Annu. Rev. Polit. Sei. 2002. 5:305-31 DOI: 10.1146/annurev.polisci.5.112701.184858 Copyright (c) 2002 by Annual Reviews. All rights reserved. CORPORATISM: The Past, Present, and Future of a Concept Oscar Molina and Martin Rhodes Department of Social and Political Science, European University Institute, 50016 San Domenico di Fiesole, Florence, Italy; e-mail: oscar.molina@iue.it, martin.rhodes@iue.it Key Words economic policy, industrial relations, bargaining, social pacts ■ Abstract Following a period of almost obsessive academic attention in the 1980s, in the early 1990s the concept of corporatism fell from favor, as its explanatory powers appeared to wane and the Keynesian welfare systems under which it had nourished apparently fell into decline. In the late 1990s, a new interest in corporatism emerged, in line with new patterns of concertation and corporatist behavior in some unexpected places—countries in which the institutional basis for collaborative, bargained methods of policy making and conflict resolution seemed distinctly unpromising. We review the extensive literature on corporatism since the 1970s and consider its applicability in the contemporary period. We argue that an excessively structural-functionalist interpretation of corporatism led many wrongly to predict its demise as a form of policy making, and that an understanding of its persistence and new manifestations today must resurrect and strengthen some early, recently neglected insights into processes of political exchange. INTRODUCTION When Schmitter asked, "Still the century of corporatism?" in the title of his seminal 1974 article, the answer seemed to be a definite "yes." Many of the countries of Western Europe had developed complex and enduring forms of interest representation and intermediation that appeared to warrant a revision of how we understood the workings of democratic polities. Alongside party systems and parliamentary politics, we had also to understand the equally important contribution to governance made by networks linking government with interest organizations, many of which had quasi-public status as key actors in policy making (Offe 1981). Although the Scandinavian countries and Austria provided the most compelling examples of this type of polity, other nations contained less complete but nevertheless important elements of neo-corporatist bargaining as integral parts of their systems of government. But almost a quarter of a century later, two seasoned analysts of European economic policy and industrial relations argued that "there is no reason to doubt that, as a strategic program for the resolution of employment issues, neo-corporatism 1094-2939/02/0615-0305$ 14.00 305 306 MOLINA ■ RHODES is moribund—defeated on the ground by the actual evolution of employment relations before reluctant abandonment by its academic proponents" (Grahl & Teague 1997, p. 418). According to this view, the corporatism concept had outlived its utility, for in a neoliberal world of freer markets and welfare state retrenchment, the practice of corporatist policy making was bound to disappear. In reality, though, reports of the death of corporatism were greatly exaggerated. Employment issues (wage bargaining and labor market regulation) continued to be at the center of the still highly coordinated Scandinavian and Austrian industrial relations systems; meanwhile, new "social pacts" had sprung up across Europe from the late 1980s onward, specifically to deal with new turbulence and new challenges in the economic environment. Even the founding fathers of corporatist studies mistakenly assumed that if the structures on which corporatism had been based were eroded (i.e., Keynesian policy making and Fordist industrial organization), then corporatist behavior and patterns of governance would also disappear (e.g., Schmitter & Streeck 1991). In the core sections of this article, we attribute this error of analysis to the ways in which the concept of corporatism was developed from the 1970s onward. Although originally endowed with both structural and procedural significance as a system and a process of decision making (albeit by different authors and streams of inquiry), by the 1980s and 1990s, corporatism as an evolutionary form of governance—and especially as a mode of macro policy making—was being neglected. Instead, there was an emphasis at the macro level on corporatism as a "system" and its degree of association with certain kinds of economic performance, as well as a concentration at meso levels on ever more detailed empirical investigations of the sectoral arena of policy making. Thus, political scientists were caught unawares—both empirically and conceptually—in the mid-1990s, when macro policy concertation (i.e., forms of routinized bargaining between governments, employers, and trade unions) reemerged. Responding to this new reality requires new attention to corporatism as a macropolitical phenomenon. Also needed, as we suggest toward the end of this article, are a revival and reinterpretation of a now neglected part of the corporatist literature—that devoted to the modes and modalities of political exchange. A BRIEF HISTORY Since it first entered the academic lexicon and debate, the term corporatism has been characterized by ambiguity, imprecision, and a liberal, rather undisciplined usage. This has been because of its initial ideological connotations (connected to twentieth-century fascism), its gradual (and at times fashion-driven) adoption to explain diverse phenomena, and its application to very different time periods and countries. Despite great attempts to clarify the term (e.g., Nedelman & Meier 1977, Lehmbruch 1979, Panitch 1980, Williamson 1989), corporatism became a multipurpose concept. Its elasticity has ensured its popularity; but simultaneously its power to explain or even characterize political systems and processes has been diluted. Historically, the use of the term corporatism had a strong normative and ideological component, adopted by fascist and communist ideologues and activists CORPORATISM 307 Oh for widely divergent reasons, but having at its core the advocacy of an institutional relationship between systems of authoritative decision making and interest representation. Accordingly, the term became synonymous with the structures of a strong and dominant state. After World War II, the work of Shonfield (1965) marked the renaissance of corporatism as a theoretical concept. He provided empirical evidence supporting the validity of Manoi'lesco's (1934) argument, which predicted the gradual "corporatization" of western capitalist economies. Shonfield observed that, in order to attain a high level of macroeconomic performance within the Keynesian framework, modern economies had promoted processes, including state planning, in which the major interest groups are brought together and encouraged to conclude a series of bargains about their future behavior, which will have the effect of moving economic events along the desired path. The plan indicates the general direction in which the interest groups, including the state in its various guises, have agreed that they want to go. (Shonfield 1965, p. 231)1 The explosion in academic interest in corporatism that took place in the 1970s S ^ was accompanied by efforts to endow the concept with greater precision. Ac- cordingly, a basic differentiation was introduced between the old "state corporatism" and the "new" or "neo" societal corporatism. The work of Schmitter (1974) marked an academic milestone in this conceptual evolution. He clearly defined neo-corporatism as a form of interest representation distinct from pluralism, statism, and syndicalism. At around the same time, Lehmbruch (1977, 1979) put greater emphasis on neo-corporatism as a form of policy making in which concertation assumed central importance.2 Despite these differences, the common concern of both was in understanding the continuous and structured participation of interest organizations in policy-making and other stages of the policy process, especially policy implementation (Williamson 1985). During the 1970s, corporatism acquired the status of a social science model—an "approach, an intellectual framework, a way of examining and analyzing corpo-ratist political phenomena across countries and time periods" (Wiarda 1997, p. 23). Corporatism, it seemed, would offer an overarching and coherent method for understanding the working of economies and societies. At the same time, there occurred a shift in the locus of corporatist literature from political theory to political economy. This shift was provoked by the large divergences observed in the responses 'In this view, the macroeconomic framework and economic conditions were very important in determining the existence of corporatist institutions and policy making. Many of the "evolutionary" corporatist authors argued that the gradual consolidation of Keynesianism is a major explanatory factor for the transition from pluralism to corporatism (Manoi'lesco 1934, Beer 1956, Shonfield 1965). Similarly, Harrison (1980) argued that this transition was driven by technological, social, and macroeconomic factors. 2Lehmbruch sees corporatism as an institutional pattern of policy formation in which large interest organizations cooperate with each other and with public authorities not only in the articulation and even intermediation of interests, but also in the authoritative allocation of values and the implementation of policies. 308 MOLINA ■ RHODES and performance of economies during the oil crisis of the 1970s. Corporatist theory might, it seemed, offer considerable insights into the success or failure of those responses, and in particular into the economic steering capacity of governments. The separation between the two conceptions of neo-corporatism elaborated by Schmitter and Lehmbruch became "official" with Schmitter's (1982) distinction between "neo-corporatism 1" (a structure of interest representation) and "neo-corporatism 2" (a system of policy making). Henceforth, the character of the actors involved in the decision-making process and the nature of their relations with the state became the principal means of distinguishing a corporatist from a pluralist system of representation. At the same time, they became the keys for establishing a link between institutional configurations, policy making, and the character of policies and policy outcomes. The positions of Crouch and Martin were representative of the two main points of view in this debate. Crouch (1983) located the difference between pluralism and corporatism in the nature of the actors involved and in their internal organization, rather than in their role in the policy machinery. Martin (1983) argued instead that "what is at stake in the distinction between pluralism and corporatism is the extent to which organized groups are integrated into the policy-making arenas of the state." This fundamental insight was subsequently downplayed in the literature. The proliferation of neo-corporatist studies in the 1980s took two main directions. First, greater efforts were made to analyze the relationship between certain neo-corporatist institutional configurations and their respective policy systems, and the ways in which these distinguished them from pluralist systems of representation and decision making. Second, there were attempts to increase and improve the empirical evidence of neo-corporatism in practice, as well as to find a relationship between neo-corporatism and macroeconomic performance. At the same time, there were two closely connected developments in the application of the concept. First, the subconcept of meso-corporatism (Wassenberg 1982) was developed to examine the role of collective actors, not as peak class associations (as in the prior literature) but as organizations that cluster around and defend the specific interests of sectors and professions. Their relationships of power dependence with state agencies could be monopolistic and exclusive, but not necessarily tripartite in the manner of, say, peak employers and labor organizations in their role as incomes policy partners (Cawson 1986). Second, the concept of private interest government was developed to refer to the collective, private self-regulation of industry, with different degrees of assistance from the state, as a possible policy alternative to either market liberalism or state interventionism (Streeck & Schmitter 1985). Thus, the 1970s literature served to define corporatism more clearly, as well as to introduce greater clarity into the debate; the 1980s saw a more extensive empirical application and diffusion of the concept. But by the early 1990s, a new variant of the literature began to focus more speculatively on the fate of neo-corporatism amid the new economic turbulence of the time—and, in particular, the so-called decline of Keynesianism. Numerous writers proclaimed the extinction of the neo-corporatist CORPORATISM 309 "beast." Corporatism, it was argued, would be eroded from below, as technological change and the decline of heavy industry undermined the foundations of old-style European industrial relations. It would also be incapacitated from above, as looser labor markets and a shift in the balance of power from unions to employers rendered tripartite macropolitical bargaining less useful (see among others Schmitter 1989, Gobeyn 1993). However, predictions of the death of corporatism were not entirely borne out by subsequent developments. First, as Wiarda (1997, p. 180) argues, there was an evolution in the form of neo-corporatism. In its early years, neo-corporatist analysis was mainly concerned with tripartite relations between labor, business, and the state. It dealt with such issues as wages, production, social programs, labor benefits, and the like; it was associated with the early or intermediary stages of postwar industrialism. During the 1980s, there was in some countries a decline in this older neo-corporatism and the emergence of new forms of neo-corporatist decision making in the postindustrial policy arenas of education, health care, welfare, and environmentalism. This change involved not merely newer issues but also new corporatized actors. But at the same time, older, traditional corporatist structures and relationships were being adapted, rather than abandoned, in those countries where they were always most important (e.g., Austria and Scandinavia) in order to address the "older" employment and social issues that were of renewed concern. Moreover, the 1980s and 1990s witnessed the emergence of such structures and relationships, albeit in less embedded and institutionalized form, in many other European countries. To varying degrees, the Netherlands, Ireland, Portugal, Italy, and Spain have all implemented social pacts, based on peak-level concertation, to adapt to new economic policy challenges. These included new competitive demands stemming from the creation of the European single market and from the urgent necessity of bringing debts, deficits, and inflation under control for membership in the economic and monetary union. Since the mid-1990s, several authors have paid attention to these experiences (Fajertag & Pochet 1997, 2000; Traxler 1997; Rhodes 1998; Regini 1999; Goetschy 2000; Negrelli 2000). The "return" of neo-corporatism has led another group of scholars to analyze its cyclical nature and the reasons for its rise, demise, and resurgence (Schmitter & Grote 1997). The new wave of concertation has also revived discussions of the relationship between these new forms of corporatism and their antecedents. However, the theoretical analysis of this new wave of corporatism remains underdeveloped—largely because of the problems of application that have long afflicted the concept. PROBLEMS OF APPLICATION In the 1980s and 1990s, the study of corporatism followed two main paths. The first was the study of corporatism as a political phenomenon, which elaborated on Schmitter's and Lehmbruch's respective characterizations of corporatism as a 310 MOLINA ■ RHODES system of representation and a process of policy making. The second, which was related but assumed a distinct place in the literature, focused on the systemic effects of corporatist institutions. This literature became methodologically separate from the first. Whereas the first path of study explored the varieties of corporatism across different countries (and sectors), the second concentrated on the links between systemic features and socioeconomic outcomes. The first path was by and large a response to the criticism that neo-corporatist theory lacked strong empirical evidence to support its claim that corporatism was genuinely distinct from pluralism (Almond 1983, Jordan 1984). Thus, much subsequent work consisted of case studies presenting qualitative evidence on the representation of organized interests and their participation in decision-making structures. Following Cawson (1986), the three main interpretations of "corporatism" in this period can be identified as follows: 1. A specific form and process of interest intermediation; a distinctive way in which interests are organized and interact with the state (e.g., Schmitter 1974, Grant 1985). 2. A differentiated model of policy making with the intervention of social partners, and potentially a novel system of political economy different from capitalism and socialism (Winkler 1976, Lehmbruch 1977). 3. A different form of state within democratic and capitalist societies, emerging alongside and then dominating the traditional parliament-centered political system. The problems in the application of the concept lies precisely in these very different uses of the term. Not only has "corporatism" been used to characterize all three major elements of political systems—i.e., polity (structures and institutions), politics (processes and mechanisms), and policy (outcomes), but it has also been applied to numerous levels of the polity (the national economy, specific policy arenas, subnational governance, and industrial sectors) in very different national contexts. As for the second path of study, the imprecision, yet clear utility, of the concept triggered efforts to provide more rigorous quantitative indicators of its existence (Coevers & van Veen 1995, Mitchell 1996). Kenworthy (2000) identifies four categories of work in this area: 1. Studies of interest groups, concerned with the centralization, concentration, and density of labor and business organizations (Schmitter 1981, Cameron 1984, Wallerstein et al. 1997, Hicks & Kenworthy 1998). 2. Studies of wage setting, focusing on the centralization, coordination, and coverage of collective bargaining (Cameron 1984, Bruno & Sachs 1985, Calmfors & Drifill 1988, Soskice 1990, Layard et al. 1991, Nickell 1997, Wallerstein et al. 1997, Hall & Franzese 1998, Traxler & Kittel 2000). 3. Studies of interest group participation in policy making (Lehmbruch 1984; Compston 1997, 1998). CORPORATISM 311 4. Studies of political-economic consensus and strike rates (Keman 1984, Paloheimo 1984, Katzenstein 1985, McCallum 1986, Crepaz 1992). As mentioned, most of these studies were devoted to the analysis of corporatist structures and systems rather than processes. Many pointed to the benefits of corporatist policy making and structures for macroeconomic performance. The adjustment of European economies to the economic crisis of the 1970s provided an excellent opportunity for assessing these effects. The need to contain inflation rates in order to safeguard competitiveness increased the importance of wage and incomes policies (Chater et al. 1981, Boyer & Dore 1995). The differences observed in the ability of certain countries to achieve wage restraint, make it compatible with monetary policy, and offset rising unemployment were explained institutionally. With the help of the quantitative indicators and rankings elaborated, most authors concluded that a positive relationship existed between macroeconomic performance and the degree of corporatism (Flanagan et al. 1983, Pekkarinen et al. 1992, Scharpf 1997). Corporatist countries were deemed more successful in terms of inflation control, employment performance, and adjustment in periods of crisis than less corporatist or noncorporatist ones (Kurzer 1991, Western 1991, Crepaz 1992). There was considerable empirical evidence for a strong and positive correlation between such "cooperative economies" and high rates of productivity and investment growth, whereas "conflictual" economies (conventionally including Canada, the United States, and Great Britain) traditionally lagged behind in both (Gordon 1996). But in recent years, much work has gone into reassessing this relationship, especially after a decade of neoliberal policies and the apparent erosion of corporatist structures (Therborn 1998, Glyn 2001). This literature suggests that neo-corporatist theory was ill-prepared to explain the new economic developments of the 1990s. Analyzing eight small countries, Woldendorp (1997) showed that countries ranking higher in neo-corporatist scales do not, as a rule, perform better than countries ranking lower. Flanagan (1999, p. 1171) found that the indicators of corporatism constructed in the 1980s have little ability to explain macroeconomic performance in the 1990s. He argues that neo-corporatist theory has to develop more carefully the relationship between changes in the macroeconomic context and the capacity of corporatist institutions and arrangements to adjust. Hemerijck (1995) was one of the first to provide a detailed study (of the Dutch case) into how positive corporatist stability could degenerate into negative immobility, demanding a critical shift in institutional arrangements if the responsiveness and potential for innovation in the system were to be regained. Seeking a better understanding of the differences in macroeconomic performance in the post-Keynesian era of non-accommodating monetary policy, certain authors began to include the monetary regime as a variable that interacted with neo-corporatist institutions (Cukierman & Lippi 1998, Hall & Franzese 1998, Iversen & Soskice 1998, Iversen 1999, Iversen et al. 2000). They have focused on the relationship between coordinated bargaining and central banks, arguing that the key to effective wage and labor-cost setting is coordination among actors 312 MOLINA ■ RHODES and an emphasis on cost containment rather than on centralization and elaborate redistributive goals (Iversen 1999). Crouch (2001) suggests that such "organized decentralization"—to use Traxler's (1995) term—might prove to be a "new" (but unusual) form of neo-corporatism, in which representative organizations accept a role of restraining their members, a role that would also prove more appropriate for an era of global restructuring and monetary discipline than a disorganized and decentralized neoliberal alternative. This shift in the literature represented a first attempt to deal with the fact that neo-corporatist bargaining and institutions were clearly surviving and adjusting, not collapsing. Analysts had to make a sudden retreat from confident predictions of corporatist demise and assumptions concerning a new phase of "conservative convergence"3 to a more cautious study of corporatist redeployment. This retreat and subsequent equivocation about the direction of corporatist developments revealed a problem not in neo-corporatist theory as such (as we suggest below, it had always contained the tools for understanding evolutionary change) so much as in the ways it had been developed in the 1980s. Broadening the agenda had provided considerable clarification of what was meant by corporatism, as well as an appreciation of its consequences for policy, but there remained a gap in our understanding of the mechanisms or processes that linked institutional structures with outcomes (Williamson 1989, pp. 18-19). There was thus a tendency to underplay the refinement of what we might term the operation of corporatism, i.e., the relationship between the corporatist polity and corporatist policies. To put it another way, there was a failure to focus on the role and characteristics of corporatist politics.4 Scholars regarded a particular set of formal policy-making practices as neo-corporatist (social pacts, tripartite negotiations at national or peak level, and forms of concertation) but offered few insights into how these functioned or adjusted over time (Parsons 1988). As Flanagan (1999, p. 1156) argued, discussions on corporatism had neglected the micro foundations of decision making within interest organizations and the government, effectively assuming identical preferences among all members or an absence of democratic processes for resolving internal conflicts. As a result, the corporatist literature lacked precision on the process and outcome of bargaining among interest groups. Thus, it is not surprising that the "dysfunctionality" of corporatist arrangements and practices for capitalism was emphasized at the expense of understanding how "functionality" might evolve. In responding to a literature that more or less agreed that the demise of Swedish centralized wage bargaining in the 1980s heralded the 3See Gerlich et al. (1988) for a critique of this trend based on the Austrian case. 4One of the few attempts to correct this problem came from Cawson (1986), who rejected Schmitter's (1982) distinction between corporatism and concertation and Cox's (1982) distinction between state corporatism and pluralism [see also the debate between Cox (1988) and Cawson (1988)]. In Cawson's view, both Schmitter and Cox tried to draw a clear distinction between the political form of the state and the nature of policy making, when in practice no such distinction can be made. CORPORATISM 313 end of corporatism everywhere (if it could not survive postindustrial pressures in its "homeland," how could it be feasible elsewhere?), Wallerstein & Golden (2000, pp. 134-35) had to remind us that the adaptability of corporatism as a process depends primarily on politics, not technological determinism: Any lasting system of wage setting must be responsive to changes in the economic environment. Adjustments in the distribution of wages and benefits, however, do not necessarily require abandoning centralized bargaining. Whether or not employers and unions are willing to cooperate in modifying central agreements to accommodate changes in technology and in market conditions depends on the political relationships that exist within and between the unions and employers' confederations. At the end of the 1990s, then, several questions remained unresolved, or at least unclear in the literature. How does corporatism evolve? How can we relate changing corporatist processes to changing corporatist structures? If indeed there has been a return to neo-corporatism after the premature announcement of its death in the 1980s, on what institutional basis has this occurred? What are the traits that differentiate it from past forms and experiences? Or to put it another way, what has been the political logic behind processes of concertation in the last decade? CONTEMPORARY CORPORATISMS As suggested above, the corporatist literature in the 1980s and 1990s was ill-prepared to answer these questions and initially responded with a structural-functionalist interpretation of change. Schmitter (1989) suggested that the erosion of traditional neo-corporatist structures lay behind the extinction of processes of concertation and macropolitical bargaining (see also Gobeyn 1993, Walsh 1995). The challenges posed to unions (Crouch 2000) and the neoliberal character of economic policies during the 1990s (Glyn 2001) had undermined the structural conditions upon which neo-corporatism had been based and developed (Schmitter 1974). Lash & Urry (1987) and Regini (1995) argued that neo-corporatist institutions were degenerating in the transition to post-Fordism and would recompose on a more flexible, decentralized basis, demanding that our analytical ttention be redirected to the micro and meso (or local and company) levels of concertation between employers and employees. Schmitter & Streeck (1991) maintained that a combination of the business cycle effect (lower growth and higher unemployment) and European integration would remove the logic underpinning successful corporatism. While looser labor markets would empower employers, an integrated European economy, with less room for discretionary national economic policies, would reduce the incentives for unions to organize collectively and deliver wage restraint in return for package deals or side payments. Gobeyn (1993, p. 20) asserted bluntly that "contemporary economic realities ... make corporatism largely unnecessary. Market 314 MOLINA ■ RHODES forces alone can presently achieve labor discipline and wage demand moderation." Kurzer (1993, pp. 244^-5) concluded from a study of European economies that social concertation was no longer feasible: "[H]igh capital mobility and deepening financial integration prompt governments to remove or alter institutions and practices objectionable to business and finance" (for a survey and critique, see Rhodes 2001b). The alleged decline of neo-corporatism was thus interpreted with the same structuralist logic that was frequently used to explain its ascendancy—a tendency already criticized in the mid-1980s by Regini (1984). If the rise of the Keynesian paradigm had created the incentives and need for inclusive and negotiated forms of economic management, the end of the Keynesian golden age of capitalism had removed them. But the institutional bias in (neo) corporatist theory meant that all of these explanations underplayed actors' rational calculation of their interests and objectives in creating corporatist institutions. The relationship between forms of neo-corporatist intermediation and processes was all too readily regarded as unidirectional; in order to have peak-level social dialogue, social pacts, or macropo-litical bargaining it was necessary to have traditional neo-corporatist structures.5 Those structures, in turn, were linked to a particular phase of the postwar political economy. To the extent that there has been any revisionism in the literature, it has often been in the direction of class or class-fraction-based explanations (e.g., Iversen et al. 2000). These retain, however, a deterministic bias that leaves little room for political contingency. Does the literature on the return of corporatism provide us with an analytical way forward? The economic crisis of the early 1990s, which coincided with the first stages of European Monetary Union (EMU), obliged European governments to adjust their economies and institutions, which produced a proliferation of new forms of concertation and tripartite social dialogue (International Labour Review 1995; Visser & Hemerijck 1997; Rhodes 1998, 2001a; Crouch 1999; Pochet 1999; Perez 1999). Schmitter & Grote (1997) present several plausible explanations for the "return" (and apparent cyclical character) of neo-corporatism, including the economic business cycle and Hirschman's notion of shifting involvements. But they find no single satisfactory answer beyond the apparent destiny of certain types of system to keep up the search for bargained, consensus-based solutions. Other authors look for an explanation in the specific context of the 1990s. Traxler (1995) and Traxler et al. (2001) argue that a shift from classic to "supply-side" or "lean" corporatism has been driven by the shift from Keynesian economic policy to a non-accommodating monetary regime under the Maastricht Agreement and EMU. Rhodes (1998, 2001a) argues that concertation through a model of "competitive" neo-corporatism has been the response of European welfare states to increasing economic internationalization and accentuated economic integration in Europe. Others (Fajertag & Pochet 1997, Pochet 1998, Pochet & Fajertag 2000) 5However, Schmitter (1989, p. 64) does accept that although neo-corporatist associabil-ity and concertative policy making are empirically interrelated, they are not necessarily co variant. CORPORATISM 315 trace the changing nature of the external context of new social pacts from the early phases of EMU (when the issue of pay restraint became central) to the more recent period when labor costs and pension reform have risen to the top of the policy-making agenda. In stressing the functional appropriateness of new forms of corporatist experimentation, such interpretations draw the opposite conclusion from the literature on corporatist decline. Although the latter saw globalization, deindustrialization, and new post-Fordist production paradigms as undermining forces, perhaps there was also something in this turbulent environment that triggered the search for new modes of concertation. This interpretation suggested, in turn, the possibility of an evolutionary, transformative understanding of corporatism, rather than one that saw it as cyclical but essentially unchanging. We argue that to identify neo-corporatism with a stable combination of Key-nesianism and Fordism is to underestimate the capacity of actors to seek and sustain its benefits in more difficult times (Rigby & Serrano 1997, Martin & Ross 1999, Rigby et al. 1999, Crouch 2000). The work of Crepaz (1992) suggested, by contrast, that if neo-corporatist systems were functional for delivering better macroeconomic performance in the 1960s and 1970s, such systems might also have the capacities and internal flexibility to deliver similar outcomes under different conditions. Traxler (1998) advances this functional but evolutionary view for the Austrian case. The advantages of a concerted approach to adjustment are also outlined by Visser & Hemerijck (1997), who see a renewal of corporatism (on a much more flexible basis than hitherto) in the Dutch case as the outcome of a search for greater stability and predictability in a competitive and turbulent environment. Based on the Danish example, Blom-Hansen (2001) explains the return (or rather survival) of concertation as a governmental strategy to guarantee continuity in policies, which, if agreed on through concerted pacts, are much less likely to be changed because of a loss of legitimation. Nevertheless, a functional, systemic explanation of new corporatist experiments may be as inadequate, on its own, as past attempts to link corporatism with Keynesianism. We accept that economic conditions are important. The high unemployment rates and slack labor markets of European economies since the 1980s have pushed issues of employment and labor market regulation to the top of the political agenda. The conditions imposed on access to EMU made budget deficit reduction a central feature of member-state policy and have limited the resources available for traditional distributive politics and side-payments. The EMU macro-economic framework also entails the loss of independent national exchange rate policies and the creation of a central monetary authority with an inflation target, which makes wage costs a key component of macroeconomic adjustment and a major determinant of labor market performance. Accordingly, three issues have figured in processes of concertation during the 1990s: pay discipline, labor market flexibility, and the restructuring of social security programs. Thus, many works have studied the impact of EMU on industrial relations and collective bargaining institutions (e.g., Kauppinen 1998, Pochet 1998, Wallerstein 1998, Crouch 1999, Martin 1999, Traxler 1999). The change in the macroeconomic 316 MOLINA ■ RHODES as well as institutional framework has led some to renew the study of the relationship between institutions and macroeconomic performance (Calmfors 1998, Iversen 1999, Soskice 1999). In most studies, it is implicitly assumed that social pacts and the return to forms of concertation are the "second-best" choices of actors responding to external pressures, seeking new positive-sum solutions between particularistic goals and certain shared macroeconomic objectives. In those cases where EMU and its stringent conditions posed serious challenges to the economy (this is the case of peripheral countries: Italy, Spain, Greece, Portugal, and Ireland), social pacts have also served to legitimize and to make socially acceptable the sacrifices required by adjustment to the single currency. But although these contextual changes may help explain the presence of new incentives for governments, employers, and unions to engage in concertation, they do not in themselves explain how or why attempts at concertation achieved greater or lesser degrees of success in different countries. For although there is agreement on the renewed importance of concertation, there have been few serious and systematic attempts to explain why this has occurred and how it has been possible. We argue that understanding the return to neo-corporatism in the 1990s (both in terms of the redeployment of existing corporatisms and the emergence of new versions) requires greater emphasis on the goals and strategic behavior of actors than on institutions and systemic variables (Therborn 1992). It is therefore important to note the differences in actors' perceptions and political resources. As far as the role of actors is concerned, the main points to highlight with regard to the corporatisms of the contemporary period are as follows: 1. The state plays an active role in negotiations. In some cases, the government has intervened simply as a third actor. In others, it has pressured unions and employers to come together. In most cases, it has been responsible for steering the bargaining process (Pochet & Fajertag 2000). Pekkarinen predicted this development in the early 1990s in suggesting that the gradual erosion of those institutions traditionally supporting neo-corporatism would increase the importance of public authorities. These would "replace, reinforce, or supplement private, centralized wage bargaining with various kinds of official intervention" (Pekkarinen 1992, pp. 18-19). 2. The "balance between the negotiating partners has shifted substantially as compared with the situation prevailing when the pacts of the 1960s and 1970s were signed" (Pochet & Fajertag 2000, p. 18). The fragmentation or weakening of the trade union movement in most of Europe over the past 20 years has weakened unions' position vis-a-vis employers and governments. Accordingly, negotiations and the political exchange involved have also been shaped by new asymmetries in the objectives and action capacities of the different actors. 3. In the presence of these new asymmetries, the centralization and extensive associational coverage of the peak interest organizations of capital and labor may not be the sine qua non of successful corporatism at all (and probably CORPORATISM 317 never was, outside of Sweden and Austria). Instead, the flux of corporatist structures will bring with it a change in the processes of bargaining, rather than their disappearance. In certain contexts, new and flexible forms of con-certation may actually be more successful when unions are less centralized and less embedded in the workplace. Thus, it is precisely the strength of the German unions that has enabled them to resist government overtures to negotiate labor market reforms that their weaker Dutch counterparts have been able to embrace (e.g., Ebbinghaus & Hassel 2000). But these changed perceptions and behavior are explainable only if we focus on the underlying process of political exchange within the new social pacts: 1. Recently, neo-corporatist concertation has not been centered on the distribution of financial resources among participating actors; the economic surplus that had facilitated compromises via side payments in the past was simply absent or was heavily restricted (Schmitter 1989, p. 70). Instead, concertation has focused on the establishment of adequate institutional frameworks for macroeconomic management and microeconomic (supply-side) reform. 2. In practice, therefore, the new social pacts no longer stem principally from an incomes policy commitment, in return, say, for an increase in the social wage, as in traditionally conceived neo-corporatism. Instead, they stem first from the unions' acceptance of pay restraint in return for an undertaking by the public authorities and employers to promote employment creation (Fajertag & Pochet 1997). A second key component of the trade-off is the opportunity for the unions to help reform the welfare state via changes that span the labor market, social security, and industrial relations. This is especially important when the social partners administer social insurance schemes. Thus, the process of exchange has also had a participation component, i.e., pay restraint in exchange for enhanced involvement in policy making and institutional design (see Traxler 1997 for an extended discussion). 3. Given the cross-party nature of welfare reform projects in many European countries, the dependence of corporatist concertation on Left governments has been replaced by a new pragmatism with regard to macroeconomic management and micro-policy reform, making newer forms of corporatism "rather neutral to government composition" (Traxler et al. 2001, p. 302). 4. Looking at the general experience of changes introduced in industrial relations systems, we observe a tendency to strike a balance between the need for competitiveness and flexibility on the one hand and the need for macro-economic stability on the other. This balance has been sought (sometimes successfully, sometimes not) via processes of articulated decentralization, in other words, by increasing the importance of sectoral bargaining at the national level while reinforcing and strengthening company-level institutions. As discussed below, this process has contributed to the "network" character of contemporary corporatisms. 318 MOLINA ■ RHODES 5. There has been an extension of the issues covered by social pacts and agreements. Accordingly, macropolitical bargaining has not been restricted to the negotiation of (short-term) incomes policies but has also involved the introduction of (longer-term) institutional reforms. As Hassel & Ebbinghaus argue (2000, p. 35), this is because "concerted social policy reform is more than a means to facilitate wage moderation; it is also an end in itself." Rising social contributions linked to increasing social expenditure can counteract the positive effect of wage agreements, and thus social security reform necessarily becomes a key part of the wages-and-competitiveness equation. The linking together of issue areas in this way has favored the conclusion of agreements, for the greater the number of topics included in the negotiations, the greater is the chance of finding a compromise through "generalized political exchange" and trade-offs (Crouch 1990)—a point we return to below. 6. There has been a tendency for the most enduring social pacts (for example, those of Ireland and Portugal) to be the subject of constant renegotiation. This is partly because of their extension to new issues (and sometimes new partners) in line with changing pressures for reform. But it is also because, in the absence of traditional institutional prerequisites, an iterative process can help consolidate concertation and embed it in institutions and behavior. A strong process may compensate for a lack of traditional corporatist prerequisites in countries where organizational structures are weak and fragmented. Compared with the institutionally embedded, traditional corporatisms of Austria and Scandinavia, these new bargains are simultaneously more flexible and, in some cases, more ambitious. They are more susceptible to periodic breakdown and renewal, but even if they are not embedded in Keynesianism, they are often just as central to the macromanagement and steering of the economy. They are also much more clearly process dependent to the extent that the process (as a means of policy development as well as conflict resolution) also becomes an objective of reform. Not only the content but also the process can be the subject of negotiation, as new policy linkages are made and as partners enter and leave to express approval or disapproval of particular reforms. As argued by Traxler (1997, p. 35), there are two prerequisites for political exchange in the process of concluding social pacts: in terms of content, there must be a settlement (or at least suspension) of the conflict of distribution between capital and labor; and in terms of procedures, there must be a reciprocal allocation of representational and organizational privileges among the social partners. It is precisely when concertation involves the distribution of concessions and sacrifices rather than economic surpluses (as has tended to be the case for the 1990s) that procedural topics gain significance. These give the partners the opportunity to exchange representational and organizational privileges as compensation for material concessions and thus enhance the chances of building a compromise. An extension of this way of thinking interprets social pacts and their ensuing reforms and institutional changes within ongoing deliberative processes of learning (Hemerijck & Schludi 2000, Teague 2000). The endeavor to modernize economic CORPORATISM 319 and social governance structures has intensified policy learning and transfer across the European Union's member states (Teague 2000, p. 447). Thus, the reshaping of employment systems is not only about the diffusion of market-oriented reforms but also about a new wave of interactions between economic and social actors committed to updating traditional welfare states. As Hemerijck & Schludi (2000) argue, policy adjustment under these conditions, with increasingly dense linkages between policy areas and with greater attention to policy sequencing, should be understood as a dynamic political process of trial and error and of puzzling about reform. Policy adjustment in many European countries is now best portrayed as a system-wide search for new, economically viable, politically feasible, and socially acceptable policy mixes in which distributive trade-offs, defined via various forms of concertation, still play a critical role. Social pacts and macropolitical bargaining in the 1990s therefore differ in several important respects from the neo-corporatist concertation of the 1960s and 1970s—or at least from common assumptions about how those systems functioned.6 These differences relate especially to procedures and content and the ways in which these are subject to political exchange. Thus—just as Nedelman & Meier (1977) recommended for the analysis of "traditional" neo-corporatism— our understanding of corporatism in the contemporary period should be less concerned with the structural nature of the phenomenon (or with misleading structural-functionalist arguments about the possibility of its occurrence) than with the processes and procedures of political exchange. In terms of Schmitter's (1982) characterization, our focus should shift from "neo-corporatism 1" (the structure of interest representation) to "neo-corporatism 2" (the system of policy making). Our key point, however, is that one crucial element remains constant in both past and present manifestations of corporatism—the existence of political exchange. Counter to the basic intuition that informed the "demise of corporatism" literature, political exchange is still viable in the contemporary period—even if the currency of that exchange has been altered (many would say "devalued"). And because political exchange is possible, so too is macro-level concertation in various forms. The problem lies in the use and adaptation of the tools we already have for analyzing and understanding it. REFINING AND REDEPLOYING THE CONCEPT OF CORPORATISM A common characteristic of the literature that deals with the link between neo-corporatist systems and policy outcomes has been the use of structural features and "favorable contexts." Thus, given the pressures generated by globalization, 6As Nedelman & Meier (1977, pp. 48-56) argued, even the Swedish corporatism of the 1970s was much more dynamic, fluid, and shifting in its organizational basis and "interaction constellation" than Schmitter's original conception of societal corporatism allowed for. For similar arguments relating to Norway and Austria, see Lehmbruch (1984) and Gerlichet al. (1988). 320 MOLINA ■ RHODES economic integration (EMU), tertiarization, etc., it was argued that neo-corporatist structures have been eroded and can no longer yield the benefits they delivered in previous decades. But as we have begun to suggest, behind this alleged misfit between corporatism and economic development in the 1990s, there is an implicit consideration of corporatism as a structure, rather than as a policy-making process. "Neo-corporatism 1" had triumphed over "Neo-corporatism 2." In the light of developments during the past decade, this approach clearly suffers from several important shortcomings. First, an emphasis on structure proceeds from a static view of corporatism. But this approach is justified neither by the historical nor the cross-country diversity of the western capitalist world. Instead, if we think of corporatism from an evolutionary point of view, we can endow it with the capacity to adapt to a changing environment and find substitutes to those structural conditions that apparently no longer exist (Flanagan 1999). In this sense, the increasing intervention of the government in the corporatist developments of the 1990s should not be seen as a sign of weakness in the system (Pochet 1998, 1999). Nor should the search for new, less centralized (although still coordinated) systems of wage bargaining be seen as the end of corporatism. They should rather be understood as the search for new ways of maintaining the positive benefits that corporatist approaches provide. As recognized by Hemerijck & Schludi (2000, p. 208) in their analysis of effective policy responses in the 1990s, many of the countries that have successfully pursued a coordinated strategy of wage restraint could not rely on the traditional prerequisites, such as the strong, centralized, hierarchically ordered interest associations, of 1970s neo-corporatism. What they relied on, instead, was institutional adaptation and the discovery of a "new politics" of corporatism, with a different set of trade-offs and innovations in the process of political exchange. What we need to understand is the nature of this change. Those who predicted the demise of corporatism did so because they thought that political exchange was no longer possible. In fact, as we have argued above, the "currency" available for exchange did not disappear but rather changed. Thus, if political exchange was still possible, so too was concertation; and as we argue below, if there is scope for concertation—even in the form of sporadic social pacts—there is also scope for the institutional embedding of such practices. Second, if instead of conceiving of corporatism as a system we approach it as a specific form of policy making, then the structural argument linking the possibility of corporatism with a particular moment of postwar Keynesianism loses much of its validity. For although it is true that some of the structures usually linked to corporatism have been eroded, we cannot conclude that a certain system has ended or that certain forms of policy making are redundant. In order to test whether in the 1990s corporatism really stopped yielding the benefits it had previously delivered, or whether indeed the nature of those benefits has changed, it would be appropriate to look to the procedural aspects of corporatism rather than just its structural components. CORPORATISM 321 Oh Once we leave aside the purported structurally necessary conditions for corporatism, what we are left with is the nature of corporatism as a process—and the need to conceptualize the politics of corporatism much more thoroughly. This perception is not a new one. Bull (1992, p. 256), among others, argued that the value of any corporatist ideal-type based on structural factors is likely to be limited because of the nature of the dynamics at the heart of the neo-corporatist process, i.e., political exchange. More recently, Siaroff (1999), in an attempt to marry structure and dynamic processes, distinguished four key elements of an ideal-type corporatist political economy: (a) structural features, mainly the characteristics of interest groups (degree of unionization, internal organization, concentration of representation, encompassment, etc.); (b) functional roles, i.e., the integration of labor and business organizations into the decision-making process; (c) behavioral patterns, i.e., the role of political exchange as a means of attaining consensus in policy making); and (d) favorable contexts, such as a tradition of consensual politics, a long-term political role of or even dominance of a united social democratic party, or high expenditure on social programs. Obviously, not all of these ideal-type features will be found in all real-world systems, most of which will contain a complex S ^ mix °f them all. As we argued above, (b) and (c) may be present without a full complement of either (a) or (J). We suggest two ways forward: (a) refocusing our inquiry on the process of political exchange and (b) adopting the notion of integration as central to our understanding of how exchange can contribute to a new structuring of policy-making systems. Once again, placing political exchange at the core of the corporatist policy-making process is far from new. In the mid-1980s, both Lehmbruch (1984) and Regini (1984) began to introduce the concept of "political exchange" as a means of producing a more dynamic account of the phenomenon. Shortly thereafter, Cawson (1986, p. 38) offered a definition of corporatism as a specific socio-political process in which organizations representing monopolistic functional interests engage in political exchange with state agencies over public policy outputs which involves those organizations in a role which combines interest representation and policy implementation through delegated self-enforcement. Somewhat later, Crouch (1990) developed an empirical analysis of corporatism based on "generalized political exchange". This, in turn, was an extension of Pizzorno's (1977) classic distinction between collective bargaining, a political manifestation of market exchange, and political exchange, a process based on functional interdependence and mutual interest among actors (for an early theoretical critique, see Mutti 1982). The notion of generalized political exchange broadens the scope of the concept of political exchange from the labor market or industrial relations to policy-making processes in general. However, both Cawson and Crouch were thinking of monopolistic or all-encompassing organizations. Thus, Crouch considers generalized political exchange a function of high levels of union strength and the existence of centralized organizations of both capital and labor. But can the idea of generalized political exchange be applied under other conditions? 322 MOLINA ■ RHODES We argue that it is possible to use exchange relations as a means of recognizing and understanding the existence and operation of corporatist policy making, even— and especially—in the absence of the traditional structural prerequisites.7 For amid greater uncertainty, the nature of political exchange will be more complex and subject to shifting interpretations by actors. Marin's (1990a, p. 40) definition is helpful here in thinking of political exchange as forms of mutually contingent, macropoliti-cal and noneconomic transaction between autonomous, organized, collective actors with divergent/competitive/antagonistic but functionally interdependent interests, the binding character of which cannot be based on law and contract. The work of Marin is also important in linking the concept of political exchange with policy network analysis. Policy networks have not only been theoretically conceived as a specific form of interest intermediation or governance (Mayntz & Scharpf 1995, Scharpf 1997) but also as an analytical framework (Kenis & Schneider 1991). By combining the two, we can link complex processes of political exchange to the institutional settings in which they take place. The process of exchange itself can occur in different institutional settings as networks evolve. Exchange can also beget institutions, as rules and norms of behavior accumulate and accrete over time. Following Marin, we can argue that whereas elementary political exchange (e.g., basic forms of wage bargaining) can take place in weakly institutionalized contexts, generalized political exchange requires a hierarchy of rules ("calculated rigidity" in Marin's terminology). And this is why the linking together of diverse policy domains in weakly institutionalized social pacts can help sustain those pacts over time, for the generalization of exchange will in and of itself generate new relations of dependence and mutual commitment—and a form of institutional hierarchy and rigidity. Thus, even in the "lean" corporatisms of contemporary, post-Keynesian Europe, actors interact within a set of interdependent and hierarchically ordered policy fields or games— and the degree of that interdependence and order can ebb and flow (cf. Traxler et al. 2001). For as well as evolving and acquiring institutional presence and durability, rules and norms can also devolve and become dysfunctional, if the intensity of exchange between actors decreases or if corporatist stability degenerates into sclerosis and immobility—a development to which environmental changes will also contribute. Understanding corporatism as a variable and constantly evolving phenomenon allows us to understand how social pacts may fade away or, as is sometimes the case in contemporary Europe, how they gradually shift from being temporary emergency solutions to acquiring the status of formal or quasiformal subsystems of policy making. A focus on how actors' perceptions and behavior relate to—but are certainly not determined by—external pressures also allows us to understand 7Again, this insight is not new—merely underdeveloped in the contemporary literature. As Regini (1984, p. 141) advised, "Further research should focus on the variability of the conditions for political exchange ... rather than on the supposed organizational or institutional prerequisites, which, for all their importance in some situations, may be shown to be neither necessary nor sufficient in others." CORPORATISM 323 the redeployment of traditional corporatisms in countries such as Sweden and Austria. As recently argued by Stephens (2000), an analysis that focuses on the perceptions and behavior of unions and employers and the power relations between them provides a better understanding of recent developments in Scandinavian corporatism than one in which structure on its own (e.g., increasing organizational diversity) plays the central role. The concept of integration is also useful here for understanding the extent to which exchange becomes politically embedded—or in some instances less embedded—over time. It also helps us to "bring structure back in." Narrowly defined, integration is equivalent to the acceptance by the public authorities of M an active role for social partners in the policy-making process. The broader view, ° ^ which Siaroff (1999, p. 189) presents as an alternative paradigm to corporatism, .1 § defines it as a > OJ ~5 -3 long-term co-operative pattern of shared economic management involving | g the social partners and existing at various levels such as plant-level man- | a, agement, sectoral wage bargaining, and joint shaping of national policies in * £ competitiveness-related matters (education, social policy, etc). o 2 * 5 We suggest that, rather than replacing corporatism, the concept of integration *S ^ •o « can be used to enrich our understanding of how corporatism works and evolves. "S § Treu (1992) provides a useful extension of an argument originally developed by o ■§j' Lehmbruch (1984, pp. 66-74), who sought to show how the instutionalization « > of corporatist systems occurred along two dimensions—the vertical (the pattern £j d of participation of social actors in policy making and implementation) and the ° |j> horizontal (the pattern of concertation between social actors and government). Treu distinguishes integration from less developed forms of concertation in processes of consultation and collective bargaining. Collective bargaining corresponds to the simple exchange required for government policies to be implemented and effective, without giving social partners policy-making power. Consultation simply provides social partners with access to information, whereas integration gives them real influence over policy design. Treu further distinguishes between quantitative and qualitative integration. Whereas quantitative integration refers to the number of partners involved, qualitative integration has both horizontal and vertical features. Qualitative integration varies horizontally in terms of the number of policy areas that are linked together in processes of concertation. The greater the number and relevance of policy areas— and by implication the greater the scope for "generalized exchange"—the higher will be the degree of integration. Qualitative integration also varies vertically in terms of the capacity of actors to influence the policy process from policy design through implementation, often via the delegation of authority from the state. So, although Regini (2000, pp. 160-62) distinguishes between political exchange and the delegation of authority as separate phenomena (arguing that whereas the former was characteristic of classic forms of corporatism, delegation is more important in recent social pacts), in reality the delegation of power and influence has always been part of the process of exchange. 324 MOLINA ■ RHODES Anyone who has tried to understand the fluid and shifting systems and levels of concertation that have emerged around Europe in the past decade or so will appreciate the contribution that these tools of analysis can make. Tracing the evolution of those systems along the dimensions of quantitative and qualitative integration (with special attention to the latter's horizontal and vertical characteristics) provides more than a means of static cross-national comparison. With due attention to the nature of political exchange involved, it also allows a better understanding of the dynamics and evolution of corporatist governance—of how the process of exchange can build and transform institutional frameworks and the nature of institutional "embeddedness" under different circumstances. It is also, crucially, a means of avoiding futile disputes over whether a system is corporatist or not and whether concertation is the equivalent of corporatism or is a weaker, less institutionalized subcategory. Instead, there exists an array of corporatist phenomena, ranging from the intensive to the extensive and from the highly to the weakly integrated, encompassing manifestations that are horizontally inclusive or exclusive and vertically shallow or deep. Particular instances of corporatism do not remain static; they evolve along these dimensions, as witnessed by developments in many European countries over the past decade or so. Although that evolution is clearly driven in part by external pressures, a genuine understanding of corporatist development must focus on the internalization of environmental constraints and opportunities in the politics of corporatist concertation and in relations between the actors in the system. Empirical evidence of the ways in which particular industrial regimes move along a "corporatist trajectory," with varying degrees of qualitative and quantitative integration, has been provided by numerous authors, and several studies have begun to illustrate that social pacts and the most recent (post 1980s) era of corporatism can best be understood in terms of a networked form of governance. Based on a study of the Dutch case, Hemerijck has provided one of the best illustrations to date of how corporatist relations can generate innovation or stasis in policy making, depending on the dynamics of societal support or institutional integration, with change taking place at "critical junctures" (Hemerijck 1995, p. 197). His "dynamic model" provides useful lessons for understanding other cases in which classic forms of corporatism have survived the crisis of Keynesianism and have been reconfigured as more complex, flexible, and networked varieties of economic governance. At a more general and comparative level, Traxler et al. (2001) have shown how labor-relations regimes can shift backward and forward over time from what they call "classic" to "lean" corporatism, via intermediary, heterogeneous stages, depending on levels of centralization, coordination, and participation in bargaining. They make a strong case for contrasting classic corporatism, as a hierarchical system of governance solidly embedded in systems of Keynesian demand management, with a more recent "lean" variety, which, in line with our argument above, they portray as a "distinct governance mode beyond hierarchy and market, something that is widely understood as the constituent property of a network." CORPORATISM 325 Weaker levels of bargaining governability in such systems are substituted by an external straightjacket—tough monetary policy (Traxler et al. 2001, p. 301). We would simply stress that classic corporatism, as argued, for example, by Lehmbruch (1984), was also a networked form of governance, albeit one with different properties from more contemporary varieties. Equally, we suggest that, far from being "beyond hierarchy," social pacts and recent examples of "lean" or "supply-side" corporatism also depend for their successful operation on a degree of hierarchy and institutional integration. In itself, a hard monetary policy as the external "functional equivalent" of internal, hierarchical discipline is insufficient for explaining the survival of such experiments. This is especially so when, as is often the case, they link, in a form of generalized political exchange, productivity goals (supply-side modernization) with distributional package deals (covering incomes policies, pensions, and social security reform). The relationship between the productivity and distributional coalitions within such pacts is complex, and tension between the two is a constant source of instability (Rhodes 2001a). Without hierarchy and integration—linking levels of bargaining and involving strong commitments (and exchange) between social partners and government—the new "networked" forms of corporatism could hardly be sustained. CONCLUSION This article began with a survey of the main streams in the corporatist literature and argued that although it is rich and extensive, by the 1980s and 1990s there had been a manifest failure to find new ways of exploring contemporary reality. The political branch of corporatist studies had either reached a dead end in its attempts to define the nature of the phenomenon, or had proliferated, but simultaneously dissipated itself, in ever more detailed sectoral case study research. At the same time, the political economy branch had exhausted its capacity for measuring corporatism and equating corporatist systems with certain types of economic performance. Although large-scale, cross-national studies of this type have imparted important insights—especially of a historical nature when quantitative work has been linked to complex narratives (e.g., Hicks 1999, Swank 2002)—there have been diminishing returns on investment in such exercises (Woldendorp 1997, Flanagan 1999). Most important for our argument, neither branch has been well adapted to the task of explaining the developments of the 1980s and 1990s. We have not presented a new theory in their stead. Nor have we attempted to redefine the concept of corporatism. We have rather argued for a resurrection, reassessment, refinement, and application of certain tools of analysis, derived in particular from theories of political exchange, that have long been present in the corporatist literature as a third branch of the study that has withered on the tree. From Lehmbruch (1977) and Pizzorno (1977) through Martin (1983) to Cawson (1986) and the work of Crouch and especially Marin in the early 1990s, there has been a deep concern with corporatism as complex process of political exchange that we believe is highly relevant for understanding the contemporary period. 326 MOLINA ■ RHODES This is particularly so because of the less formal, less institutionalized, and less predictable nature of the new types of corporatist concertation. Their structures and actors should be understood in terms of networks; their logics in terms of the processes that underpin them; their fate in terms of the evolution of integration and the changing "currency" of exchange. To date, with few exceptions (e.g., Traxler 1997, Regini 1999, Hassel & Ebbinghaus 2000), those who have begun studying the new neo-corporatisms or systems of concertation of the late twentieth and early twenty-first centuries have barely moved beyond description and speculation. If we go back to some of the earlier insights in the neo-corporatist literature that have recently been neglected, our attempts to understand contemporary corporatist practices and experimentation can be enriched. The Annual Review of Political Science is online at http://polisci.annualreviews.org LITERATURE CITED Almond G. 1983. Corporatism, pluralism and professional memory. World Polit. 35:245-60 Beer S. 1956. Pressure groups and parties in Britain. Am. Polit. Sci. Rev. 50:1-23 Blom-Hansen J. 2001. Organized interests and the state: a disintegrating relationship? Evidence from Denmark. Eur. J. Polit. Res. 39 (3):391^16 Boyer R, Dore R. 1995. Les Politiques des Revenues in Europe. Paris: La Decouverte Bruno M, Sachs J. 1985. Economics of Worldwide Stagflation. Cambridge, MA: Harvard Univ. Press Bull M. 1992. The corporatist ideal-type and political exchange. Polit. Stud. 40:255-72 Calmfors L. 1998. Macroeconomic policy, wage-setting and employment—what differences does the EMU make? Oxf. Rev. Econ. Policy 14(3):125-51 Calmfors L, Driffill J. 1988. Bargaining structure, corporatism and macroeconomic performance. Econ. Policy 6:13-61 Cameron D. 1984. Social democracy, corporatism, labour quiescence and the representation of economic interests in advanced capitalist society. See Goldthorpe 1984, pp. 143-78 Cawson A. 1986. Corporatism and Political Theory. London: Blackwell Cawson A. 1988. In defence of the New Testament: a reply to Andrew Cox, "The Old and New Testaments of corporatism." Polit. Stud. 36:309-15 Chater R, Dean A, Elliot R. 1981. Incomes Policy. Oxford, UK: Clarendon Coevers F, van Veen T. 1995. On the measurement of corporatism. Labour 9(3):423^2 Compston H. 1997. Union power, policy making and unemployment in Western Europe, 1972-1993. Comp. Polit. Stud. 30:732-51 Compston H. 1998. The end of national policy concertation? Western Europe since the Single European Act. /. Eur. Public Policy 5(3):507-26 Cox A, ed. 1982. Politics, Policy and the European Recession. London: Macmillan Cox A. 1988. The Old and New Testaments of corporatism: Is it a political form or a method of policy-making? Polit. Stud. 36:294-308 Crepaz ML. 1992. Corporatism in decline? An empirical analysis of the impact of corporatism on macroeconomic performance and industrial disputes in 18 industrialized economies. Comp. Polit. Stud. 25(2): 139-68 Crouch C. 1983. Pluralism and the new corporatism: a rejoinder. Polit. Stud. 31:452-60 Crouch C. 1990. Generalised political exchange in industrial relations in Europe during the CORPORATISM 327