which conflicts are settled. Political exchange involves two ‘actors’. This means that any determinant of corporatism must account for the presence of two ‘actors’:monopolistic interest zy zyxwvutsrqp zyxwv Apart from appearing in his definition @. 38). the expression is used only three other times in CorporarismundPohIical Theory when describing thedifferent levels ofcorporatism @p. 74,89, I 10). ’I A. Pinorno, ‘Political exchange and collective identity in industrial conflict’, in C. Crouch and A. Pinorno (eds), The Resurgence of Clan Conflict in Western Europe since 1968 (London, Macmillan, 1978). For a useful summary of the literature see L. Pam, ‘Political exchange in the Italian debate’, EUl Working Papers, 85: 174 (1985) (European University Institute, Florence). 264 zyx zyxwvutsr zy The Corporatist Ideal- Type and Political Exchange groups and the state (or state bodies) as actors with interests to pursue. For Cawson, the determinant of political exchange is to be sought in one structural variable only, the concentration of political interests; political exchange is activated as a result of a trend towards monopoly amongst interest groups. This assumes (since political exchange, by definition, involves two actors) that the emergence of the state as an actor is dependent upon the growth of monopolistic interests or that accompanying a trend towards monopoly is a growth in state power. This is a causal link which is open to question. It is quite feasible for a country to have monopolistic interests but a state which is lacking in power and is, in fact, controlled by those interests, or is in a position where it is too weak to deliver its part of the exchange. Conversely, it is quite feasible for a strong state to exist in the absence of monopolistic interests. The structural variable of the concentration of political interests is, therefore, insufficient to explain the emergence of political exchange. A second structural variable is required to account for political exchange: state power. Cawson fails to include this as a variable despite his claim elsewhere that the major element which the new corporatist writing added to the earlier insights of writers such as Beer and Rokkan was a conception of ‘the state’ as distinct from ‘government’; that it makes sense to speak of a ‘state interest’; and that it is necessary to recognize ‘the central presence of the state in corporatism, and to interpret the state as a system of power relationships . . .r.36 His ideal-type assumes state power as a ‘given’ in the modern industrial state instead of using it as a structural variable which will vary. Paradoxically, Cawson’s error lies in creating - in the process of identifying the determinants of corporatism - the very distinction he is attempting to overcome (and for which he criticizes Cox) in the pursuit of a comprehensive definition of corporatism: his definition merges the political form of the state and policy-making but his determinant is only a determinant of the political form of the state. There are two probable reasons for this omission. The first lies in the attempt to free corporatism from authoritarianism and therefore any association with state power. While Cox argues that the only structural variable is state power, Cawson wishes to emphasize that societal corporatism is the result of an autonomous development towards monopoly interests, one in which the state plays no role. This is insufficient, however, for political exchange to occur in the absence of a state which is strong enough and cohesive enough to activate political exchange and then to deliver its part of the exchange. The second reason is that Cawson, like Martin and Crouch before him, limits his perspective to creating a unilinear continuum between pluralism and corporatism. If a second structural variable is included, the idea of a continuum must be abandoned (see Figure 1). Breaking with the idea of a continuum is not new in state-group typologies. In contrast to previous attempts, however, the typology in Figure 1 builds directly on the theoretical work carried out by Cawson, at the same time as avoiding concept inflation and terminological ambiguity.” Stare rule results from a zyxw zyx zyxwvu Cawson, Corporatism and Political Theory, pp. 39, 45, 57 and 148. Chapter 3, in fact, deals entirely with the question of the state. ” Examples of other typologies include M. Golden, ‘Interest representation, party systems and the state in comparative perspective’,Comparative Politics, 3: I8 (April 1986) (six types); P.G .Cerny, The Changing Architecture of Polirics: Structure. Agency, and the Future of the Stare (London, Sage, 1990). C h s 3 and 6 (nine types and nine subtypes); P. Lange and M. Regini, ‘Introduction:interests and institutions - forms of social regulation and public policy-making’, in P. Lange and M . Regini (eds), State. Market and Social Regulation: New Perspectives on Ituly (Cambridge, Cambridge )6 zyxw zy zyxw zyxwv zyxwvut zyxw zyxwvu MARTIN J. BULL 265 - S 3 2w I- d I- rA Strong LOW High State rule Corporatism zyxwvutsrqp wed Pluralism Rule by private interests situation where a strong and cohesive state is confronted with a fragmented interest structure; corporatism where a similarly strong state is confronted by monopolistic interests; pluralism where a weak and essentially non-existent state is confronted with a fragmented interest structure; and rule by private interests where a weak state is captured by monopolistic interests. It should be stressed that these are ideal-types and that most national situations in practice will fall between the extremes. Moreover, most national situations will be spread across the different types according to the policy areas under study. The now rich literature originally prompted by the quest for ‘mesoCorporatism’ has shown the diversity of modes of interest intermediation and policy-making across different policy sectors in the same political system.3*In some policy sectors, state bodies may be almost authoritarian in their ability to make and implement policy; in others they may be very heavily dependent on the active participation of interest groups. In short, it is only by breaking with the idea of a continuum that a myriad of national and intra-national situations can be adequately understood. Two final points should be made about the typology. First, what appears to be missing from it is ‘state corporatism’. What happens to the distinction between state and societal corporatism in this typology? The key to understanding the absence of state corporatism in the typology is that the trend towards monopoly interests is autonomous or self-generated. Under state corporatism those interests are created and organized by state coercion. Consequently state corporatism is a sub-type of ‘state rule’ and not ‘corporatism’. Mussolini’s Italy is the typical example. Confronted with a fragmented and weak interest-group structure, he remoulded it according to the demands and needs of the fascist state.39Secondly, University Press, 1989)(three types). Spacedoes not pcrmit adetailedcritiqueof these typologies, but only the third one clearly builds on the existing literature (despite the use of different terminology) and attempts to make the sea less, rather than more, ‘grey and uninviting’. However, the exclusion of a fourth type in a two-by-two matrix (for reasons of empirical analysis) makes it difficult to view the authors’ third type (which extends right across one dimension) as a rigorous ideal-type. The apparent complexity of breaking with the idea of a continuum is illustrated by Sainsbury who, despite being sceptical about a continuum, concludes rather pessimistically that ideal-types based on more than one variable ‘involve increased difficulties in research and observation, especially in comparative analysis where more than one case is involved’; see Sainsbury, ‘Corporatism and pluralism’, p. 102. Lange and Regini (eds), State. Market, and Social Regulation, which analyses the case of Italy, is one of the best examples of this approach. Historians such as Maier have pointed out that the new corporatism has closer parallels with nineteenthcentury developments than the inter-war yean. See C. S. Maier. “‘Fictitious bonds ...of zyxwvu ’’ zyxwv zyxw zyxwvutsrq zyxwvut 266 The Corporatist Ia2al- Type and Political Exchange some pluralists may feel unhappy with the categorization of pluralism as consisting of a ‘weak’ state and a ‘low’ self-generated concentration of interests. If so, they should perhaps reconsider whether the cause of searching for a theoretically consistent pluralist ideal-type is advanced by the tendency to regard all that was written about liberal democracies before Schmitter’s seminal article in 1974 as necessarily ‘pluralist’. Even if pluralist writers have successfully qualified the apparent originality of Schmitter’s observations, they have not responded adequately to the implication of those observations: that it may be unsatisfactory to continue calling all of this ‘pluralism’.40 The establishment of the above typology is not sufficient to test the completeness of the corporatist ideal-type. Two structural variables (state power and the concentration of political interests) have been identified to explain the emergence of corporatism. These two variables would be sufficient to explain corporatism as a political form of the state. If it is accepted, however, that corporatism cannot be considered purely as a political form of the state but must be viewed as a distinctive mode of policy-making as well, can the structural variables adequately explain the concept which is at the heart of the corporatist process - political exchange? If it is accepted that ideal-types should go beyond merely defining distinct institutional traits, then they must explain not only the organizational nature of different actors but also their behaviour. Corporatism is more demanding in these terms than the other ideal-types in that it seeks to explain why the actors engage in political exchange. Do the structural variables cited explain the activation of political exchange? To answer this question the dynamics of political exchange must be explored. zyxwvut zyx T h e Dynamics of Political Exchange Cawson has a ‘static’ view of political exchange. In responding to Cox’s claim that if the state did not dominate the relationship interest groups would be under no compulsion to enter it, he states that under societal corporatism the element of compulsion does not arise from statefiat but from . . . the logic of membership and the logic of influence of interest associations. Simply put, both state actors and groups can be pushed into a closed corporatist relationship by a convergence of organizational imperatives.“ Cawson chooses his words carefully because they are essential to the maintenance of his structurally determined ideal-type. The consequence of the concentration of political interests reaching a certain level is the activation of political exchange. The adoption of this strategy is not ‘voluntary’ on the part of the actors but the result of the development of structural factors which constrain zyxwvuts zyxwvut wealth and law”: on the theory and practice of interest representation’, in S. D. Berger (ed.), Organising Interests in Western Europe: Pluralism, Corporatism and the Trmjormation of Politics (Cambridge, Cambridge University Press, 1981). Ch.1. Of corporatist analysts, Lehmbruch now finds the distinction between state and societal corporatism of limited use and even, in some respects, misleading; see G. Lehrnbruch, ‘Concentration and the structure of corporatist networks’, in J. H. Goldthorpe (ed.),Order and Conflict in Contemporary Capitalism: Studies in the Political Economy of Western European Nations (Oxford, Clarendon Press, 1984). Ch.3. Schmitter. ‘Still the century of corporatism? “ Cawson, ‘In defence of the new testament’, p. 3 10 (emphasis added). MARTINJ. BULL zyxw zy zy 261 the actors into behaving in this manner. As a result of the change in the structure of political interests, there is no longer a need to compete; rather, collusion is in the interest of both the monopoly groups and the state, so ‘closure’ occurs. As Dunleavy and O’Leary note: In corporatist reasoning elites colludeand collaboraterather than compete ... a closed process of accommodationbetween government,business and other institutional elites directs strategic policy in line with a shared conception of the national interest.42 The problem with this position is that it assumes that a change in the nature of interest representation (and, on the basis of the arguments in the previous section, the emergence of state power) will ‘push’ both actors into political exchange because the need to compete will be replaced by a mutual interest to collude. This is something rejected by Schmitter himself. Having established the structural changes relating to representation and control over members associated with the development of corporatism, he writes that: zyxw Even where these properties exist with respect to members, interlocutors (especially State authorities) may refuse to grant them corresponding corporatist ‘rights’. They may nor officially sanction existing monopolies, establish formal systems of guaranteed representation or depend on associational approval for the taking of policy measures. And even where they do this, when it comes to the actual allocation of goods or administration of regulations, they may not use associations to govern the complianceof affected interests,make interest associationsco-responsiblefor zyxwv subsequentdecisionsor devolveupon them authority to carry out directly the necessary tasks.” It might be added that the converse is also possible. In a situation where the state wishes to extend corporatist ‘rights’ to monopolistic interest groups, those groups may refuse either for ideological reasons or because they see their interests served better via other methods. In short, a view that political exchange can exist and persist as a result of structural factors appears to be limited in explanatory power. A more dynamic approach, therefore, has to be sought, which takes into account both the interests of the different actors, and the fact that the underlying basis of political exchange may be conflictual and not cooperative or collusive in nature. An approach based on rational choice theory responds to these demands. As Anthony Heath has noted, ‘providing there is some scope for choice, there is some scope for a rationalist approach’.” An outline of rational choice theory as an organizing concept in relation to political exchange demonstrates the zyxwv zyx 42 P. Dunleavy and B. O k r y , Theories of the State: the Politics of Liberal Democracy (London, Macmillan, 1987). pp. 143 and 330. ” Schmitter, ‘Democratic theory and neocorporatist practice’,p. 169. 44 A. Heath, Rational Choice and Social Exchange: a Critique of Exchange Theory (Cambridge, Cambridge University Press, 1976), p. 60. Footnote 13 referred to two other approaches to corporatism. Winkler’sapproach is primarily based on the notion of ‘statecontrol’ and is ‘static’ in nature. Jessop’s approach, and Marxist approaches in general, while obviously responding to the demands of ‘interests’ and ‘conflict’, nonetheless remain fundamentally structural in their explanations. 268 zyx zyxw zyx zyxwvu The Corporatist Ideal-Type and Political Exchange limitations of a pure structural approach to political e~change.~’ If state agencies and interest groups are regarded as ‘actors’, the determinants of an exchange relationship being activated between them will be, at the most simple level, the willingness and capability of the actors to do so. This raises two questions. First, under what conditions will the actors be willing to enter into a political exchange relationship? Secondly, on the assumption that this willingness exists, under what conditions will the actors be capable of doing so? Cawson effectively answers the first question by reference to the second. In other words, the willingness to enter into political exchange is explained by the actors’ capability; this capability is explained in terms of an increasing concentration of political interests and greater control of leaders over members inside the interest groups. ‘Organizational imperatives’, then, ‘push’ the actors into changing their strategies, because their interests are apparently best served this way. An analysis of willingness and capability, however, reveals the limits in this approach. From a pure rational choice perspective, the question of willingness should be answered in terms of the best means to obtain the goals sought. Rational choice theory does not concern itself with the nature of the goals sought but the means to achieve them. The test of an actor behaving rationally is whether or not the actor has chosen the most efficient means to achieve his goal.4 This is an important point because much of the corporatist literature has been concerned with the question of the interests satisfied by corporatism, often on the assumption that there must have been a change in the goals of the different actors, and particularly on the part of one of them. Cox and Hayward, for example, argue that the trade unions’ participation in a corporatist system needs to be explained because it ‘negates the primary function of trade unions: the protection of the material living standard of their members’.‘’ The implication here is that one of the actors’ main goals has been abandoned, but it remains possible for political exchange to be activated while actors’ goals remain, in fact, unchanged. Political exchange may simply represent the current best means to achieve those goals compared with the existing alternatives of action.& A pure form of rational zyxw zyxwvutsr zyxwvuts zyxwv &S Space does not permit a more sophisticated account of rational choice theory here. Nor is one necessary, however, for the purpose of the argument; see Heath, Rational Choice and Social Exchange. For use of rational choice arguments in the empirical analysis of corporatism see, for example. M.Regini, ‘The conditions for political exchange: how concertation emerged and collapsed in Italy and Great Britain’, in J. H. Goldthorpe’(ed.), Order andConJict in Contemporary Capitalism: Studies in the Political Economy of Western European Nations (Oxford, Clarendon Press, 1984). Ch.8 and, to a lesser extent, A. Cox and J. E.S.Hayward, ‘The inapplicability of the corporatist model in Britain and France’, International Political Science Review, 4 2 (1983). The broader debate over workers’ acceptance of wage regulation and the capitalist system is also relevant and hiis been the focus for rational choice arguments. See. for example, P. Lange, ‘Unions, workers and wage regulation: the rational bases of consent’, in J. H. Goldthorpe (ed.), Order and Conflict in Contemporary Capitalism: Studies in the Political Economy of Western European Nations (Oxford, Clarendon Press, 1984). Ch. 5; A. Przeworski, Capitalism and Social Democracy (Cambridge, Cambridge University Press, 1985) and D. S. King and M.Wickman-Jones, ‘Review article: social democracy and rational workers’, British Journal ofPolitical Science, 20 (1990). 387413. a A. Heath, Rational Choice and Social Exchange, p. 79. ” Cox and Hayward, ‘The inapplicability of the corporatist model in Britain and France’, p. 219. U See, for example, M.J. Bull, ‘From pluralism to pluralism: Italy and thecorporatist debate’, in A. Cox and N. OSullivan (eds), The Corporate State: CorporatLvn andthe State Tradiiion in Western Europe (London, Elgar, 1988). Ch. 4 and Regini, ‘The conditions of political exchange’. This may explain why Schmitter has noted that the outcomes produced by corporatism and pluralism are often similar; see Schmitter, ‘Stillthe century ofcorporatism?’, p. 102. The debate over the interests served in the eventual policies produced by corporatism has tended to detract from a comprehensive analysis of what brings the actors together in corporatism. The two need not necessarily be the same thing. MARTINJ. BULL zy zy 269 zy zyx zy choice argument, then, would explain the willingness of state agencies and interest groups to embark on political exchange as simply the result of a mutual change in the means to obtain unchanged goals. True as this may be, it is too simplistic to be consistently applicable, and Cox and Hayward’s point cannot be so easily dismissed. A further application of rational choice argument may be necessary to explain the activation of political exchange in many cases. Before an actor chooses the most efficient means to his goal, an actor (in rational choice theory) must decide what his goal or goals are. In the event of him identifying more than one goal he prioritizes them and then proceeds to evaluate the best means to maximize the goals sought. It is at this stage that explaining the willingness to embark on political exchange becomes more complex for two reasons. First, although the goals are given an order of priority some may not be zero-sum but partially achievable (making a clear-cut choice between which goals to achieve more difficult). Secondly, different methods to achieve the goals (partially or completely)may forseeablygenerate other gains which become goals in themselves. In short, it is more likely that actors, rather than adopting a twostage process of deciding on goals and then on the means to achieve them, will review different means they can pursue and the different type of gains each may generate, and then decide on the best course of action to pursue. From this perspective, then, the question of willingness to embark on political exchange should be answered, in the first instance, in relation to what the actors stand to secure from the exchange.49This adaptation of rational choice argument is particularly relevant in view of the fact that the actors are clearly losing something from entering into a political exchange relationship. Both actors lose their autonomy of action in that they agree to pursue policies which they may not otherwise have done had they not entered into such a relationship. The type of gains the actors stand to make from political exchange are twofold: direct and indirect.= The direct gains are those which relate to the goods embodied in the political exchange relationship: for the interest groups, preferential access to the decision-making process with an increased ability to affect market outcomes and protect their monopoly status; for the state, increased effectivenessin the implementation of its policies and greater legitimacy accorded to its actions. The indirect gains relate to organizational needs. In the case of the interest groups, the leaders have the opportunity of expanding their own patronage, power and prestige, and of expanding membership either by developing ‘selective incentives’ or through realizing the direct gains of the political exchange relationship.” The state, on the other hand, may benefit from the creation of ‘insider’ groups, giving them influence over decision-making for political purposes and excluding other groups; or, it may benefit from coopting groups into decision-making procedures and giving the leaders the illusion of having influence as a means to social control. zyxw zyxw ‘’ It should be added that the perception of the possible gains will be affected by what the other actor is doing or is expected to do. yI Obviously a comprehensive rational choice approach to this question would have to consider all the alternative means of action to political exchange and the type of gains they would generate, and the likely response of the actors to each other in adoptingcertain courses of action. It would also have to place a different value on different types of gains such as short-term versus long-term gains. The purpose here is simply to identify the type of gains that are possibIe from political exchange to explain a willingness to embark on it. ’Selective incentives’ is a reference to M. Olson, The Logic of Collecrive Action: Public Goodr and (he Theory ofCroups (New York, Schocken, 1968). pp. 51 andpassim. ’’ 270 zyx zyxwvutsr The Corporatist Ideal- Type and Political Exchange The precise combination of gains which are sought will obviously vary with each situation and further specification is unnecessary. Clearly, the willingness to embark on political exchange on the part of state agencies and interest groupscan be explained from the perspective of a rational calculation of costs and benefits on the part of state agencies and interest groups. Such willingness, however, may still be marred by capability. Under what conditions will the actors be capable of successfully fulfilling a political exchange relationship, thus ensuring that they obtain the benefits from the relationship without incurring costs greater than the loss of their independence of action? Three types of condition can be identified. The first type relates to the actors themselves, and specifically to their organizational abilities. These are the structural conditions which were developed earlier in this paper: a high concentration of interests and the existence of a state with sufficientcoherence and power. These conditions allow the actors to act in a unitary manner. The second type of condition relates to the actors’ external environment. Certain types of condition may either be necessary for political exchange to be activated or may affect the duration of political exchange. Some corporatist writers have attempted to identify these conditions in a generic manner with the implication that they have a structural nature: political conditions (the presence of a prolabour government); ideological/cultural conditions (a low level of political polarization and a political/cultural tradition of consensus over conflict); and economic conditions (stable economic growth providing resources for redistribution). Empirical analysis, however, has shown that these conditions need not necessarily apply in particular countries.52All that can be stated, therefore, is that political, economic, institutional and cultural conditions in each country may affect the possibility and outcome of political exchange. The third type of condition relates to the nature of the exchange itself. The success of a political exchange relationship will depend upon the ‘quality’ of the content of the particular exchange and its expected and actual outcome. The above arguments outline an alternative framework within which to analyse the emergence and persistence of political exchange. Using rational choice theory it would be possible to develop an alternative approach which negated the value of structural variables per se. Political exchange would be viewed as no more than a rational calculation of costs and benefits on particular issues by the actors involved. Such a pure approach, however, would have its limitations, which are inherent in any rational action model.” Underlying these arguments are two different models which effectively talk past one another. Nonetheless, they are not mutually exclusive. On the contrary, it could be argued that it is a merging of the insights of the two approaches which comes closest to explaining the dynamics of political exchange. While certain structural conditions may be posited for the existence of political exchange, the activation of such exchange relationships, and particularly stable exchange relationships, may depend as much on the exact nature or ‘quality’ of the content of the exchange and other external factors as on the structural factors themselves. Even if structural conditions are a prerequisite, the activation and persistence of political exchange cannot be assumed as a given. Since political exchange constitutes the zyxwvu zyx s2 M. J. Bull, ‘From pluralism to pluralism’, for example, shows that the conditions may be precisely the opposite to those normally cited for political exchange to flourish. It is not necessary to identify these limitations here. An example would be that a pure rational choice approach would not take into account ideological constraints on the behaviour of the actors. MARTINJ. BULL zyxw zy 271 heart of the corporatist process, this creates insurmountable problems in the construction of a structurally determined ideal-type which merges policy-making and the political form of the state. Corporatism represents a particular type of intermediation between monopolistic interest groups and the state but it should be recognized that other forms of intermediation between these two actors remains possible. zyxwv The limits to existing corporatist theoretical reasoning seem clear and relate to the failure to present a convincing theoretical ideal-type. Cawson’s ideal-type which is structurally defined by one independent variable is inadequate because his (correct) merging of the political form of the state with a distinctive mode of policy-making introduces a dynamic element which he subsequently fails to explain. Moreover, structural variables per se are inadequate to account for corporatism. The dynamics of political exchange indicate that, despite the necessity of structural determinants of an organizational nature, and the pressures these bring to bear on the actors involved, state-group relationships remain ultimately voluntary. The stability and persistence of a corporatist system depends, therefore, as much on the nature and outcome of the particular exchanges as on conditions which may be identified as structural. For corporatist theorists in search of ‘independent variables’ on which to construct a corporatist ideal-type, this appears to be an intractable problem. Can corporatism escape from this impasse? This question is best answered by first noting it is not necessarily corporatism’s impasse because it raises questions about ideal-types in general. A thorough analysis of other ideal-types may reveal similar problems to those present in corporatism. A tempting response to these problems would be to limit ideal-types to political forms of the state. The problem with this is that it would remove the dynamic element which explains the functioning of particular forms of the state. In the case of corporatism, it would involve removing political exchange which is precisely what makes corporatism distinctive. Corporatism would be left as a characterization for any relationship between monopolistic groups and the state. Cawson’s original point, then, that corporatism, to be viable, must combine forms of representation and policy intervention loses none of its validity. Yet as soon as ideal-types are expanded beyond the political form of the state to include policy-making, structural variables appear to be no longer always adequate. This does not mean that ideal-types should necessarily be abandoned, but simply that their limitations should be recognized. Dunleavy and O’Leary note that zyxw the usefulness of any ideal-type should be gauged from the contribution it makes to simplifying complex realities, allowing social scientists to derive more precise and testable hypotheses, which can then be checked empirically.Y Ideal-typesbased on non-structural as well as structural variables lose none of the clarity and utility of their structurally determined counterparts, but they do Dunleavy and O’Lcary, Theories of the State, p. 148. 272 zyxw zyxw The Corporatist Ideal- Type and Political Exchange reduce the value of typologies based on structural variables. The value of a ‘pluralist<orporatist’ continuum based on only one structural variable should be abandoned.