The globalization debate

With the above antecedents, it took the conjuncture of the collapse of Soviet socialism in the early 1990s, the financial crisis of 1997 that destabilized the East Asian model, and the seeming stability of globalizing Western capitalism until the early-mid 2000s to consolidate the research agenda of comparing capitalism with capitalism. Ever since the 1990s, a central question for comparativists was how globalization impacted upon the institutions, strategies and performance of different political economies. While studies of its factors and consequences abound, many authors saw globalization as a powerful transformative force that would bring about convergence on a single liberal model characteristic of the Anglo-American world (Ohmae Reference Ohmae1994; Gray Reference Gray1998).

The convergence logic posits that globalization-induced capital mobility leads to a “race to the bottom” of wages, social standards, and corporate taxes. To maximize returns, investors locate production to sites where labor is cheapest and taxes are lowest. This sets in motion a downward spiral, as governments engage in new rounds of reforms to reduce the burdens on capital and ensure national competitiveness. States are also constrained by the careful scrutiny of international financial markets forcing them to prioritize price stability and fiscal balance. This “input weighted” convergence model (Hay Reference Hay2004, p. 233) ignores the fact that domestic institutional and political factors mediate the pressures of globalization. This precisely is the tenet of comparative political economy. Over the 1990s, comparativists started to challenge the idea of convergence with ever more sophisticated conceptual tools and empirical analyses of capitalism's persistent variants.

The recent wave of comparative political economy literature shares a number of core assumptions and related theoretical innovations (Hall Reference Hall, Lichbach and Zuckerman1997; Crouch Reference Crouch2005: Ch.1; Jackson and Deeg Reference Jackson and Deeg2006). The first contribution is the identification of a number of key institutional areas and their systematization as different configurations serving as building blocks for political economy models. Financial systems, corporate governance, industrial relations, institutions of skill formation and (elements of) welfare regimes are most often evoked as core institutions. Second, a peculiar economics-driven revival of a holistic view of society has been underway, shifting the focus of attention from single or few institutions to the internal logic of whole institutional configurations, which have been termed complementary insofar as their elements reinforced each other. Third, institutional complementarity limits the number of plausible configurations. Proposed typologies range from the neatest bipolar models to those recognizing half a dozen different clusters (Hall and Soskice Reference Hall and Soskice2001; Kitschelt et al. Reference Kitschelt, Lange, Marks, Stephens, Kitschelt, Lange, Marks and Stephens1999; Schmidt Reference Schmidt2002; Amable Reference Amable2003; Whitley Reference Whitley2007). Finally, while a group of scholars has traced capitalisms’ performance to varied national models, others have theorized the system on grounds of diversity on sub-national, or trans- and supranational levels (Schmitter Reference Schmitter, Brunetta and Dell-Arringa1990; Hollingsworth et al. Reference Hollingsworth, Schmitter and Streeck1994; Djelic and Quack 2003).

In an assessment of the field, Hall concluded that the study of advanced capitalism “can look back on considerable accomplishments in recent years” (1997, p. 198). By the second half of the 1990s, however, the field also seemed to have reached a crossroads, since it was replete with hard to reconcile divisions. For instance, by rediscovering the social whole through the theoretical lens borrowed from economists, historical institutionalism has in the view of some become receptive to the a-historical idea that the function of “non-market institutions [is to] make the market possible” (Crouch Reference Crouch2005, p. 15). Accordingly, the choice between historically grounded but methodologically laxer, and rigorous but alleged functionalist, perspectives has created a dividing line, and ultimately widened the gap between studies of social dynamics and statics. To be sure, even prior to its encounter with economics, historical institutionalism was not adequately equipped to capture dynamic social processes. Indeed, the prevailing “punctuated equilibrium” model of path-dependent change bred dissatisfaction and criticism, which led to an intensive search for more dynamic concepts (Pontusson Reference Pontusson1995; Thelen Reference Thelen2004; Streeck and Thelen, Reference Wade2005; Crouch Reference Crouch2005).

In a similar vein, pressure for parsimony raised the stakes involved in settling on an “optimal” number of models. The drawback of neat binary logics was that too many countries stayed in the “grey zone” of mixed cases or fell out of the typology altogether. Conversely, giving specific status to countries in poorly defined intermediate positions entailed the risk of ending up with “as many models of national capitalism as there are nations” (Hall Reference Hall, Lichbach and Zuckerman1997, p. 183). Finally, craving for parsimony and rigor tended to tip the balance between reductionism and eclecticism in favor of the former, as it justified the preference for assuming rational agency motivated by economic self-interests at the neglect of political and cultural influences (ibid. 190).

VoC's contributions

In light of the above tendencies and divisions we can better grasp VoC's secret of success. It entered the field in the right historical moment to galvanize partly deadlocked important debates. By clearly taking side in favor of parsimony and rigor, and skillfully incorporating concepts and tools from economics, it has pushed this agenda further than any other approach, this way exposing it in all its inherent strengths and weaknesses.

The Hall-Soskice volume makes several contributions. By giving micro-foundations to capitalist institutions and their adjustment it seeks to combine historical institutional and rational-individualist perspectives albeit with a stronger emphasis on the latter. It is a firm-centered approach. It suggests that firms, in order to develop and exploit core competences, need to coordinate with other actors including their own employees, unions, share – and stake holders, clients, and competitors. To solve problems that inevitably occur, enterprises rely on institutions, defined as sets of formal and informal rules to regulate behavior. Five spheres are considered: industrial relations, vocational training and education, corporate governance, regulations of standard-setting, research and development (R&D) and technology transfer; and institutions coordinating managerial incentives and the interaction between managers and employees. Among these, strong emphasis is placed on the institutions of skill formation and their links to broader forms of social protection.

The approach compares national political economies by how firms resolve the coordination problems they face in the above areas. Firms are not considered as autonomous actors capable of taking their fate in their own hands. Rather, in the VoC perspective strategy follows structure. Existing institutions orient firm behavior in systematically different directions in “coordinated market economies” (CMEs) – Germany being the prime example –, and “liberal market economies” (LMEs), such as the US. In the former, firms coordinate their behavior with other actors mainly through non-market means, such as incomplete contracting, networks, and collaborative relationships, whereas in the latter, firms’ behavior is primarily governed by markets. These institutional differences also impact on patterns of innovation and world market specialization, thus providing for comparative institutional advantages.

More than other approaches, VoC adamantly stresses tightly coupled institutional complementarity among the five domains. As parts of an interlocking totality institutions assure each others’ efficient functioning as well as that of the economy as a whole. VoC's understanding of institutional complementarity is also specific in that it is equaled with isomorphism. According to VoC, hybrid or mixed models of capitalism do not work as efficiently as pure cases of coordinated or liberal market economies.

VoC assumes a strong notion of path dependent change. The first source of change is an evolutionary logic, which predicts that once a specific coordination mode has emerged in one area, it pushes for complementary institutions in other spheres. The second source stems from outside, the world economy with its ever changing technologies, products and tastes, which challenge existing national institutions and firm practices. VoC expects that in order to sustain comparative institutional advantages, firms engage in recreating their foundations. Heightened international competition thus leads to adjustment strategies to restore earlier equilibriums, with LME firms pressing for further deregulation and those in CME for renewed coordination. Even capital mobility, commonly assumed to destabilize organized capitalism, reinforces differences in national institutional frameworks. Transnational corporations (TNCs) engage in “institutional arbitrage” by shifting particular activities to countries whose setup either offers the same kind of advantages they enjoy back at home, or supports specializations that could be less efficiently pursued in their country of origin. While institutional complementarity is strongly associated with path dependent gradual change, there is one exception. Deregulation of financial markets in CMEs might snowball in a way that ultimately decomposes the system as a whole.

VoC's seductive forces

A recent volume which summarizes and reflects upon the reception of VoC concludes that it “has revolutionized the study of contemporary political economy” (Hancké et al. 2007, p. 36). Within the field, the VoC approach stands out indeed. A cursory look at Google Scholar reveals that the Hall-Soskice volume has been quoted more than two thousand times, thus leaving far behind any other recent book that compares advanced capitalist states. This indicates a remarkable impact even if some reservation concerning the magic of citation numbers is justified. What has made the approach so influential?

Our explanation is that it effectively combines a scientific appeal with a normative concern.

Its parsimony is truly attractive. If the central problem for comparative capitalism is the persistence of systemic diversity against the background of external pressures, then Occam's razor makes VoC the preferable approach. Its arguments in support of viable capitalist models are simple, straightforward, and testable. Complexity is also reduced to a minimum when it comes to the number of varieties or the logic of institutional complementarity. Another source of appeal is the incorporation of tools from economics to a greater extent than is the case with other approaches, giving VoC a competitive edge in political science and sociology struggling hard to lose their image as representatives of “soft sciences”, and looking up to economics as the role model to follow. The approach rigorously links individual choices (the micro-foundations) with institutional and macroeconomic outcomes, and thus lends itself to game theoretical analysis and modeling.

Beyond that, by building on the difference between general and specific skills, VoC has also advanced the asset theory of actors’ preferences to new areas of social policy and the choice of electoral systems. This way, it seeks to complete a paradigmatic shift in comparative political economy, which undermines long-held ideas about the significance of worker agency for the origins of generous welfare states and coordinated forms of capitalism. Earlier, welfare state generosity was viewed as a product of “politics against markets”, reflecting patterns of working class mobilization, mediated by political competition and left-wing governments, and associated with redistributive justice (Esping-Andersen Reference Esping-Andersen1990). In turn, the VoC interpretation sheds a whole new light on the welfare state as a product of “politics with markets” and supported by cross-class coalitions of capital and labor seeking to hedge risks associated with their investments in “co-specific skills” (Ebbinghaus and Manow Reference Ebbinghaus and Manow2001; Estevez Abe et al. Reference Estevez-Abe, Iversen, Soskice, Hall and Soskice2001; Iversen Reference Iversen2005). Welfare arrangements might still entail redistribution, but what really carries weight is that they are efficient solutions to the problem of specific skills provision, and thus functionally complement the CME production regime.

In a similar vein, the conventional view of the origins of electoral systems of proportional representation (PR) associates these with a defensive move on part of a divided right, which tried to prevent electoral elimination by a strong left (Rokkan Reference Pontusson1970). Some VoC authors, in contrast, see a close connection between the PR system and the concern of business and labor to secure a regulatory framework that allows investment in specific skills. PR is best suited to this purpose, as it allows for the representation of special interests while also fostering consensual policy making. The redistributive bias of PR systems is a price capital is willing to pay in exchange for the benefits accrued from consensus oriented regulatory policies (Cusack et al. Reference Cusack, Iversen and Soskice2007, Iversen Reference Iversen2005).

With this, VoC has indeed pushed firm-centeredness to the limits. Such a perspective seems certainly appropriate in the era of globalization, which has restored the power of firms vis-à-vis states and labor. But VoC has gone much further than merely highlighting the firm-centered nature of contemporary capitalism. It has also suggested that CMEs and generous welfare states are not dependent on any force – be it labor, the state or other social actors – that can counter capital, as the latter has all interest in preserving these institutions for reasons of efficiency. Thus, the floor is wide open for a reinterpretation of history. Viewed through the new lens, the emergence of labor-inclusive political economies does not require worker mobilization let alone class struggle, since generous welfare states are (co)built by, and partly for, employers. Taken to the extreme, what this adds up to is that labor, which has de facto been tamed in the brave new world of contemporary capitalism, is now also bound to be eradicated from the (political economy) history books – at least in so far as its role as a countervailing power in the history of industrialization, democratization and welfare state building is concerned. The jury is still out as to whether this interpretation will replace other perspectives, or whether an important point has been pushed too far (Kitschelt Reference Kitschelt2006).

Finally, VoC shares with much of contemporary political economy a normative concern about the future of CMEs which, in a world left bare of alternatives to capitalism, are often seen as representatives of its gentler and preferable forms. VoC conveys in a new form a comforting message to all of those who are worried about CMEs’ future in the light of globalization and the decline of labor. This reads that even when firms are placed at the center of analysis, CMEs are here to stay, since (save for the threat of financial deregulation) their tightly coupled national institutions are not bound to converge any time soon, especially as they perform equally well - albeit in different market segments and niches from those of LMEs. While some gladly subscribe to this view, it has provoked fierce debates in other quarters.

Empirical substantiation and level of analysis

VoC scholars assert that although both coordination types appear in every capitalist economy, their frequency tends to vary in a patterned rather than random way. Yet, some critics have raised the essential empirical question: what counts as sufficient proof for the existence of systematic variation in the frequency of coordination modes? They have argued that real-world economies cannot be reliably classified as LME or CME before data are provided on the “rough proportions” in which they display alternative forms of coordination in any given period of time. Unless the temporal changes in these incidences are captured, it is difficult to know whether an economy maintains membership in its original cluster or shifts to a different type (Crouch Reference Crouch2005). VoC accepts that the dominance of strategic hierarchical and network coordination in the CME political economy does not eliminate actors’ preference for market transactions in particular “sub-economies” defined by sector, locality, labor market segment, or firm size (Hancké et al. 2007, 21ff). Conversely, even in liberal market economies, hierarchical or network methods are likely to be preferred in cases of asymmetric market power common to oligopolies, oligopsonies, and certain supplier chains (Hall and Soskice Reference Hall and Soskice2001, 9, fn. 8).

Yet, if relative frequency rather than all-encompassing prevalence of market versus strategic coordination makes all the difference, then the substantiation of patterned variation becomes even more important. Since conclusive evidence on this aspect is in short supply, the effort of collecting more robust data cannot be spared by offering logically plausible solutions (Crouch Reference Crouch2005, p. 26).

The question of where to search for supportive evidence raises the problem of the proper level of analysis. Critics contend that VoC's national-level analysis is inapt to grasp the system's contemporary logics and driving forces. First, as long ago argued by Schmitter, changes in technology, market structure, and public policy have turned sectors and localities into some of the main sites “at which exchanges between producers are ‘socially constructed’ [and] public policies are ‘effectively administered’” (1990: 12-13). Along similar lines, the literature on business and social systems of production has found that firms’ transactions are governed in varied ways in different sectors, localities, and periods (Hollingsworth and Boyer Reference Hollingsworth and Boyer1997; Whitley Reference Whitley2007). Second, it has been equally convincingly put that currently many firms must coordinate their activities not in national but above all transnational or global contexts, within which a multitude of oligopolistic, oligopsonic, and perfect markets are incorporated into transnational production, commerce, and finance. Therefore, the global commodity chains’ interference with national capitalisms’ logics deserves serious attention – especially as they are characterized by multiple and permanently shifting loci and forms of profits, control, and coordination (Gereffi Reference Gereffi and Stallings1995).

The common thread underlying the above concerns is that, even if only market and strategic modes of coordination are assumed, the inference that any of these constitutes a whole political economy cannot be made unambiguously on grounds of their dominance on a single level. Rather, judgment on this matter hinges upon reasonable analytic procedures for aggregating the impact of varied coordination mechanisms prevalent in the multi-level worlds of capitalist economies.

Perils of institutional homogeneity and advantages of heterogeneity

To the extent that it exists, does complementarity originate in the homogeneity or heterogeneity of the core properties of national capitalist institutions? Are there good reasons for accepting that VoC's coherent ensembles promise superior performance to that of their hybrids or different capitalisms, as asserted by Hall and Gingerich (Reference Hall and Gingerich2004)? While the available evidence is inconclusive, critics contend that institutions might bring about greater aggregate returns precisely by virtue of their differences, which compensate for individual elements’ shortcomings in constituting the whole (Crouch Reference Crouch2005). The substantive issue involved is that organizing all key aspects of coordination uniformly by market or strategic interactive logics produces over-specialized environments, which are likely to run into trouble when facing certain challenges. Hence the contention that although VoC scholars take pride in proposing not just one but two best ways for capitalist success, they tend to overestimate the advantages of pure over hybrid regimes or, as Crouch has put it, of the “pedigree animal” versus the “mongrel” (2005, p. 362).

To highlight why convergence on homogenous coordination might turn into an obstacle to enhanced performance, it is helpful to re-state the key distinctive features of LME and CME using Hirschman's (1970) concepts of exit and voice. From this perspective, the first peculiarity of the VoC approach lies in the proposition that the institutional configuration of LME is uniformly geared towards perfecting the exit mechanism of recuperation from performance lapses. The CME architecture operates on the opposite principle, as it turns the exercise of voice and mobilized loyalty into routine remedies for deficiencies.

In concrete terms, in the LME, public information on impaired profitability prompts capital flight from firms with declining shareholder value. Top management has both the incentives and concentrated authority to react fast by firing workers and searching for new market opportunities. Deprived of organizational resources, labor is unable to put up significant resistance. At the same time, portable skills keep workers’ costs associated with forced exit and re-employment minimal, and make contentious collective action unnecessary. For firms and employees who fail to react fast by evasive action, there are drastic punishments in store. Public regulation clears the market and enforces reallocation of production factors through encouraging mergers, acquisitions, and hostile takeover. Meager welfare regimes do not allow workers to be overly “picky” when new employment opportunities are in sight. Finally, technology transfer is accelerated by licensing, market sale, and fluid labor markets easing the exit and new entry of R&D personnel.

In contrast, the CME keeps capital loyal via an instituted investor voice on matters of strategy and management. Vocal capitalists, rather than “voting with their feet”, use inside information to monitor, compare, and sanction performance. Such practices grant management grace periods for hammering out longer-term recovery plans to restore investor confidence not least through accommodation of vocal clients’, suppliers’, and labor's preferences. Management's incentive schemes and constrained authority contribute as much to the emerging pattern of longer time horizons, consensus building and compromise, as do labor's organizational strength, long employment tenure, and firm- and sector-specific skill structure reproduced by vocational training. Corporate leaders cannot get away easily with identifying redundant labor as the root cause of declining performance, but are forced to scrutinize a broader array of possible deficiencies. Further, generous welfare arrangements grant fired workers a measure of protection, which allows them to be vocal on the terms of acceptable new employment. Technology development occurs via collaboration among a multitude of public and private actors in standard setting, networking, and funding R&D.

Similar to authors of VoC, Hirschman enquires into the character of institutions that improve the efficacy of exit or voice. Yet he retains more skepticism about the ability of even the best designed institutions to make firms, organizations, and political economies immune to performance lapses altogether. Importantly, he argues that both exit and voice can be “overdone,” as their advantages are inevitably counterbalanced by corresponding disadvantages. Exit is individual, fast, and costless relative to voice, while the latter, whether its ultimate purpose is contention or deliberation, takes effort and time to evolve, and is subject to collective action problems (Olson Reference Olson1971). Yet, while voice conveys rich information, exit is less informative about the factors of decline and thus less helpful in devising optimal strategies for recovery (Hirschman Reference Hirschman1981, p. 244).

But if in the above sense “there is no free lunch in life”, it follows that declining efficiency is likelier to occur in homogenous than in properly mixed institutional environments. Hence the reason for Hirschman's interest in heterogeneous configurations, which leave room for occasional and/or partial recourse to voice in dominantly exit-prone environments and vice versa; allow for cyclical alternation between exit and voice as principal reaction modes; or facilitate institution building to achieve improved combinations (1970, p. 126). Yet, since all good things may not go hand in hand, and despite overall positive impact on efficiency, hybrids may prove less stable than pure arrangements. Hirschman admits that his approach “does not satisfy our craving for equilibrium, harmony, and final response” (1981, p. 237).

Against this background, a second peculiarity becomes apparent: VoC's polar types are assumed to combine efficiency with stability. In the original formulation at least, neither the LME nor CME depends for superior performance on occasional, partial, or cyclical adoption of its counterpart's recuperating mechanism, as tight coupling of their elements enables them to magnify gains and control costs simultaneously. Ironically, then, “free lunches” of largely costless adaptation are not only available but abundantly “served” in the VoC world, which is quite a surprising twist for an economics-driven reasoning. Hence ideals of perfect competition and unproblematic collective action appear to be close to the heart of VoC logics. VoC considers how agents would behave if key institutions were capable of bringing markets to ultimate perfection and making strategic coordination redundant. In turn, CME actors are perceived to respond to situations in which proper institutionalization of collective action helps transcend its “paradoxical” character.

On the above grounds, the VoC models – sometimes criticized for a-theoretically building types on the “generalization of the empirical properties of specific national models” (Crouch Reference Crouch2005, p. 33; Amable in Crouch et al. Reference Crouch, Streeck, Boyer, Amable and Jackson2005, p. 372) –, do appear to have a theoretical base that justifies their characterization as Weberian ideal types. Indeed, international trade theory complements their theoretical foundations. VoC draws on the Stolper-Samuelson theorem in which factors of production are perfectly mobile, and the Ricardo-Viner model that assumes physical and human capital to be specific to certain economic uses (Frieden and Rogowski Reference Frieden, Rogowski, Keohane and Milner1996; Hiscox Reference Hiscox2002). These models inform VoC on many of the issues to which we now turn.

The need for a theory of gradual transformative change

The weakness of the punctuated equilibrium model, whereby either the status quo is reproduced through changes to “more of the same” or erosion on one institutional dimension may decompose the whole system, has alerted comparativists about the need for conceptualizing a third type of “institutional change, including major change, short of a comprehensive redesign of the system as a whole” (Streeck in Crouch et al. Reference Crouch2005, p. 366). The implied research agenda has been vigorously pursued (Amable Reference Amable2003; Thelen Reference Thelen2004; Crouch Reference Crouch2005; Streeck and Thelen 2005; Deeg and Jackson Reference Jackson and Deeg2006; Amable and Palombarini Reference Amable and Palombarini2009; Hall and Thelen Reference Hall and Thelen2009), and has revealed that the Hall-Soskice approach is ill-suited to theorizing the missing link of endogenously driven transformative change. The rigorous effort to make any potential source of dynamism endogenous and thus subordinated to inherent equilibrating forces effectively turns VoC into an intellectual barrier to dynamic perceptions of society.

In the VoC world, public policies can be effective only to the extent that “they are incentive compatible, namely consistent with the kind of coordinating capacities embedded in the existing political economy” (Hall and Soskice Reference Hall and Soskice2001, p. 46). Firms respond positively to policies that improve the coordination mode they usually rely on, but resist imposition of alien mechanisms. This implies that the public authority lacks ability to act as an autonomous force of institutional change. Nor will transformative change emanate from the sphere of democratic politics. As argued by Wood (2001), economic actors and organizations will trust and support parties and their governments only if they can control them. Their veto power either stems from authority positions within parties and access to policy making centers, which make sure that their voices are heard, or from the exit option that allows them to punish “deviant” politicians.

Accordingly, Hall and Gingerich (Reference Hall and Gingerich2004, 32ff) invoke Hirschman's (Reference Hirschman1970) notion to characterize the distinct patterns of exit-prone versus voice-driven politics of class-conflicts (in the Stolper-Samuelson world) versus sectoral clashes (in the Ricardo-Viner world) generated by the two ideal types. Faced with exogenous shocks (including the unwelcome kind of state intervention), LMEs’ dominant mobile asset holders tend to exit from areas of declining returns and seek better opportunities, which provokes class conflict between capital and labor. In turn, CMEs’ specific asset holders are likelier to resort to voice in defense of the public policies favoring existing specializations, and align in a multi-class coalition against representatives of rival preferences. Yet, no matter how conflict-prone, interest group politics is unlikely to become a source of gradual and transformative change, since complementary institutions act to reproduce the dominance of mobile asset holders in LMEs and the rule of specific asset holders in CMEs. Furthermore, if a co-evolution between the LME and the Westminster system and CME and consociational or corporatist regimes is assumed, then the form of democratic institutions will also act in favor of the status quo.

As hinted above, recent literature has advanced the agenda of endogenizing the forms of democratic institutions suggesting that historically, the choices between majoritarian and PR systems can be traced to “the emergence of distinct capitalisms [that] predated PR” (Cusack et al. Reference Cusack, Iversen and Soskice2007, 1). In this account, later differences in electoral systems had been due to the prior existence of a densely organized local economy. Its manifestations included strong guilds with a monopoly of skill training that unions feared to challenge, the dominance of businesses that built on a skilled workforce, and nascent organizations of capital and labor predisposed to cooperation. As the industrial revolution required unified regulation of wages, labor control, social protection and skill formation, capital and labor were in search of a political formula that could extend their local collaboration to the level of the national political economy. PR proved to be that solution, since it allowed right-wing parties and their social base, capital, to gain from consensual national regulations even if left-dominated governments engaged in redistribution at capital's cost.

In contrast, in countries where dense local economic organization, strong guilds, and the implied pattern of collaborative relationships at the workplace level and beyond had been absent at the advent of industrial revolution, the right-wing had all the incentives to “nationalize” the overwhelming local tradition of non-cooperation. To achieve the desired outcome, left-wing parties had to be excluded by a majoritarian system that allowed the undivided right to “win and take it all.”

With this account, VoC has also started to address the question of how and why institutions have come to be complementary in the first place, an issue it had long been silent about. This line of reasoning ultimately rests on the suggestion that the embryonic forms of capitalisms which eventually became the LMEs and CMEs were present long before the industrial revolution elevated their characteristic local regulations to the national level. Further, it is also proposed that these local pre-industrial “VoC” had given birth to their own premature quasi-democracies – namely single member district electoral systems with and without consensus –, prior to the radical extension of franchise after the First World War. While judging its historical accuracy is beyond our competence, this vision of the “bottom-up” direction of capitalism's economic and political development offers a path-dependent solution for the puzzle of divergent trajectories of institutional change. From the very moment that factor-based and specific asset-based models are imputed into history, they set in motion a “perpetuum mobile” of systemic logics, which then allow LMEs and CMEs to survive as clear alternatives world wars, global economic crises, and political cataclysms.

Not all authors in the VoC tradition, however, seem to be content with putting ever more flesh on the framework's bare bones. Taking on board some of the criticism, Hall's recent work instead loosens up some of the original formulations. He has reinterpreted the VoC models “as institutional ecologies built up gradually over time”, whereby the configurations devised in one era shape “the challenges they were to face in later periods” (2007, p. 80). His view indicates a return to the idea of historical institutionalist pioneers Gerschenkron (1962 [1976], p. 26) and Shonfield (Reference Shonfield1965, p. 72), that institutions which had been part of the solution to problems in certain states and periods, might become a problem in other times and countries.

Hall also goes far in accepting that, since instances of abrupt shifts from one setup to another are rare, in the typical case even transformative change should be gradual. This being the case, the implied experimentation inevitably will involve mixing old with new elements. As a corollary, all transforming economies ought to rest on mixed rather than pure institutional configurations. But the question follows, how to distinguish transforming from “stable” economies except by reliable measures of change, and adequate analytic procedures that aggregate its advance in varied domains? Hall's study addresses the frequently raised concern that the doyen of CMEs, Germany, has had a difficult time in weathering globalization, and in some areas has drifted away from strategic coordination. Nevertheless, he argues, partial liberalization – such as the loosening of sectoral coordination in Germany –, should not necessarily corrode strategic action across the board: in certain areas it may even enhance its effectivity. The implied research agenda, then, “is to establish which kinds of moves towards more intense market competition erode existing capacities for strategic coordination and which are compatible with them” (2007, p. 78).

While this shift in the perspective certainly facilitates the study of social dynamics, it seems to be discomforting from the viewpoint of the VoC paradigm, as it effectively questions a key aspect of its original types. If Germany can drift towards a mixed economy with the newly adopted liberal forms not undermining but helping to preserve or enhance its strategic capacities, then how can one defend the original proposition that for mutual reinforcements among coherent institutions pure types tend to out-perform mixed ones? Should students of economic success still search for its secrets in homogenous market and strategic modes of coordination – or in their properly calibrated mixtures? Further, which are the mechanisms of calibration if the operation of equilibrating forces that provides for VoC's paradigmatic identity can no longer be assumed?

The concessions made to critics seem to challenge the VoC framework from a second related angle too. Hall finds that France and Germany do better than the UK in raising productivity and keeping inflation at bay, but are less successful in creating new employment in the service sector (ibid. p. 79, p. 82). At a closer look, then, the “free lunches“ promised by VoC all but disappear. VoC's original formulation is, of course, not wholly inimical to the notion of trade-off. After all, LMEs and CMEs are predicted to do well in radical versus incremental forms of innovation but not in both. Arguably, however, the successful pursuit of any of these agendas also represents absolute strengths relative to the rest of the world, which barely innovates at all. In contrast, the observation that superior inflation control or productivity come at the expense of inferior job-creation, seems to confirm the critics’ point that the pure types’ advantages are counterbalanced by corresponding disadvantages of over-specialization. Since it is far from evident that disinflation at the cost of joblessness or the rise of a service economy at the expense of meager wages and low productivity indicate superior performance in comparison to that achieved by mixed types, the whole debate on the ultimate superiority of pure over mixed cases is re-opened.

With respect to the question of institutional change, then, VoC seems to be facing a dilemma. On the one hand, by tracing the origins of LME and CME to their embryonic local manifestations in the distant past, Cusack et al. (Reference Cusack, Iversen and Soskice2007) come close to suggesting that their emergence is neither problematic nor particularly puzzling. Yet, the assumption that in some form the two varieties of capitalism had always existed – even long before the industrial revolution gave birth to capitalism as a system –, makes VoC vulnerable to the sarcastic objection that comparativists today tend to routinely but erroneously assume that “most things will basically remain the way they are” (Streeck Reference Streeck2009, p. 14). On the other hand, by allowing the approach to incorporate more dynamics, VoC faces the danger of losing its paradigmatic identity (ibid., pp. 19-20). Specifically, by allowing for the possibility that strengthening the liberal aspects of a CME may actually help preserve or even enhance the quality of its still remaining strategic coordination, Hall (Reference Hall and Hancké2007) questions the very rationale of the existence of the coordinated ideal type. If liberalization can do good to strategic coordination, then how to make sense of the fact that prior to globalization “institutional entrepreneurs” (Crouch Reference Crouch2005) paid so little attention to perfecting CMEs through larger doses of market coordination?

What capitalism?

Those concerned with how far the VoC models succeed in accentuating key attributes without losing sight of the systemic unity within which capitalist diversity ought to be studied, opened a new round of debate on the micro-foundations of the VoC approach and its inability to capture the systemic character of its types. In this vein, Amable was the first to doubt that the uncompromising market logic would ever produce a viable social model (Amable in Crouch et al. Reference Crouch, Streeck, Boyer, Amable and Jackson2005, p. 372). Recently, Polanyi's (1944 [1957]) major concern about a fundamental conflict between the operation of free markets and the essential human need for social stability has been invoked as theoretical support for a rejection of the VoC approach as a valid theory of the capitalist system (Streeck Reference Streeck2009, p. 249).

The gist of Streeck's criticism is that the approach fails to integrate the chief micro-motifs, fear and greed, which drive human action in a capitalist society. This also results in losing sight of politics “as an independent autonomous force […] not […] designed or devoted to bringing capitalist markets and transactions into equilibrium, or advance ‘national competitiveness’” (ibid., p. 251; see also Amable Reference Amable2003). Furthermore, by considering politics as endogenous to the economy, VoC fails to grasp the contradictory factors and logics of institution building and societal change. Unless fear and greed with all their consequences are factored in, the effort to theorize capitalisms’ dynamics remains a futile exercise, as no convincing theory of capitalisms can be elaborated without settling first on an adequate theory of capitalism.

It is not the case though that the VoC models are wholly inapt at taking these motifs seriously. Rather, they suffer from asymmetric weaknesses, which stem from downplaying the destabilizing potential of fear and greed, respectively. While hunger for profits is central to the LME logic, workers’ fears from dislocation or capitalists’ worries about their resistance are not assumed to play a constitutive role. Conversely, greed is given little room in the CME model, which is all about taming “the unruly restlessness of the model capitalist actor” (Streeck Reference Streeck2009: 256), and pre-empting contention through compromise.

VoC's optimism that anxiety about dislocation or insatiable hunger for profits can be controlled, can be traced to the salience attributed to human skills and the institutions of education, training, and socialization reproducing these skills. When viewed through the VoC lens, capitalism primarily appears in the variants of a knowledge-based economy. In LMEs, institutions of skill formation permanently replenish a pool of human capital that facilitates individual entrepreneurship and rapid structural change. Ordinary people have no problem in adapting, since they are educated and socialized to think and act like capitalists. In the CME, education, training, and instituted deliberation all conspire in favor of a different adjustment process. Here, capitalists (just as workers) are trained and socialized in thinking as collective entrepreneurs. Accordingly, both models vary on the theme of capitalism with a “human face”. In contrast to Polanyi's warnings of the destructive forces inherent in self-regulating markets, VoC's LME depicts dislocation as largely costless, while the CME virtually abolishes it as a serious threat. Unlike in the Marxian concept of “free labor”, workers are not seen as separated from but united with the factor of production that is central to the VoC logics: generally or specifically usable human capital.

In Streeck's view, these models are based on unrealistic expectations. Notwithstanding adequate education and training, workers will remain anxious about the “satanic mill” that undermines the stability of their existence by abruptly and permanently changing their professional and social status and identity (Polanyi 1944 [Reference Polanyi1957], p. 33). Conversely, capitalists making “a living by specializing in the subversion of social constraints” (Streeck Reference Streeck2009, p. 242), will not feel once and for all content with the social limits of their individualism. Rather, apart from exceptional periods, workers’ fear and capitalists’ greed will provoke defection from and/or revolt against any “order in equilibrium” (ibid., p. 246). It is this instability driven by the dynamic interplay between countervailing forces to which the trajectories of capitalist institutions are to be traced.

Institutions and actors beyond the nation state

VoC's discussion of the consequences of international regime building and globalization rests upon a limited investigation of these processes themselves. The approach is less concerned with globalization and European integration as such, than with refuting what it perceives as the conventional view of these processes. The rejected perspective is reminiscent of the early convergence literature (see above), and rests on three pillars. First, firms are similar across nations in terms of their strategic orientation. Second, predominantly motivated by lowering production costs, firms will relentlessly relocate their operations to cheaper sites. Third, the threat of exit puts governments under pressure to deregulate labor markets and lower social standards and taxes.

In our view, by grossly oversimplifying the globalization debate VoC misses an important dimension of contemporary capitalism, and the source of much institutional change. In recent years, a large body of literature has tried to come to terms with an emergent institutional and social reality beyond the confines of the nation state, and its repercussions on national institutions. Especially in the field of international political economy, authors have pointed to the emergence of social networks as well as transnational elite and class formation, and highlighted the multiple levels and asymmetric forms of power relations this process has entailed. More recently, globalization has also attracted the interest of institutionalist research. Its findings defy VoC's double claim that globalization today is mostly theorized in terms of footloose capital scanning the globe for cheap production sites, and that international regimes can be controlled by national actors who seek to maintain domestic institutional advantages. In the following we outline the building blocks of an institutionalist approach which allows for a richer account of globalization and European integration. We start with the work of Djelic and Quack (2003), who see globalization as a deeply inter-twined double process of institution-building on the transnational level and institutional change on the domestic level. Their view has a number of advantages.

First, it captures the historical novelty of contemporary globalization by arguing that, in comparison to earlier phases of transborder expansion, it is much more embedded in a dense institutional structuring of the transnational space. Second, this perspective offers a nuanced understanding of how transnational institutions have come about. Three processes are distinguished. In the “dominant mode”, transnational institutions reflect the power and preferences of a dominant nation. Thus, many international regimes created after the Second World War bear the mark of US American design. A second mode is “negotiated”, where transnational institutions result from confrontation and bargaining among different, state and non-state, actors. Many EU institutions and legal acts stem from this mode. Finally, the “emergent mode” rests upon a long-term transnationalization process that has created actors whose identities and interests are no longer exclusively shaped by national, but also by transnational institutions. This mode has advanced within transnational contexts, which had their origins in the dominant or the negotiated mode, and led to the new regimes’ partial emancipation from national interests and institutions.

Third, globalization is systematically linked to institutional change. Djelic and Quack view globalization as a “circular interplay” among actors, rules, and processes of institution building on transnational, national, and sub-national levels (2005, p. 9). The interaction is likely to result in institutional change when internal challenges or opportunities come up against external triggers for change and can draw on alternative institutions. Faced with new challenges, domestic actors are unlikely to confine themselves to seeking solutions within the domestic arena, but also exploit the opportunities provided for on the transnational plane. In line with some of the above reviewed literature on institutional change, Djelic and Quack argue that the resulting changes are gradual rather than radical, but highly consequential nevertheless. Unlike in VoC, transnational institutions are conceived as alternatives to “incumbent” national institutions, since institutional entrepreneurs make use of the available or actively contribute to the adoption of new setups.

This brings us, fourth, to the question of actors. Djelic and Quack do not only stress the role of institutional entrepreneurs who might be interested in institutional changes but also address specifically the question of transnational actors, especially corporations, operating from within the nation state. In contrast to VoC that sees firms merely engaged in cross-country institutional arbitrage, Djelic and Quack acknowledge the possibility that these seek to alter the host country's institutions, in order, for instance, to enhance market opportunities. Accordingly, domestic firms that have gone transnational are also likely to promote institutional change at home.

Liberalization, exit and voice

Although Quack and Djelic stress the link between globalization and institutional change, they are less articulate about its direction. However, there seems to be little doubt in the literature that change gravitates towards liberalization and strengthening of market forces with supranational institutions fostering rather than mitigating this trend globally and in Europe (Scharpf Reference Scharpf1997; Bartolini Reference Bartolini2007; Streeck Reference Streeck2009). How does the general trend towards liberalization affect the micro-logics of exit and voice, so crucial for reproducing the varieties of capitalism? As we saw above, according to VoC, firms in LMEs react to the external shock of globalization by resorting to exit while in CMEs they use voice to help recovery of domestic institutions.

With this reasoning VoC departs, however, in important ways from Hirschman's key intuition about a negative association between exit and voice, namely that the “presence of the exit alternative can […] atrophy the development of the art of voice” (1970, fn. 1, 43). The emergence of voice as a dominant “coordination” mechanism on the macro-level is inextricably linked to the creation and maintenance of boundaries, which raise barriers for exit. Historically, the formation of the territorial nation state locked-in economic and social actors, and thus shifted their preferences towards expressing their grievances through voice. In turn, to the degree that European integration and globalization abolishes existing economic and legal boundaries, we are likely to witness an increased role of exit across the board (Bartolini Reference Bartolini2007).

No longer “sentenced” to loyalty to the national economic space and to fixing problems of domestic institutions by voice, firms can more easily exploit evasive options. Most prominently figure enhanced opportunities to physically exit the home country by producing abroad. In addition, European integration opens the possibility to explore alternative regulations, material, and jurisdictional resources without actually physically moving (ibid.). Firms can for instance directly access global financial markets, recruit skilled workforce abroad and thus dispense with the effort of domestic skill formation, or appeal to the European Court of Justice in order to enforce national change. All in all, the horizontally and vertically extended opportunities for institutional arbitrage are much more multifaceted than envisaged in VoC.

Firms’ magnified opportunities for exit, in turn, have repercussions for “the art” of voice. As the exit options are unequally distributed in that they favor resourceful players, political actors are likelier to become attentive to them than to others who in the absence of easy alternatives will stay put. In this context, voice is becoming overall less efficient a mechanism for remedying institutional lapses. Even actors who are capable and supportive of voice might learn to prefer the exit option. It is not difficult to see how this snowball effect can lead to reinforced institutional change, and to predict its political impact.

What this all amounts to is that comparativists cannot spare the effort of incorporating globalization and European integration into their analyses of capitalism and its varieties. The capitalist world is adopting a new shape in which economic and social actors are confronted with and engaged in creating multiple levels of institutions. While the national level might still be dominant, its systemic stability and autonomy is being disrupted by new exit opportunities and weakening effectiveness of voice on multiple levels. This has two consequences for the study of capitalist diversity. First, when it comes to varieties, the question of change and instability resulting from multilevel interactions has to come to the fore. Institutionalists may still be arguing about whether partial liberalization enhances the effectiveness of strategic coordination in other parts of the economy or rather corrodes them. But they cannot afford to ignore the unraveling of established institutions and behavioral patterns that globalization and European integration have brought about. The golden age, as it were, lies behind us. Second, when it comes to studying capitalism, its transnational and structured expansion cannot be treated as an exogenous shock, but has to be understood as an inherent tendency to capitalist accumulation, and thus made part and parcel of the analysis of national (and sub-national) levels.

Extending the framework eastwards

Globalization, finally, has also raised the question of the rise of new dynamic economies outside the capitalist core on the agenda. Under the auspices of foreign capital and technology, a number of peripheral economies have recently started to upset traditional patterns of international division of labor, making their inroad into international specializations formerly considered typical for LMEs or CMEs. A most prominent example is that of the East European newcomers to the EU. To what extent can these political economies be fruitfully analyzed in the VoC framework?

Over the last years, a number of authors have indeed tried to stretch VoC to Eastern Europe, looking for similar types of capitalism there. Following the lead of Hall and Gingerich (Reference Hall and Thelen2004), one study has applied factor analysis to map the post-socialist world, and found that Belarus, the Ukraine, Slovenia, and Croatia are the most coordinated, and Russia, Estonia and Armenia the most liberal market economies (Knell and Srholec Reference Knell, Srholec, Lane and Myant2007). Given that more than half of these countries just barely (if at all) qualify as market economies, this finding does not strike us as particularly meaningful. Other studies, only looking at East-Central Europe, claim a convergence on the CME type (McMenamin Reference McMenamin2004; Lane Reference Lane2005), or, to the contrary, on LMEs (Cernat Reference Cernat2002). While still other researchers see the region populated by mixed types of capitalism, which combine features of both LMEs and CMEs (e.g. Mykhnenko Reference Mykhnenko and Hancké2007). All in all, this suggests confusion when it comes to applying VoC's core types to post-socialist capitalism.

On more substantive grounds, Buchen (Reference Buchen, Lane and Myant2007) and Feldmann (Reference Feldmann and Hancké2007) have found that the institutional configurations off two East-Central European countries bear resemblance to those of advanced Western economies, with Estonia qualifying as LME, and Slovenia as CME. Yet, these attempts to export VoC categories to a region outside of the OECD world have their own flaws. On the one hand, even if the borrowed approach might capture some aspects of two extreme cases, the overwhelming majority of countries in the region still defy a classification along VoC's ideal types. This raises questions about the specificities which set Estonia and Slovenia apart. On the other hand, it is doubtful whether Estonia can really be classified as LME, as many of its features seem to be at odds with the ideal type. For example, liberal institutions have hardly been able to shape the country's industrial specialization the predicted way. Market radicalism has not led to radical innovation but to specialization in low-tech, or the low-skill segment of high-tech, sectors.

Is it by chance that the attempt to extend VoC eastward has not brought about convincing results so far, or is it the approach's assumptions that make it difficult to capture the character of capitalism outside the core? We argue the latter is the case. Most importantly, VoC's “dual convergence thesis” (Hay Reference Hay2004, p. 235) – also challenged in its validity for advanced capitalist countries – seems to be of limited use when it comes to analyzing societies outside the OECD world. As well established in earlier literature, the spectacular instances of late development did not originate in the emulation of, but rather deviation from, the frontrunners’ institutions and ideas. Geschenkron (1962 [1976]) drove home this point by analyzing France's and Germany's success in catching up with Great Britain. This account can be reconciled with VoC on the basis that the emergence of CMEs was due to the timing of industrialization and the specific sectors through which it advanced.

If that is the case, however, then VoC has to allow for more than just two types of capitalism. After all, the rising “Rest” of Asian, Latin American, and most recently East-Central European countries have challenged Western capitalism without converging on either of its two types (Amsden Reference Amsden2001). Rather then looking for convergence with either LMEs or CMEs, comparative political economy should therefore be interested in the specificities of the newcomers to global capitalism. It is not by chance that many East Europeanists remain convinced that the region's institutional configurations are truly unique (Stark and Bruszt Reference Stark and Bruszt1998; Bohle and Greskovits Reference Bohle and Greskovits2007; Nölke and Vliegenthart Reference Nölke and Vliegenthart2009).

The dual convergence thesis is challenged in yet another way. More than in the case of other latecomers, East-Central European institutions have been thoroughly shaped by transnational actors, as the majority of the post-socialist institutions were not yet in place before these economies became exposed to global and European influences. In order for VoC's dual convergence thesis to hold, transnational actors should be actively engaged in recreating the setup to which they have been accustomed. The literature tells otherwise. Especially transnational firms originating in CMEs are using the newly available exit options to escape the institutional configurations of their home country, and to experiment with new arrangements in the host countries which do not reflect much of their CME origins (Bluhm Reference Bluhm2007). For instance, CME-based TNCs have hardly urged the establishment of negotiated industrial relations beyond the firm level in East-Central Europe, or backed the post-socialist welfare states’ efforts to provide them with skilled labor forces.

This in turn can give rise to a new research agenda, which is concerned with the emergence of institutional complementarities across rather than within countries. For instance, building on Nölke's and Vliegenthart's (2009) characterization of East-Central European capitalism as a distinct third type, a “dependent market economy” (DME), it can be asked whether the interaction between CMEs and DMEs entails complementary rather than comparative institutional advantages. In this case, DME's institutions would foster specialization in low cost and relatively low-skilled segments of the production chain, allowing CMEs to deepen their own specialization in the high-skill and high cost production segments.

More realistically, however, firms’ relocation to alternative spaces of regulation is likely to intensify pressures for further liberalization in CMEs, without leading to the construction of coherent models of capitalism in the new locations. VoC, even in its amended form which incorporates more than two types, lacks the conceptual tools to grasp the dynamic and interdependent process of relocation, restructuring, and social dislocation resulting in unstable and contested institutional configurations.