Conceptualizing Cooperation

Conceptualizing Cooperation on Neo-Corporatist Foundations

Contemporary debates over economic cooperation developed in reaction to the neo-corporatist scholarship of the 1970s and 1980s. In sharp contrast to pluralist competition between relatively fragmented social groups in the United Kingdom and the United States, neo-corporatism was defined as a system of interest intermediation characterized by large, centralized, compulsory, hierarchical, and encompassing producer associations (Schmitter, 1974). Conceptualized in an era of acute, distributional conflict, these hierarchical and monopolistic interest associations were hypothesized to behave more “responsibly,” because they could internalize the cost of their demands (Calmfors & Driffill, 1988, p. 15). In practice, neo-corporatism was associated with “generalized political exchange” (Pizzorno, 1978) in which governments traded countercyclical fiscal policies or more generous social benefits for wage restraint and industrial peace (Cameron, 1984, p. 170). As more Western European countries embarked on corporatist experiments, the concept of neo-corporatism was stretched to encompass a wide range of distinct phenomena, from joint policy formulation (Lehmbruch, 1984) to collaboration in production (Streeck, 1991), and even a general spirit of consensus (Katzenstein, 1985). Whereas many countries, from the United Kingdom to Italy, flirted with non-liberal patterns of cooperation (Regini, 1984), the Scandinavian countries—Denmark, Norway, and Sweden—proved particularly successful (Hall, 1999). ⁵ Here, encompassing producer associations settled conflicts in a wide array of domains from industrial relations to policy making (Pekkarinen, Pohjola, & Rowthorn, 1992).

By the 1990s, however, rapid technological change, economic internationalization, tertiarization, and European integration challenged neo-corporatism in Scandinavia and illustrated the hazards of conflating distinct empirical phenomena (Streeck, 2006). Most notably in Sweden, these changes appeared to generate discord rather than consensus among hierarchical interest associations. Employers, for example, abruptly withdrew from centralized collective wage bargaining and other tripartite boards in 1991 (Pontusson & Swenson, 1996). At the same time, cooperation flourished in countries with smaller, less centralized interest associations such as Germany (Soskice, 1990) and Ireland (Rhodes, 2001). Because the literature on neo-corporatism conflated centralized structures with cooperative processes, it was poorly positioned to explain these puzzling developments (Molina & Rhodes, 2002). Many scholars consequently shifted their focus away from the structural attributes of interest associations, especially labor, to the relationship between them. They did so, however, in different ways, with some analysts focusing on concertation or cooperative economic policy making (Baccaro, 2003) and others examining coordination or collaborative production strategies (Soskice, 1990).

Scholarship on concertation evolved directly out of neo-corporatist analysis (Lehmbruch, 1979) and has focused primarily on social policy formulation and implementation (Cameron, 1984; Ebbinghaus & Hassel, 2000; Regini, 1984). Scholars have also studied concertation in reforms of labor market regulation, housing policy, taxation, and even collective wage bargaining as an “incomes policy” tool supporting macroeconomic adjustment (Hardiman, 2002; Rhodes, 2001; Thomas, 2003). Given its intellectual lineage, scholarship on concertation shares with research on “organized capitalism” (Höpner, 2007; Streeck, 2009) and “macrocorporatism” (Martin & Swank, 2012) a concern with the influence of political (rather than market) forces on managerial decision making, but it has more resolutely focused on policy-making processes rather than organized interests’ structural attributes. ⁶ Thus, although the earliest work on concertation emphasized that policy makers could ameliorate class conflict by incorporating large trade unions into policy making, evolving scholarship acknowledges the important roles that less encompassing producer organizations (Baccaro, 2003) and non-class-based associations (O’Donnell, 2001) can play. As demonstrated by successful concertation in the Irish case, the organization of actors within associations and the institutionalization of access to multiple policy domains can produce patterns of political engagement that differ significantly from the competitive lobbying that predominates under pluralism such as in the United States (Hacker & Pierson, 2011). At the same time, the outcomes of concertation contrast with those from earlier neo-corporatist adjustment in allowing cooperative patterns of policy making to enhance rather than merely stifle market competition.

Other scholars have instead focused on cooperation in the act of production. Reinterpreting the economic success of the German case, the literature on coordination originally sought to understand how enterprises organize collective action without relying on large, centralized producer associations (Soskice, 1990). ⁷ Inspired by new institutional economics, the Varieties-of-Capitalism (VoC) literature examines how firms in “coordinated” market economies (CMEs) support institutional arrangements that facilitate cooperation—among themselves and with other actors—in providing such collective production inputs as specialized equipment, human capital, technology, and common standards (Hall & Soskice, 2001; Streeck, 1991). ⁸ Although more sensitive to enterprise- and industry-level cooperation than concertation, the coordination concept is nonetheless distinct from “micro-corporatism” (Teubner, 1988) and “meso-corporatism” (Wassenberg, 1982) in its emphasis on cooperative processes. Moreover, it differs from both corporatism and concertation in its conceptualization of the drivers behind these processes. Rather than emphasizing socio-political pressures, coordination is driven by businesses’ self-interest in leveraging non-market production governance for higher micro-economic efficiency. Depending on the employers’ preferences, coordination can center on different loci—including the firm, sector, or nation—and involve a variety of collective actors such as employers’ own associations, trade unions, or banking blocs (Kitschelt, Lange, Marks, & Stephens, 1999). Before systematically comparing these two processes of concertation and coordination, Table 1 recapitulates the two concepts’ relationship to both theorizing on classical neo-corporatism and liberalism, with country examples given as illustration.

OPEN IN VIEWER

Table 1.

Patterns of Cooperation.

		Cooperation in policy-making
		Institutionalized	Limited
Cooperation in production	Institutionalized	Neo-corporatism (pre-1990s Sweden) a	Coordination (Germany)
	Limited	Concertation (Ireland)	Liberalism (United Kingdom)

a

As described below, Sweden became less corporatist during the 1990s, and patterns of cooperation more closely resemble Germany today.

Although scholars of concertation and coordination have been careful to highlight how each concept differs from both neo-corporatism and liberalism (Baccaro, 2003; Hall & Soskice, 2001), potential contrasts between the loci and dynamics of concertation and coordination remain underexplored. ⁹ Scholars of concertation are highly sensitive to the economic implications of policy cooperation, but seldom examine cooperation in the act of production itself. For example, whereas Avdagic, Rhodes, and Visser (2011); Baccaro and Simoni (2008); Regini (1984); and Rhodes (2001) discuss how wage restraint and labor market reform enhance productivity, they do not examine inter-firm or industry-labor cooperation in the research, design, assembly, or distribution of goods and services.

Early work on coordination, by contrast, either omitted employer participation in policy making from measures of coordination or viewed it as merely a mechanism to protect asset-specific investments and lock-in the benefits of coordination (Hall & Gingerich, 2009; Hall & Soskice, 2001). More recent research in the VoC tradition has acknowledged that policy process integration may be conceptually distinct from cooperation in production and more sensitive to state intervention (Hassel, 2007; Martin & Swank, 2012). However, VoC authors tend to view all instances of cooperation through the conceptual lens of coordination. Largely retaining a focus on employers’ desire to solve collective action problems, they have broadened the scope of actions originally associated with the concept to include national-level politics. Paying less attention to how patterns of cooperation might differ between realms of collaboration, these scholars have emphasized that coordination processes at different levels of analysis (national versus firm level) display distinct dynamics (Martin & Thelen, 2007). Although this analytic broadening of coordination as a concept allows the VoC literature to talk about a wider range of phenomena, it comes at the potential cost of lower explanatory leverage for its core analytic principle.

A third group of scholars acknowledges the distinct dynamics of different types of cooperation (Hicks & Kenworthy, 1998; Höpner, 2007; Streeck, 2009; Thelen, 2012), but work in this vein could do more to integrate the concertation literature’s findings. Constructing a two-dimensional framework, these authors introduce “organization” (Höpner, 2007; Streeck, 2009) or “social solidarity” (Thelen, 2012) as a second analytic category to complement the analysis of coordination. However, measures of these new concepts retain neo-corporatism’s focus on the structural attributes of interest organization, emphasizing unionization rates and associated patterns of group incorporation. As a result, these approaches provide little leverage for understanding the breakdown of policy concertation in highly coordinated Sweden and its emergence in liberal Ireland. Moreover, these approaches disagree about the trajectory of coordination, in part due to a differential focus on class-based (Streeck, 2009) and intra-class (Thelen, 2012) conflict. We argue that more attention to the dynamics of concertation could illuminate such puzzling outcomes and create new analytic space to conceptualize changing patterns of coordination. More specifically, scholars could avoid coding coordination along a one-dimensional continuum, acknowledging that coordination, although still widely used by businesses to solve their collective action problems, has become more liberal in character, having expanded managerial prerogatives in setting the terms of employment (Baccaro & Howell, 2011). In turn, one could examine how changing patterns of coordination between companies, from organized decentralization in collective wage bargaining to modularization in training programs, have changed the power distribution between labor and capital on the firm level, underwriting shifts in the character and outcomes of local concertation processes. ¹⁰

The contrasting dynamics of the two cooperation processes become obvious when we compare concertation and coordination more systematically. On the causes of concertation, we argue that this form of cooperation is relatively easy to establish. Of course, concertation is neither a natural nor a universal phenomenon. For instance, in pluralist societies such as the United States, fragmentation among weak employer associations and trade unions inhibits collective action and tends to translate into governments receiving fewer benefits from engaging producer associations (Avdagic, 2010; Baccaro & Simoni, 2008; Hassel, 2007). From the perspective of firms, however, the costs of concertation are relatively modest. Although enterprises expend resources to cooperate in social, tax, and incomes policy and may be adversely affected by higher taxes and burdensome regulations, concertation does not systematically undermine managerial autonomy and rarely requires firms to share sensitive information about revenue, remuneration, their equipment, their products, or their technologies. ¹¹ Moreover, in the contemporary neoliberal political climate, businesses can often expect major concessions from labor or the government through “social pacts,” which may make concertation less financially demanding for businesses than coordination.

The state plays a central role in concertation. This is not surprising as policy formulation and implementation are state prerogatives (Martin & Swank, 2012). But the lower barriers to employer participation in concertation also enable governments to shape concertation using a variety of instruments. For example, policy makers can broker agreements by offering generous side payments (Regini, 1984), confronting stakeholders with the threat of unilateral reform (Rhodes, 2001), or merely engaging in a constructive dialogue (O’Donnell, 2001). Although only those governments that are not powerful enough to favor unilateral action are prone to engage in concertation (Avdagic, 2010; Baccaro & Simoni, 2008), concertation processes benefit from a relatively strong, capacious executive that can credibly promise to reward or punish its bargaining partners (Regan, 2012; Rhodes, 2001; Visser & Hemerijck, 1997). ¹² Consequently, concertation is also highly sensitive to shifts in government partisanship (Avdagic, 2010; Hassel, 2007; Regini, 1984).

Closer attention to the consequences of concertation, however, highlights its limits. Although policy makers can use concertation to manage distributional conflict and facilitate legislative reform, they have struggled to leverage cooperation in policy making for investment in high-quality productive inputs like specialized equipment, risk capital, and skills (Hall, 1999). Government bureaucrats simply lack the information to unilaterally invest in specialized, supply-side resources (Culpepper, 2003), and businesses will only commit to the provision of collective goods when supported by additional institutions in the realm of production (Hall & Gingerich, 2009; Soskice, 1990). As a result, concertation influences production only indirectly, by securing industrial peace (Cameron, 1984), promoting wage restraint (Hancké & Rhodes, 2005; Hassel, 2007; Regan, 2012), or facilitating market-oriented reform (Baccaro & Howell, 2011; Rhodes, 2001). As concertation’s restricted scope tends to preclude the construction of more durable forms of comparative advantage that could generate businesses’ self-interest in cooperation, employer engagement typically remains dependent on either public pressure or explicit buy-ins, leaving concertation open to easy destabilization.

Comparing the dynamics of concertation with the causes and consequences of coordination further clarifies the distinctive character of each form of cooperation. The institutional preconditions for cooperation in production are significantly higher than they are for policy making. Typical features of coordination, including cross-shareholding and long-term relationships with large banks, labor representation on the supervisory board, and shopfloor representation, all impinge on managerial autonomy in significant and systematic ways. Moreover, even less structured forms of collaboration on pay, human capital formation, technical standards, and innovation require firms to share highly sensitive information about their work processes and product strategies with competitors. To the extent that collaboration requires firms to make significant, asset-specific investments in production, the barriers to cooperation are even higher. Coordination thus requires strong interest, exceptionally high levels of trust and a dense network of complementary institutions to monitor and sanction opportunistic behavior (Hall & Gingerich, 2009; Soskice, 1990).

As a result, coordination is extremely difficult to create. Successful instances of contemporary coordination can often be traced back to historical preconditions such as a robust guild system (Iversen & Soskice, 2009) or critical junctures like industrialization (Ornston, 2012a) and democratization (Martin & Swank, 2012). Of course, the specific elements of coordination may change, even significantly, over time (Thelen, 2004). Efforts to create coordination de novo, however, are rare. Contemporary examples are often limited to the regional or local level, where interpersonal networks facilitate trust-building (Culpepper, 2003; Herrmann, 2009; Sabel, 1996). ¹³

Relatedly, governments play a more limited role in this process. Although the governments can support coordination by subsidizing private sector associations (Streeck, 1997) and serving as employers in their own right (Martin & Thelen, 2007), it is difficult to mandate coordination (Hall, 1999). In sharp contrast to processes of concertation, a powerful, autonomous executive may actually undermine coordination by making firms more reluctant to share information and enter into long-term commitments. As a result, collaboration in the realm of production is more commonly associated with fragmented, veto-prone governments (Cusack et al., 2007; Hall & Soskice, 2001). ¹⁴ Although firms in CMEs may pressure governments to invest in social programs to protect asset-specific investments in human capital (Estevez-Abe, Iversen, & Soskice, 2001), we contend that coordination is not particularly susceptible to shifts in government partisanship (Streeck, 2009; Thelen, 2012).

Examining the consequences of coordination further clarifies why cooperation in production is strongly path-dependent. In addition to requiring low levels of state agency to establish and maintain, employers’ interest in coordination grows as it is institutionalized (Hall, 1999; Hall & Soskice, 2001). Once the initial steps of coordination have facilitated collective investment in the provision of high-quality inputs, firms will be able to defend high value-added niches (Hall & Soskice, 2001; Streeck, 1991). Having long sustained employer interest in sustaining coordination, corporate success in quality-based, upmarket movement can insulate societies from many disruptive effects of cost competition and radical technological innovation. Admittedly, as the share of industry in the economy has declined and production has become further cross-nationally integrated in recent years, the breath (rather than the depth) of coordination has weakened in some CMEs. However, it has done so only gradually (Höpner, 2007; Thelen, 2012) and in stark contrast to the dramatic breakdown of concertation in Italy (Regini, 1984), Sweden (Pontusson & Swenson, 1996), and Ireland (Regan, 2012).

The conceptual distinction between processes of concertation and coordination can resolve continuing scholarly disagreements about the causal processes associated with cooperation. Often, continuing debates are simply the product of analysts choosing different conceptual lenses. VoC scholars may thus be analytically primed to overlook how liberal economies have leveraged the productivity-enhancing benefits of policy cooperation in social pacts to facilitate market-oriented reforms (Hall & Gingerich, 2009). Similarly, as analysts of concertation have pointed out (Baccaro & Howell, 2011), VoC scholars’ conceptual filters may prompt them to downplay the neoliberal consequences of episodes of policy concertation in CMEs. Conversely, scholars of concertation are naturally inclined to disagree with the VoC literature’s emphasis on path dependence (Hardiman, 2002) or criticize it for “forgetting” the state (Hemerijck & Vail, 2006). Although the latter charge is true to the extent that the VoC literature has been slow to acknowledge the distinct causes and consequences of policy concertation, it should not distract from the VoC literature’s insights about the state’s limited role in fostering coordination and the process’ path-dependent characteristics.

These differences would likely be less consequential if concertation and coordination emerged in tandem. The preceding analysis, however, suggests that concertation and coordination are only loosely coupled and exist in tension. The same conditions that facilitate concertation, such as a strong executive, may discourage firms from sharing information with policy makers or one another (Hall, 1999). Conversely, the well-organized societal actors that underpin coordination (Streeck, 1997) may use their entrenched position to resist concertation (Ebbinghaus & Hassel, 2000). In fact, concerted policy reform may prove less necessary to the extent that societal actors can autonomously organize production in efficient and effective ways. For example, countries with high levels of labor market coordination are less likely to strike social pacts as societal actors can autonomously restrain wages (Hancké & Rhodes, 2005; Hassel, 2007) and rely less heavily on cost-based competition (Streeck, 1991). Conversely, successful cost-based competition may hinder coordination by reducing firms’ interests in high-quality inputs (Ornston, 2012b). This is not to argue that concertation and coordination never coincide. The Nordic countries continue to feature both to a high degree. But we suggest this overlap is less common and less stable than is usually assumed (Hall & Gingerich, 2009; Hall & Soskice, 2001; Hicks & Kenworthy, 1998). In Scandinavia, for example, levels of concertation have declined visibly over the last two decades.

In the next two sections, we examine our claims about the processes of cooperation in two diverse cases. Germany, a country with a high level of coordination but low concertation, is frequently viewed as an archetypical CME (Hall & Soskice, 2001). Ireland, by contrast, ranks poorly on measures of coordination (Höpner, 2007), but has relied heavily on policy concertation to manage economic adjustment (Thomas, 2003). Consistent with our theoretical claims, the case studies demonstrate that concertation was more amenable to state intervention and easier to create, but did not necessarily lead to upmarket movement and proved less stable. By contrast, coordination, a usually profoundly path-dependent process, evolved with limited state intervention but resulted in a more durable form of comparative advantage that has blunted institutional change. The analysis is complemented with brief comparisons to other European countries, whose experiences help illustrate how these causal relationships hold beyond the sample. These shadow cases include the United Kingdom and France as countries having long featured only limited cooperation in both policy making and production, as well as the Nordic countries, which have historically displayed high levels of both.

Germany: The Politics of Coordination

Having provided much of the inspiration behind the literature on the varieties of capitalism, Germany’s high levels of coordination are widely acknowledged (Baccaro & Howell, 2011; Hall & Gingerich, 2009). Although the country’s society—particularly the employer camp—has long been highly organized (Höpner, 2007; Katzenstein, 1987) and involved in social policy implementation as a form of “welfare corporatism,” Germany’s history of policy concertation has been checkered. For instance, between 1967 and 1977, arguably the high point of successful corporatist adjustment in neighboring countries, Germany’s “concerted action” (Konzertierte Aktion) never developed into a full-blown corporatist forum and rode mainly on the shoulders of social democratic politicians. Having assumed government responsibility for the first time after the end of World War II, the Social Democrats were keen to break with the country’s postwar ordoliberal orthodoxy supported by previous center-right governments and leverage policy cooperation for implementing Keynesian ideas. However, conflicts between the trade union movement and the Social Democrats in government ran high (Streeck, 2006), and wage moderation was achieved instead by effective coordination (Hall, 1994).

Coordination spans many areas of the German economy, including a cooperative system of industrial relations. In contrast to state-led concertation in Irish wage negotiations—which often required government side payments to succeed—the German social partners reach agreements at the regional sectoral level without state interference (Tarifautonomie). Moreover, at the company level, the workforce enjoys extensive rights of co-governance in works councils and supervisory boards. Given the high degree of collective agreements’ coverage, the system has effectively suppressed wage-based competition within sectors and allowed businesses to invest in their employees’ skills without having to fear that competitors would reap the benefit through worker poaching. Moreover, in the vocational education system, businesses train to jointly agreed standards (Thelen, 2004). Although unions and the state also play important roles in this collective skill system, policy makers have been careful to accommodate business preferences so as not to endanger their participation. In turn, the German state has been far less proactive than other European public authorities, and the Federal Institute for Vocational Education and Training (Bundesinstitut für Berufsbildung, (BIBB)) focuses on providing research and advising. ¹⁵ Finally, when it comes to accessing capital, many companies continue to strongly rely on financing from their house banks (Deeg, 2010).

The roots of cooperation in production reach far back to industrialization during the 19th century, when society’s self-organizing efforts shaped decentralized regional cooperation and enabled processes of creative and strategic institutional re-composition. Horizontal connections between networks of companies sustained extra-firm arrangements and policies for training, competition, finance, and dispute resolution that allowed clusters of small- and medium-sized specialized producers to prosper on the basis of reformed versions of pre-industrial craft production (Herrigel, 1996). Moreover, during and after World War I, larger industrial companies started to collectively organize, both to reduce the ruinous aspects of competition among them and to better position themselves for dealing with the increasingly powerful labor movement (Feldman, 1981). Throughout this process, the relative weakness of Germany’s national public authorities vis-à-vis the Prussian bureaucracy allowed business cooperation to strengthen at the sectoral and regional levels (Martin & Swank, 2012). These high levels of organization in turn allowed companies of varying sizes to settle on common approaches to business challenges and cooperate with labor in a plethora of increasingly democratized parapublic institutions. With the social partners playing an important role in the social insurance funds and the federal labor office, the country’s highly organized society left the decentralized state effectively “semi-sovereign” after World War II (Katzenstein, 1987). In turn, scholars characterized cooperation in Germany as “sectoral corporatism,” “corporatism without the state,” and “horizontal coordination” (Manow & Seils, 2000, p. 140).

These high levels of coordination facilitated “high skills, trustful cooperation at the workplace, flexible internal labor markets, rapid adjustment to new technology, [and] successful technology transfer to small and medium-sized firms” (Streeck, 2006, p. 22). With the associated labor rights effectively foreclosing cost competition, companies embraced patterns of “diversified quality production” (Streeck, 1991). In addition, they went to great lengths to stabilize the high level of coordination necessary for such “high-road” competitive strategies through leveraging their power in the social insurance funds and labor offices (Manow, 1997).

Successful coordination among the social partners, however, has translated into disinterest in concertation. Moreover, the weakness of the semi-sovereign German state—which had facilitated the growth of coordination—prevented it from committing the social partners to productive participation in negotiations about policy reforms. This became obvious when consecutive governments during the mid-1990 were unable to enlist business and labor in tripartite negotiations about countering the growth in unemployment post-unification. Nine “Chancellor-rounds” under the center-right Kohl government produced no initiatives; and neither did even more elaborate and institutionalized attempts at concerted reform under the label of an “Alliance for Jobs” later in Kohl’s tenure and under the early center-left Schröder government. The social partners felt no need to legitimize themselves through participation in concertation given their continuing and highly effective coordination; and policy makers had little leverage to make them sign up (Schulze-Cleven, 2009). Thus, eventually, Schröder decided to unilaterally implement a series of far-reaching “Hartz” labor market reform laws that took such steps as shortening eligibility periods for unemployment benefits, extending possibilities for temporary agency labor, and curtailing the social partners’ institutional power. Importantly, while doing so, the government was very careful not to endanger continuing coordination. In that spirit, it left the employment conditions for employers’ core workforces largely intact (Thelen, 2012).

Ultimately, as demonstrated by the labor market’s impressive performance during and after the recent financial crisis, the continuing strength of coordination in Germany makes the usage of concerted means for policy reform far less necessary than in other countries. When the crisis hit, Germany was greatly affected, with output in the export-dependent manufacturing sector falling by more than a fourth between the third quarter of 2008 and the first quarter of 2009, and the economy contracting by 4.7% in 2009 alone (International Labor Organization [ILO], 2011). However, rather than unemployment rising strongly, increases were temporary, and it soon fell to 6.6% in December 2010, well below the averages of 8.1% in the Organization for Economic Co-operation and Development (OECD) and 9.6% in the European Union (Reisenbichler & Morgan, 2012). In contrast to competitors in other countries, German companies were able to hold on to—rather being forced to shed—their skilled labor during the economic crisis by using the tools they had developed in cooperation with worker representatives to nurture internal flexibility. Throughout 2009, companies prevented layoffs by reducing working hours, cutting overtime, drawing down working-time accounts, and increasing holidays. In combination, these four measures reduced average yearly working time per employee by twice as much as the government’s extensions of short-time work (ILO, 2011). ¹⁶

Long before the crisis, coordination underwrote the restructuring of production processes, including the definition (and training to) new skill profiles but also wage restraint and sourcing more inputs from abroad, which improved the country’s export performance relative to its main trading partners (Carlin & Soskice, 2009). Coordination continues to effectively support businesses’ success with incremental innovation in high-value products such as cars or machine tools. It is thus no wonder that firms are keen to further update the vocational training system to ready future workers for the “industry 4.0” production strategies that managers envision will make use of cyber-physical systems communicating via Internet protocols. ¹⁷

As already hinted, patterns of coordination have recently become more liberal in character, leaving additional room for local decisions and becoming less inclusive. For example, in collective bargaining, a slow shrinkage in the reach of and shifts to lower levels of aggregation have been under way for some time. Between 1996 and 2011, the share of employees working under sectoral collective bargaining coverage has fallen from 70% to 54% in the West and from 56% to 37% in the East (Ellguth & Kohaut, 2012). Moreover, works councils have gained in power vis-à-vis central union offices (Holst, 2013). Although even the government-sponsored financial liberalization during the 2000s failed to significantly undercut employers’ self-interest in continued coordination (Deeg, 2010), tendencies toward dualization are undeniable. They express themselves in the growing contrast between capital–labor cooperation continuing strongly for some workers, particularly at local level and increasingly on the terms set by management, and increasing numbers of workers being cut off from the benefit of such cooperation. ¹⁸ For instance, as manufacturing companies have made ample use of the Schröder government’s labor market reforms, they hired more temporary agency labor as part of their quest to gain organizational flexibility throughout the business cycle while keeping the benefits of cooperation with core workforces on locally adapted but sectorally coordinated standards. In addition, given that the social partners continue to oppose state activism and concertation in vocational education, large companies are increasingly getting their way, with the training system developing in a more segmentalist direction and featuring less standardization of curricula (Thelen & Busemeyer, 2012). In sum, although coordination has become less encompassing than in previous decades, it remains a defining characteristic of the German economy, and recent reforms have arguably had the effect of entrenching coordination within key sectors (Thelen, 2012, 154). This relative stability represents a sharp contrast to the abrupt, across-the-board collapse of concertation in Ireland during the Great Recession.

Of course, the particulars of the interaction between concertation and coordination in Germany are influenced by the specific features of the country’s political economy, including the unique weakness of the semi-sovereign state. One could argue that concertation is uniquely underdeveloped in Germany. But similar patterns prevail in other CMEs, including the nationally coordinated Scandinavian states. ¹⁹ Historically, these countries combined high levels of concertation and coordination, with policy cooperation delivering collective goods that employers craved, including macroeconomic stability (Pekkarinen et al., 1992; Scharpf, 1984). Concertation, however, proved less useful in delivering the kinds of specialized collective goods that employers sought to navigate rapid technological change, economic internationalization and tertiarization. As a result, the Scandinavian countries increasingly resemble Germany in their reliance on coordination rather than concertation.

This shift is most visible in Sweden. As hinted at before, the breakdown in concertation was influenced by government policy. Employer attitudes toward policy cooperation soured in the 1980s, following a leftward shift in public policy and the introduction of the controversial wage earner funds in particular (Vartiainen, 1998). Concertation did not completely collapse, however, until the election of the center-right government that not only failed to support institutionalized cooperation in policy making but actively opposed it (Pestoff, 2002). Although employers returned to the bargaining table in a more conducive political environment, the Alliance for Growth sponsored by the Social Democrats in the late 1990s proved short-lived, and concertation in Sweden has remained relatively weak (Blom-Hansen, 2000).

Swedish coordination, on the other hand, has proven more robust despite the Bildt administration’s hostility to organized capitalism in the early 1990s (Rehn, 1996, 189). To be certain, Sweden has witnessed important changes, including a weakening of the universal banks, a decline in trade union density and the decentralization of collective wage bargaining over the course of the 1980s and 1990s. Wages, however, are still coordinated at the sectoral level, where well-organized industry associations continue to engage labor unions over working hours, shopfloor conditions, and other issues related to production (Lindgren, 2011). Why did coordination persist, whereas concertation collapsed? Businesses were able to draw on the benefits of a high level of coordination among themselves, as well as on the strong, bilateral relationships with organized labor that these inter-business linkages supported at the sectoral and shopfloor level. In turn, concertation was less necessary for business. They could abandon concertation without endangering access to secure specialized inputs by negotiating independently with organized labor. ²⁰

Although this breakdown in concertation is less conspicuous in Denmark, the evolution of cooperation in Sweden’s neighbor is characterized by parallel developments. Like Sweden, Denmark historically relied on concertation, most recently to navigate the economic shocks of the 1980s and 1990s. Policy cooperation, however, declined over time (Blom-Hansen, 2000). This trend accelerated in the 2000s, as a center-right government reduced the social partners’ influence in policy making and implementation (Jørgensen, 2009). Coordination remains robust, however, extending from sectoral-level wage setting to ambitious agreements over working hours and training. In fact, coordination has proven so successful that it undermines concertation. The center-right government reduced cooperation in policy making because it viewed the social partners as so successful in advancing their collective interests that it undermined the state’s fiscal priorities (Schulze-Cleven & Weishaupt, in press).

Ireland: The Politics of Concertation

Unfortunately, the literature on coordination does not explain developments elsewhere in Europe. More specifically, it fails to explain the experience of tripartite bargaining in countries with a historically weak history of cooperation along the European periphery (Rhodes, 2001). Some countries like France and the United Kingdom, with very powerful executives and single-party governments that militated against both coordination and concertation, evolved in a liberal, pluralist direction. Other countries, with relatively pluralist traditions, but weaker executives, responded differently. Irish policy makers, for example, struck a series of three-year collective agreements with industry representatives and trade unions that were subsequently broadened to tackle taxation, social benefits, training, and a range of other issues (Hardiman, 2002; O’Donnell, 2001; Thomas, 2003). The emergence of Irish “social partnership” fits poorly within the VoC framework, which predicts that cooperation is difficult to create, particularly in a historically liberal market economy (Hall & Soskice, 2001). The Irish case is far less puzzling, however, when examined not as a form of coordination (Culpepper, 2002; Hardiman, 2002) but rather as an outbreak of concertation (Baccaro, 2003; Baccaro & Lim, 2007).

In fact, Ireland represents an archetypical example of concertation because Irish social partnership engaged producer associations, including the Irish Confederation of Trade Unions and the Irish Business and Employers Confederation in incomes, fiscal, social, and industrial policies (Thomas, 2003). Although the three-year wage agreements that underpinned social partnership could be seen as a form of coordination, negotiations were heavily dependent on policy concessions, particularly income tax reductions (Hardiman, 2002). ²¹ In fact, social partnership was so dependent on political tradeoffs that societal actors expressed concern about the proliferation of policy commitments and the unsustainable demands this placed on the social partners during the 2000s. ²²

Collaboration in the act of production, however, remained sharply circumscribed. Employers resisted initiatives to recognize, much less to actively cooperate with labor (European Industrial Relations Observatory, 2002), with the result that trade union density actually declined during two decades of social partnership (Regan, 2012). Disputes over trade union recognition and employer reluctance to cooperate with existing unions led analysts to dismiss experiments with enterprise-level cooperation as “more rhetoric than reality” (Gunnigle, 1998, p. 195).

Attempts at inter-firm network-building were no less problematic, as efforts to link indigenous and multinational enterprises were shuttered after less than three years (O’Malley, Kennedy, & O’Donnell, 1992), whereas other networking programs were criticized for their lack of effectiveness (Sterne, 2004). Policy makers and industry representatives, even those embedded within network-based initiatives such as the Centers for Science, Engineering, and Technology, confessed that there was little cooperation in research, training, marketing, or other elements of production. ²³ Whereas some representatives insisted that inter-firm networking was important, they usually emphasized lobbying when pressed to supply examples. ²⁴ For example, one interviewee remarked, “We’re continually networking. [There is] the American Chamber of Commerce, which has a very high political profile. [It] works on the corporate taxation rate.” ²⁵

Precisely because Ireland relied on concertation rather than coordination as in Germany, Irish scholars have generated very different theories about cooperation than their German counterparts. Analysts, for example, often highlight the ease with which stakeholders were able to construct social partnership during the 1980s (Hardiman, 2002; O’Donnell, 2001). Within a historically liberal market economy, Irish stakeholders had attempted and failed to promote labor market cooperation in earlier decades (Hardiman, 1988). ²⁶ But the 1987 Program for National Recovery successfully broke from Ireland’s conflictual and voluntarist tradition of industrial relations by forging a tripartite agreement to reduce government spending and moderate wage growth over a three-year period (O’Donnell, 2001).

In contrast to earlier deals, the Program for National Recovery established a precedent for subsequent three-year agreements. Furthermore, cooperation was not limited to a single issue or domain, but rapidly adapted an austerity pact into a strategy to tackle unemployment, social exclusion, and investment. Subsequent deals broadened cooperation to the extent that stakeholders were bargaining over

tax policy; fiscal policy; exchange rate and monetary policy; social welfare payments and social policy; local economic development; active labour market policies and training; public service modernization; health policy; firm and macro-level industrial relations; race relations; equality; housing; transport; education and agriculture and rural development. (Thomas, 2003, p. 194)

At the same time, Irish social partnership proved more fragile than institutionalized cooperation in other countries such as Germany. Although cooperation was broadened to an impressive array of policy domains (Thomas, 2003) and had converted even its staunchest critics into proponents (Thomas, 2003), collective bargaining abruptly collapsed in the wake of the financial crisis in 2009 (Regan, 2012).

In explaining Irish social partnership, scholars have emphasized the central role of the state, which established an institutional home and focal point for negotiations (O’Donnell, 2001), confronted producer associations such as trade unions with the credible threat of unilateral fiscal retrenchment (Allen, 2000), and underwrote collective wage agreements by providing side payments, most notably tax concessions (Hardiman, 2002). Indeed, some analysts highlight the leadership of individual politicians, most notably former Taoiseach Bertie Ahern, in brokering an agreement among industry and labor representatives (Thomas, 2003). To this end, the content of social partnership proved sensitive to shifts in the partisan composition of government. For example, center-right governments made little effort to target low-income groups in reducing taxes as part of the 1987 Program for National Recovery and the 1990 Programme for Economic and Social Progress (Hardiman, 1998). The Fine Gael-Labour government that followed, however, devoted greater attention to social exclusion and inequality. For example, the 1993 Partnership for Competitiveness and Work targeted low-income groups (Hardiman, 1998) and created new employment services (Taylor, 2002). Moreover, the Partnership 2000 agreement from 1996 expanded these initiatives, formally incorporating the unemployed, women’s groups and charities into “official designations of Social Partnership” (Rush, 1999). By the same token, however, the government’s ability to refashion Irish social partnership rendered it vulnerable. Cooperation proved harder to sustain once the charismatic Bernie Ahern was no longer Taoiseach and policy makers lacked the resources to provide generous side payments (Regan, 2012).

One source of strain on sustaining cooperation was concertation’s inability to deliver the same collective goods as German-style coordination. Although collective agreements moderated fiscal spending, contained wage growth, reduced taxation, and secured industrial peace (Hardiman, 2002; Haughton, 1998; Thomas, 2003), investments in high-quality inputs remained at comparatively low levels. For example, although the social partners agreed to tackle unemployment and social exclusion through active labor market measures (Sexton et al., 1997), training was not integrated into companies’ production strategies as in CMEs like Denmark (Martin, 2005). Active labor market measures largely revolved around public employment and wage subsidies (Sexton et al., 1997), with employers viewing publicly provided training as neither particularly relevant nor useful (Boyle, 2005; Fitzgerald, Kearney, Morgenroth, & Smyth, 1999). In fact, the Training and Employment Authority, which was always viewed critically by Irish industry, was dissolved following a financial scandal in 2008. ²⁷

Although Ireland increased investments in other collective goods such as research and higher education, its commitments continued to trail its West European peers (Breznitz, 2007; OECD, 2006). Without inter-firm networks that would enable small- and medium-sized (indigenous) enterprises to capitalize on public investment in basic research (Breznitz, 2007), the country’s expanding high-technology sector remained characterized by relatively low profitability and proved vulnerable to increasing cost competition (Breznitz, 2007). After 2000, economic growth was driven by non-tradable services, most notably residential investment (Honohan, 2009). ²⁸ The resulting boom and bust further reduced the fiscal resources available to policy makers to underwrite continuing policy concertation (Regan, 2012).

The dynamics of concertation in the Irish case appear to apply more broadly to outbreaks of cooperation during the 1980s and 1990s in such countries as the Netherlands, Italy, Portugal, Spain, and Eastern Europe. As in Ireland, policy deals in these countries indicate that collaboration is much easier to create than the VoC literature suggests, in part because of the contrasting role of the state in concertation and coordination (Avdagic, 2010; Ebbinghaus & Hassel, 2000; Iankova, 2002; Rhodes, 2001). At the same time, the national agreements in these cases were based largely on policy cooperation, either within prominent social pacts or more modest, “negotiated” social policy reforms. Cooperation in productive activities such as training, technological development, standard-setting, or marketing remained highly path-dependent. Virtually all of the countries that had to construct cooperation from scratch turned to wage restraint, fiscal retrenchment, or market-oriented social benefit reform to promote competitiveness (Hassel, 2007; Ost, 2000; Rhodes, 2001).

Conclusion: Conceptualizing Cooperation and Collective Capabilities

The German and Irish cases suggest that concertation and coordination represent distinct forms of cooperation, associated with unique causal logics. Although this article used a cross-national comparison to highlight these differences, within-case analysis also supports these claims, including both the differential role played by public authorities in the two processes of collective action and the processes’ contrasting relative stability. Analysts exploring German social policy reform are more likely to highlight the transformative role of the state (Vail, 2009) than those studying issues more closely linked to production such as vocational training (Thelen & Busemeyer, 2012) or finance (Deeg, 2010). By the same token, scholars studying Irish incomes policies are more likely to stress state capacity and institutional innovation (Hardiman, 2002) than those focusing on the continuing low degrees of shopfloor collaboration (Gunnigle, 1998). Moreover, parallel dynamics prevail in other countries, such as Sweden, where employer participation in policy concertation proved surprisingly vulnerable to partisan shifts (Anthonsen & Lindvall, 2009), whereas coordination in the act of production proved more resilient (Lindgren, 2011).

A combination of acknowledging the diversity of collective action, appropriately matching concepts to empirical material, and exercising caution in generalizing from one form of cooperation to another promises to resolve several of the continuing scholarly disagreements about the dynamics of cooperation. Greater conceptual clarity would also reduce confusion among policy makers who consume the literature on comparative political economy. For instance, Irish policy makers, emboldened by the rapid institutionalization of Irish social partnership and encouraged by the literature on coordination in Germany and Nordic Europe, believed that they could use concertation to foster quality-based competition (National Economic and Social Council, 2005). Finally, the preceding analysis encourages anyone interested in increasing communities’ ability to organize collective responses to syncretic processes of neoliberal transformation to draw more nuanced lessons about the causes and consequences of different forms of social partnership.

Of course, there are limits to the simple distinction between concertation and coordination elaborated here. For instance, our analysis glosses over potential differences between coordination at different levels. Within the VoC literature, scholars have theorized distinct roles for the state (Martin & Thelen, 2007) and distributional consequences for company-, sectoral-, and national-level forms of coordination (Martin & Swank, 2012). We support such distinctions and do not advocate a single framework to analyze capitalist diversity. Rather, we question the tendency to analyze all forms of cooperation through the lens of coordination and advocate leveraging the insights of the underappreciated literature on concertation. In line with other recent conceptual work that introduces new concepts to assess the evolution of coordination, including its drivers and consequences (Höpner, 2007; Thelen, 2012), we believe that a multi-dimensional approach is the best way forward to capture the evolving dynamics of cooperation. Among other promising leads, we believe that such an approach to the study of cooperation could yield new insights into the changing institutional foundations of comparative advantage (Hall & Soskice, 2001), help probe the distributional effects of coordination’s evolving scope, and rationalize the diverging cross-national fate of concertation in the face of both organized interests’ membership erosion and recession-induced austerity (Alfonso, 2013).

In acknowledging a broader, more diverse “varieties of cooperation,” it is soon evident that economies can combine different forms of collaboration. For example, although we document a general retreat from policy cooperation in Scandinavia, economic shocks have triggered episodes of concertation in recent years. Even here, however, we argue that clearer analytical distinctions between concertation and coordination allow scholars to better understand how their interactions have driven the evolution of cooperation. For example, the surprising resilience of concertation in Finland during the 1990s reflects the failure of traditional coordinating institutions to grapple with depression-like conditions. Closer attention to the literature on concertation reveals why and how the Finnish government was able to radically restructure coordinating institutions, threatening producer associations in ways that were not possible in Germany and other highly coordinated economies. ²⁹ At the same time, scholarship on coordination helps explain why Finnish concertation diverged from Irish social partnership. Whereas Irish policy makers prioritized wage restraint and market-oriented reform, Finnish policy makers could rely on a legacy of private–public and inter-firm cooperation to boost investment in a variety of high-quality collective goods, including education, early stage risk capital, and research (Ornston, 2012b). ³⁰ Of course, the Finnish example represents but one of many possible interactions between the power-based dynamics of concertation and processes of interest-based coordination, either one of which can center on different levels of analysis. We hope that future scholarship will probe this interplay to uncover the potential synergies and blockages that might be associated with specific combinations of cooperation in different institutional contexts.