Orchestrators of coordination: Towards a new role of the state in coordinated capitalism?

Abstract

Liberalization poses significant challenges for the continued provision of collective goods within coordinated market economies (CME). Extant scholarship suggests two dominant sets of responses. Either CMEs continue to rely on employer coordination, but only for a privileged core, leading to dualization. Or, in cases where the state enjoys high capacity, the state instead compensates for liberalization but ends up crowding out employer coordination. In both cases, the result is decreasing employer coordination. We argue that in CMEs, the state may also play the role of “orchestrator” by supporting the revitalization of employer coordination. It does so through the deployment of ideational and institutional resources that mobilize employers’ associations on a voluntary basis. Applying our framework to a core area of coordinated capitalism, vocational education and training, we show that in both Germany and Switzerland, this indirect and soft form of state intervention was instrumental for turning around their crisis-stricken vocational training systems.

Introduction

In recent decades, structural pressures linked to the rise of the knowledge economy have challenged the capacity of coordinated market economies (CME) to provide an alternative to liberal models. Although scholars have identified different responses to these challenges, they agree that the degree of employer coordination, that is employers’ capacity to act collectively to achieve joint gains in relation to their production (Ornston and Schulze-Cleven, 2015), is declining. Instead, they have observed the upholding of employer coordination for a privileged core only, underwritten by reactive state policy and with negative consequences for collective good provision (usually dubbed “dualization”). Alternatively, scholars have pointed to the massive expansion of direct state provision of collective goods while pursuing liberalization, which, however, crowds out employer coordination (Baccaro and Howell, 2017; Emmenegger et al., 2012; Thelen, 2014).

In contrast, we argue that states play a more important role than previously recognized (most notably in the Varieties of Capitalism framework, Hall and Soskice, 2001) in actively supporting employer coordination. In this way, states partially moderate dualization without recurring to hierarchical forms of state intervention. Triggered by increasing political pressure borne out of self-undermining feedback effects from dualization (Jacobs and Weaver, 2015), this new form of state agency aims to reinvigorate institutions for employer coordination that play a key role in collective good provision. Rather than exhibiting the extremes of a largely passive state or one that takes over at the expense of employer coordination, we identify a different approach in our case studies. Within this approach, which we call “orchestration,” the state deploys ideational and institutional resources to mobilize intermediary actors on a voluntary basis in pursuit of joint governance goals. The term is borrowed from recent literature in international relations (Abbott et al., 2015) and describes the behavior of organizations that rely on indirect and soft forms of governance. We argue that this perspective, which also resonates with work on “metagovernance” of policy networks in public administration (Sørensen and Torfing, 2009) and soft governance modes in European integration studies (Barcevicius et al., 2014), may be fruitfully adapted to the literature on coordinated capitalism.

Orchestration denotes a less intrusive form of state intervention than traditional forms of state activity (e.g., regulation or fiscal policy) or macro-policy concertation with peak-level representatives of business and labor focused on public policymaking, which presuppose the capacity to coerce and/or the availability of sufficient material resources (e.g., Molina and Rhodes, 2002; Ornston and Schulze-Cleven, 2015; Schmidt, 2009). It is also different from more recent “social pacts” that have in many cases primarily served to mobilize societal support for austerity reforms (Regan, 2017) rather than to revitalize employer coordination for the pursuit of collective goals. Moreover, even though scholarship on classic neo-corporatism has certainly pointed to the role of state actors (e.g., Berger, 1981; Katzenstein, 1985; Streeck and Schmitter, 1985), most work in this tradition has evaded a deeper reflection of the state as an active “orchestrator” (Hemerijck and Vail, 2006; Ornston, 2012; Schmidt, 2009).

For some time, this lack of attention seemed justified by the state’s passive role casting a “shadow of hierarchy” (Scharpf, 1997) while granting non-state actors autonomy in managing their own affairs (Streeck and Schmitter, 1985). Yet traditional forms of neo-corporatist interest mediation have come under pressure due to union decline and the empowering of employers in a globalized economy, which has shifted the balance of power between business, labor, and the state. The challenge for policymakers in CMEs is how to maintain and further employer coordination in this changing socio-economic environment. With the state’s ability to coerce employers to cooperate at an all-time low and limited material resources to engage in generalized political exchanges as in classic neo-corporatism, the state must adopt a more pro-active role in facilitating employer coordination. For us, the notion of “orchestration” is particularly apt to describe these new forms of state intervention in the context of traditionally corporatist countries, which (necessarily) rely on the voluntarist contribution of “intermediaries” (Abbott et al., 2015) in exchange for public legitimacy. Given its dwindling supply of “hard” currency in corporatist practices of generalized political exchange (both in terms of coercive capacity and material resources), state actors take on a more active, moderating, and skillful role as orchestrators compared to the rather passive, reactive, but ultimately more resourceful stance of state actors in classic neo-corporatism.

We develop this argument in the context of collective skill formation systems that rely on firms and their intermediary associations to participate in the financing and administration of training, which presupposes employer coordination (Busemeyer and Trampusch, 2012). Given the centrality of vocational education and training (VET) as the “showpiece of coordinated market economies” (Culpepper, 2001: 275), analyses of this policy area speak more broadly to debates about dynamics of institutional change in political economies. In the empirical section, we study how established neo-corporatist practices are transformed, evolving into a more “orchestrated” approach of the state in skill formation policies. We do so through an examination of changes in two CMEs with long corporatist traditions of sharing governing space with private actors, Germany, and Switzerland. Our study of recent policy developments in these two countries shows how strategies of state-led orchestration are quite successful in maintaining employer commitment to societal goals such as the training of youths, but we also explore the extent to which these new strategies indicate significant deviations from the past.

In the conclusion, we discuss the broader relevance of the concept of orchestration for analyzing institutional change in advanced political economies. We argue that state orchestration of employer coordination may be feasible in other policy fields besides skill formation, even though its concrete manifestations are likely to vary.

The role of the state in CMEs

Even though scholarship in comparative political economy usually focuses more on firms, producer groups, and their representatives, there is a growing interest in the field in better understanding the role of the state. However, so far, state actors are generally perceived as, for instance, active promoters of liberalization policies (Howell, 2019) or as implicit supporters of dualization (Palier and Thelen, 2010). Only recently, there is evidence that at least in some cases, states in CMEs are increasingly pushing in the opposite direction, for example, actively promoting more employer coordination rather than dismantling it (Graf et al., 2021). To account for this shift, scholarship has pointed to self-undermining feedback effects from liberalization (Jacobs and Weaver, 2015), and how even in economies with little tradition for state involvement in industrial relations or training, the state is called upon to bring greater balance to the economy. This new role of the state in CMEs, however, lacks further theoretical scaffolding that may provide tools for analyzing the state’s more pro-active stance.

A key reason for this lacuna may be the way the role of the state was featured in classic theories of neo-corporatism. Broadly speaking, the literature on neo-corporatism traditionally paid limited attention to state actors in actively fostering employer coordination, even though this is the core characteristic distinguishing coordinated from liberal market economies (Hall and Soskice, 2001). This body of work generally depicts the role of the state as mostly passive and reactive, allowing non-state actors considerable autonomy in managing their own affairs, while still maintaining sufficient state capacity to cast a shadow of hierarchy (Scharpf, 1997) in case non-state actors do not uphold their commitment to maintain a certain degree of collective good provision (Streeck and Schmitter, 1985). Thus, classic neo-corporatism depended on the existence of state capacities to intervene if necessary, a relatively equalized balance of power between business and labor, and the capacity of unions and employers’ associations to enforce collective decisions among their members (Streeck, 1992). Taken together, neo-corporatist institutions constituted a regime of “beneficial constraints” that furthered employer coordination by limiting the ability of individual employers to pursue short-term profit maximization strategies, while at the same time forcing them to contribute to (but also benefit from) joint investments in collective goods such as a well-trained labor force.

However, globalization and the transition towards service-based knowledge economies pose significant challenges for established systems of corporatist interest intermediation, necessitating a re-thinking of theoretical perspectives of corporatism. The decline of union power, the decentralization of collective negotiations, and the weakening capacity of intermediary associations to enforce joint decisions among their members are key factors contributing to the diminished potential for employer coordination (Baccaro and Howell, 2017).

In a situation of weakening corporatist institutions, what are the possibilities for the state to counter negative effects of structural pressures on employer coordination? The influential work of Thelen (2014) carves out two policy responses for CMEs to these challenges, namely dualization and embedded flexibilization. Dualization restricts the scope of collective good provision to a core of large companies that still benefit from employer coordination, whereas more peripheral companies are increasingly left out. Dualization conceives of the role of the state as a largely reactive and passive “underwriter” of the growing dualism in labor markets by insulating the privileged core by means of public policy (Palier and Thelen, 2010). In skill formation more specifically, the negative side effects of such dualization were the rationing of access to training, the creation of second-class training, and declining firm participation (Thelen, 2014: 10-11, 86).

In contrast, embedded flexibilization implies a more active role for the state, but this role comes in the form of direct state intervention that compensates for the negative side effects of liberalization through public policies and macro-corporatist concertation. This form of concertation, which denotes tripartite cooperation with peak-level representatives of business and labor in public policymaking (Ornston and Schulze-Cleven, 2015: 556), however, comes at the expense of employer coordination in the production sphere, so that direct state provision increasingly substitutes for collective goods previously provided through employer coordination (Martin and Thelen, 2007). In the realm of training policies, this is achieved, for instance, by establishing state-sponsored training schemes and vocational schools. In the short term, these measures ensure a sufficient supply of training places for youths. In the long run, however, they can prove counterproductive as employers are reluctant to hire youths trained in systems that only to a limited extent rely on their input (Busemeyer, 2015). From this perspective, direct state provision endangers employers’ willingness to comply because companies are reluctant to give up proprietary and strategic information to the state (Culpepper, 2003; Ornston, 2012).

This dominant view in the literature thus suggests a dilemma for state actors: Giving priority to maintaining employer coordination leads to dualization with the state becoming a passive underwriter of these developments. Vice versa, adopting a more interventionist stance leads to a crowding out of employer coordination, surrendering a key comparative institutional advantage of CME. Our notion of state-led orchestration suggests that there may be a way out of this dilemma, namely for state actors to adopt a more activist, but still largely non-interventionist approach that is focused on furthering employer coordination rather than replacing it with direct state provision.

Our argument builds on a small, but growing literature that attributes a central role to the state in funding, certifying, and governing collectivist VET systems (Busemeyer and Trampusch, 2012; Carstensen and Ibsen, 2021; Graf et al., 2021). In the following sections, we develop our more encompassing theoretical framework on state-led orchestration that can be used to understand the changing role of the state in collective skill formation systems.

The state as orchestrator

This paper suggests that a more pro-active role of the state is starting to emerge in economies that have previously been characterized by dualization. The main aim of this renewed role, which we dub orchestration, is a state-led revitalization of employer coordination.

Orchestration differs in important ways from the two primary responses to the rise of the knowledge economy identified by Thelen (2014). Different from dualization, orchestration pays more attention to labor market outcomes of those beyond the core of political economies. This goes along with a more active role of the state in facilitating deliberations between a larger set of actors compared to classic neo-corporatist decision-making bodies—while still being focused on maintaining employer coordination. Different from embedded flexibilization and concertation, orchestration is a form of state agency that relies on soft and indirect but no less consequential forms of governance. Brought on by state actors lacking the capacity to enforce hierarchical means of intervention (or because state actors oppose such means for ideological reasons), states have adopted a different, less interventionist approach to governance. Rather than turning to state provision, the state deploys intermediaries (employers’ associations, unions, and other non-state actors) that get institutional and ideational resources to further the private provision of collective goods through employer coordination. ¹

State orchestration as a form of state agency

We borrow the concept of orchestration from recent work in international relations and governance theory. This body of work theorizes under which conditions actors with limited access to hard forms of governance can project power by relying on soft and indirect forms of governance (Barcevicius et al., 2014; Sørensen and Torfing, 2009). Along these lines, (Abbott et al. 2015: 722) define orchestration as the “mobilization of an intermediary by an orchestrator on a voluntary basis in pursuit of a joint governance goal.” Rather than conceiving of non-state intermediaries as being in the subordinate role of a hierarchical relationship with the state, orchestration implies a more cooperative approach between state and non-state actors. State actors make use of intermediaries to get access to expertise and specialized knowledge about pressing issues and to get legitimacy for policy choices and their effective implementation.

We argue that the concept of orchestration may be fruitfully adapted to the literature on neo-corporatism and coordinated capitalism. Taking place within a nation-state, neo-corporatist structures clearly offer a different setting for cooperation than the one most commonly in place at the international level, but still, the ability of the state to intervene hierarchically is limited, especially in corporatist countries, where the state depends on employers’ continued contribution to the provision of collective goods. One example of this is the area of collective skill formation. Here the state could, in principle, use its coercive power to set up a statist VET system but doing so would leave the state without crucial resources from employers. To maintain an efficient and relevant VET system, the state is thus dependent on employers to provide training positions and define the skill requirements of the economy (Bonoli and Emmenegger, 2021; Busemeyer, 2012). Without the ability to impose constraints on employers, and with unions no longer able to act as a counterbalance to employers, states have become ever more dependent on soft forms of governance to keep employers involved. Essentially, the state needs to adopt a leading and more active role in remodeling corporatist decision-making, including reaching out to societal stakeholders that have traditionally not been part of corporatist decision-making.

Orchestration implies that state actors make a conscious effort to support and promote employer coordination by balancing out competing interests. As noted above, the state has partially done this already in classic neo-corporatism, but these past arrangements crucially depended on prerequisites, which can no longer be taken for granted. Thus, due to the changed socio-economic environment, the involvement of the state is necessarily less interventionist and more focused on setting positive incentives for intermediary associations to engage with the state and each other, voluntarily accepting the state’s leadership in setting and defining joint governance goals in collective good provision (here firm-based apprenticeship training). In the words of Abbott et al. (2015: 722), “by assuming ideational leadership over the intermediary’s problem definitions and policy priorities, the orchestrator can nudge the intermediary toward governance goals that are compatible with its own goals.” Admittedly, this is a weaker form of state intervention compared to embedded flexibilization, concertation, and classic neo-corporatism. The importance of the state is based less on its capacity to impose and enforce rules but rather in taking over responsibility to coordinate, that is, orchestrate, the various activities of different stakeholders and thereby facilitate employer coordination.

The commitment of intermediaries, which is vital for the effectiveness of orchestration as a governance strategy, hinges on intermediaries perceiving the state as a neutral and honest broker (Seitzl and Emmenegger, 2019). That is, although political parties of course shape state priorities, they must be committed to the promotion of collective goods such as the training of youths. Maintaining the state’s political neutrality is supported by giving bureaucratic actors (state agencies) a dominant role in orchestration efforts. Agencies care about the effectiveness of policies in “their” domain—either out of professional ethos or because their agency is likely to foot the bill in terms of public expenditures or take the political blame in case of policy failure (Martin and Thelen, 2007). This also means that state orchestration of employer coordination is more likely to be successful if there are established networks of institutionalized trust between the state and intermediaries, that is, employers’ associations, unions, and other stakeholders (Crouch, 1993; Lehmbruch, 1984). Thus, state orchestration is more likely to succeed in countries with corporatist traditions, but it is certainly feasible in other contexts as well because it is less demanding in terms of societal actors’ collective action capacities compared to classic neo-corporatism.

What do orchestrators do?

Successful orchestration of employer coordination depends on the deployment of resources that enable intermediaries to shape the actions of the agents they target. What kind of resources, then, are useful for the state in reaching its goals? As already noted, the resources that states provide to intermediaries through orchestration are primarily ideational and institutional.

Ideational support provides public legitimacy to intermediaries by placing them centrally in the governance of the system (Offe, 1981). Ideational support also facilitates the sharing of information and data (Culpepper, 2003) as well as the deliberative construction of narratives that otherwise dispersed actors can rally behind (Schmidt, 2009; Sørensen and Torfing, 2009) in a joint effort to produce creative responses (Ornston, 2012) to the ever-evolving challenges to skill formation systems. Such narratives include the portrayal of collective skill formation systems as a success story—not least due to the efforts of firms, which increases the systems’ overall attractiveness and visibility. Different from classic neo-corporatism, state actors as orchestrators need to be more actively engaged and involved in the construction of these overarching narratives to generate and maintain public legitimacy (e.g., by referring to national skills strategies). Moreover, through repeated interactions, state orchestration holds the potential to transform narrow preferences of employers by signaling continued support for coordinated efforts (Martin, 2000).

The provision of institutional resources involves the alignment of the political institutional setting that shapes the reform process, the effective bringing together of diverging interests within a common institutional frame of negotiation, the sharing of “public space” with organized interests (Crouch, 1993), and may even include the creation of new intermediary associations (Graf et al., 2021). Hence, different from ideational support discussed above, the support of institutional resources involves the delegation of actual decision-making power to non-state actors. Of course, this was also a central feature of classic neo-corporatism. But orchestration is different in two ways. First, state-led orchestration involves a more active role of the state in balancing out the increasing power asymmetries between unions and business associations as well as between small and large firms. State orchestration therefore entails the maintaining of consensual and inclusive decision-making, while, at the same time, respecting that non-state actors enjoy considerable autonomy in managing their own affairs. Of course, given that orchestration is a soft form of state intervention, there are limits to what the state can do in terms of balancing out power asymmetries. One example is to grant institutionally weak intermediaries a voice of equal weight in deliberations to those of economically powerful actors. Another example is active state support in the creation of intermediary associations that further employer coordination.

Second, another important difference between orchestration and classic neo-corporatism is the broader range of actors that are included in deliberative processes. As will be shown below, this broader range of actors includes other societal stakeholders as well as representatives from different parts of the state architecture beyond the traditional corporatist actors (unions and employers). The extent to which this is done depends on policy legacies and political circumstances, but we clearly recognize a willingness of state actors to promote more inclusive deliberation forums to reap the benefits of cross-class cooperation across the whole economy rather than in a shrinking core only to be able to credibly construct collective narratives.

Of course, orchestrating states can also offer material resources, although in stark contrast to concertation, the lack of material resources is often the very reason why orchestrating states refrain from engaging in corporatist practices of generalized political exchange. The more limited material support characteristic of orchestration may include assistance for intermediaries’ management activities, the provision of apprenticeship positions in the public sector (as a regular employer, not as an alternative to employer-provided training), and public subsidies for new experimental projects (Culpepper, 2003; Strebel et al., 2021). However, compared to the ideational and institutional resources, these material resources are less important.

Empirical evidence

In this section, we demonstrate how orchestration works in practice. To do so, we examine changes in the field of VET policy in Germany and Switzerland. The starting point of our case studies are the apprenticeship crises of the 1990s and early 2000s, when policymakers in both countries struggled with a significant lack of training places. The first reaction to the crisis on the apprenticeship market was “dualization” (Thelen, 2014) – often called “segmentalism” in the more specific literature on skill formation (Emmenegger and Seitzl, 2019; Thelen and Busemeyer, 2012). At first, scholars studying “dualization” expected it to be a rather stable political constellation, because actors deriving benefits from a preferential treatment are also politically more influential. However, over the years, the negative side effects of segmentalism—notably, an increasing stratification of training and employment opportunities with dualizing consequences for labor market outcomes—have become more apparent. Consequently, and due to public pressure, policymakers have started to change track by attempting to develop more inclusive governance models, that is, what we call orchestration. It is still too early to tell whether orchestration will be sufficiently effective in completely mitigating the dualizing effects of segmentalism, but it certainly represents a change in the state’s approach to employer coordination.

Methodologically, our case studies provide a parallel demonstration of theory (Skocpol and Somers, 1980). Rather than explaining differences in outcomes as is commonly done in a most similar case study design, our aim is to spell out the theory’s implications in specific settings while also demonstrating the commonalities between these settings. This provides readers with a much fuller understanding of the operationalization of key concepts and variables and shows how the theory works on the ground. As mentioned, we focus on VET institutions, which are defining features of CMEs (Culpepper, 2001). We concentrate on the dual apprenticeship training system, which remains at the heart of the non-tertiary German and Swiss skill formation systems.

Germany ²

The German apprenticeship training system faced a deep crisis in the late 1990s and early 2000s, when the number of available apprenticeships slots was by far insufficient to meet the demand on the part of youths and school leavers for training opportunities due to the economic aftershocks of, first, German reunification and, second, the dotcom bubble. The initial reaction to this crisis was one of dualization, which showed itself as the maintenance of “strong coordination on the employer side but in the context of a distinct narrowing of firms and workers covered under the resulting arrangement” in combination with an “unorganized and unregulated periphery [that] is allowed to grow” (Thelen, 2014: 14). In case of VET, dualization is usefully described as a process of increasing segmentalism, which captures the state’s increasing willingness to meet the demands of large firms to ensure the training participation of the remaining firms (Thelen and Busemeyer, 2012). This focus on large firms was partly the result of the state’s passive reaction to the secular decline in the training participation of small and medium-sized firms (Busemeyer, 2009) but also reflected the weak organizational capacities of small and medium-sized in Eastern Germany (Culpepper, 2003: 102–107). ³

In the context of debates about how to deal with the crisis on the training market in the early 2000s, the left wing of the SPD and some left-leaning trade unions pushed for the introduction of a training levy (i.e., collecting a levy from firms that under-invest in training). The “modernist” wing of the SPD eventually shied away from this form of direct state intervention (which was also heavily opposed by the opposition and employers) but passed the first “Training Pact” in 2004 as an alternative. This “pact,” which was more of a joint training initiative than a social pact in the social scientific usage of the term, was signed by representatives of the federal government and employers’ associations. Notably, trade unions did not sign the agreement out of opposition to the liberalizing agenda of the government.

In essence, this joint initiative can be regarded as another element in the dualization approach, as the state adopted a rather passive stance towards the changing training market. The first iteration of the joint initiative contained only soft commitments on the part of employers and the federal government to create additional training places and to establish new policy measures geared to smoothen the transition from school to dual training for youths (Ausbildungspakt, 2004), documenting the rather informal and non-committal nature of the initiative. Most importantly, the traditional corporatist decision-making governance structure in the training reform process remained in place. Hence, unions may have been less concerned about their non-participation, which, for them, was also a way to express their political opposition to the government’s whole approach. However, in the long term, this first joint initiative set in motion a process of transformation in corporatist decision-making processes in the domain of skill formation, which has also led to a re-definition of the state’s role.

The first joint initiative was followed by several further joint initiatives over the years: the second one lasted from 2007 to 2010 and the third one from 2010 to 2014. Although these two initiatives continued to some extent along the lines set out by the first initiative in 2004, they also indicate an increasingly pro-active approach by state actors to further employer coordination. A first important change was the expansion of signatories, signaling a more inclusive approach to skill formation policies that goes beyond the traditional set of actors. As mentioned above, the first joint initiative was only signed by representatives of the large employers’ associations and two federal ministers. The second iteration added another federal ministry, the Federal Employment Agency, and the Association of Free Vocations. The third joint initiative included representatives from the education ministries at the level of states (Länder) and the federal state secretary for migration affairs.

The content of the initiatives changed too. Rather than dealing with short-term crises on the apprenticeship market only, later iterations adopted a broader perspective on training policy, being concerned, for instance, with mitigating “matching” problems on the training market (i.e., persistent inequalities in the relationship between demand and supply of training places across different occupations and regions) and the integration of youths with migration background. The changing nature of the initiatives through their various iterations therefore signals how the former dualization approach was incrementally transformed into a more inclusive state-led orchestration approach. This new approach primarily focuses on furthering employer coordination and cooperation between the involved stakeholders more generally, but not policy formulation as in classic forms of concertation.

The move towards orchestration was completed in 2014, when the joint initiative was re-modeled into an “Alliance for Initial and Further Education” (Allianz, 2014). Even though the Alliance was and still is a temporary institution (the 2014 agreement would run until 2018, to be followed by yet another agreement in 2019), its internal governance structure is more geared than the previous initiatives towards establishing a permanent discussion forum among a broad set of stakeholders. The federal government provides the resources (i.e., financial support for the management office) to facilitate these efforts, whereas the previous initiatives were largely informal affairs, the Alliance has a political steering committee as well as several working groups. The former includes the political heads of the various participating organizations, whereas members of the working groups usually come from lower levels in the bureaucratic hierarchies to work on concrete issues and problems. In the following years, the federal Alliance became the role model for the establishment of similar initiatives at the local and regional level, so that by now similar initiatives exist in all 16 federal states. ⁴ This signals that the Alliance approach is not an ephemeral political initiative, but a genuine new model for the interaction between the state and other stakeholders in the domain of skill formation.

The Alliance also brought the trade unions back into the boat as signatories (interview with DGB representative, June 2019). The impetus for this change came both from the SPD who had just entered another Grand Coalition with the CDU as well as the unions themselves, realizing that the Alliance might not be merely a short-term crisis measure but could evolve into a lasting feature of the governance structure in the skill formation domain. In addition, the Alliance brought in further partners, again documenting a more inclusive and “orchestrated” approach compared to classic corporatism. These new partners are the ministries for economic affairs and the ministries for labor and social affairs from the Länder level (interview with KMK representative, June 2019). In terms of content, the Alliance covers a broad set of topics, including labor market measures and further training. The inclusion of unions in the Alliance has also contributed to firmer commitments on the part of employers to create additional apprenticeship places (interview with BDA representative, June 2019).

To what extent does the model of corporatism enshrined in the Alliance represent a new approach that might be called “orchestration?” First, different from the previous dualization approach, the Alliance is decidedly inclusive by bringing together all important stakeholders in the domain of skill formation. Furthermore, by involving a broader range of stakeholders, the Alliance provided added value compared to the traditional corporatist decision-making bodies in VET, which still exist alongside the Alliance (interview with ZDH representative, June 2019). Thus, the Alliance signals an important change in direction away from the segmentalist approach of the previous phase. At the same time, the Alliance is not an instrument of direct state intervention. As in traditional neo-corporatism, the consensus principle rules. In addition, since the Alliance runs parallel to existing decision-making bodies, it has no formal powers in the domain of policy formulation (interview with KMK, BMBF representatives, June 2019). In this regard, it differs from traditional models of concertation, which are directed at the policymaking level. In contrast, the Alliance is clearly focused on directly promoting employer coordination as well as coordination among the involved stakeholders more broadly.

Second, different from the traditional neo-corporatist model, the Alliance is decidedly public, highlighting its important function to boost public support for skill formation policies. The internal governance structure of the Alliance mirrors its goals: The working groups focus on concrete issues and challenges in the implementation of training policies, whereas the political steering committee increases the public visibility of collective skill formation (interview with DGB representative, June 2019). For both employers and unions, to be involved in the Alliance signals to the public a certain commitment to the provision of collective goods, that is, providing training opportunities for young people. Hence, the “currency” in terms of ideational resources that state actors offer to employers to stay committed to the corporatist model is public legitimacy.

Third, the Alliance indicates a re-definition of the role of the state in furthering employer coordination. In the traditional, neo-corporatist model, the state would adopt a rather passive role while leaving social partners in the driving seat. In contrast, in the “orchestrated” model of employer coordination embodied in the Alliance, state actors adopt a more active role in bringing together different stakeholders and in setting the agenda for common discussions. Different from the dualization phase, when state actors tended to side with large firms in the export-oriented sectors of the economy (Busemeyer, 2012), state orchestration pursues a more inclusive approach by involving a large range of stakeholders. Furthermore, the model of orchestration also implies that state actors need to walk a fine line by balancing out power asymmetries between unions and employers as well as between small and large firms, while keeping participation in corporatist processes attractive for employers.

Thus, the Alliance could signal a potentially transformative change in state involvement in the domain of skill formation in Germany, moving from dualization to orchestration. As mentioned above, this is also evidenced by the fact that the Alliance model is proliferating from the federal to the regional and local level. However, the Alliance exists alongside the established institutions and procedures in skill formation, which are not abolished. Therefore, its main added value is to create inclusive discussion forums, mobilize ideational resources, promote employer coordination, and create public legitimacy and visibility. Thus, the Alliance does not fundamentally change the fact that in firm-based training systems, states ultimately depend on the continued commitment of employers to provide training opportunities (Busemeyer 2012). As this dependence is likely to increase over time, not least during labor market downturns, it remains an open question whether state orchestration of employer coordination will be a sufficiently powerful remedy against dualizing pressures emerging from the basic structure of the system.

Switzerland

Like Germany, Switzerland faced an apprenticeship crisis in the 1990s due to declining levels of employer coordination. The number of apprenticeship contracts decreased by 17% in the period from 1985 to 1995, while circa 20% of school leavers had no training position or were in transition schemes (Strahm, 2008: 316). The prolonged economic crisis of the 1990s explains part of these struggles. However, this dualization process also marked the high point of decades of state passivity, in which the state had left the task of VET governance entirely to other stakeholders, in particular employers’ associations (Berner, 2013; Trampusch, 2010). Yet by the mid-1990s it was obvious that these intermediary associations were failing to ensure a sufficient degree of collective good provision.

The apprenticeship crisis, and the political pressure that followed it, finally pushed the state into action. At first, the state remained passive, offering a meager 10 million Swiss Francs to finance additional “developmental measures” (Strahm, 2008: 319). However, political pressure kept mounting. In 1997, trade unions started working on a popular initiative (Lehrstellen-Initiative), which demanded a constitutional right to a training position and wanted to require non-training firms to financially support training firms (Sigerist, 2008: 295). In parallel, several cross-partisan initiatives were launched in parliament. The media began to accuse the government of “not doing its homework” (Strahm, 2008: 317).

In response to mounting pressure, the state adopted an orchestration approach to further employer coordination. Providing ideational, institutional, and, to a limited extent, material resources, the state tried to improve the capacity of employers to provide collective goods such as a well-trained labor force. As we show below, the orchestrating state mostly refrained from relying on hard and direct forms of governance in this process. Instead, state intervention aimed to strengthen employer coordination by using soft forms of governance and trying to influence firms’ training behavior through intermediary associations.

Importantly, this more orchestrating state did not meet much resistance from employers, which reflects both declining levels of employer coordination and the VET system’s legitimacy crisis. In particular large firms had become critical of the VET system, demanding a stronger focus on general skills and more emphasis on post-secondary training (Emmenegger and Seitzl, 2019). Consequently, large firms were pushing for transformative change, as reflected, for instance, in the creation of universities of applied sciences (Trampusch, 2010). In contrast, they felt less compelled to preserve their traditional prerogatives.

In a first step to further employer coordination, the government brought two resolutions to parliament (Lehrstellenbeschluss I in 1997 and II in 1999), which were supposed to make it more attractive to firms to engage in training and to facilitate the establishment of sector-specific training programs, which were to be mainly paid by employers but were partly subsidized by the state. In addition, the resolutions aimed to facilitate the VET system’s extension to new economic sectors, most notably information and communications technology (ICT), where state agencies also assisted in the creation of new intermediary associations to govern the new occupational field. With these measures, the orchestrating state tried to make VET attractive for more talented youths (Gonon and Maurer, 2012: 141).

In parallel, the government announced the fundamental revision of the VET act to revitalize the system. In 1999, it set up a highly inclusive reform commission, consisting of all relevant stakeholders, which presented the draft law in 2000. Although the draft law was subject to several changes, it already passed in 2002—with unanimity in both parliamentary chambers (Berner, 2013). The reform “was accompanied by a broad consensus, not only between the social partners, but also between political parties” (Trampusch, 2010: 199).

The 2002 reform retained the focus on dual apprenticeships, in which training takes place both in schools and firms but provides occupational skills that are portable beyond the firm level. In addition, firms and their intermediary associations remain key actors in the financing and administration of collective training. Yet there were also important elements of change, which, however, primarily aimed to strengthen (rather than replace) the collective training regime (Trampusch, 2010). In the following, we highlight three important areas of reform activity: the role of the state, collective governance, and the extension of the VET system to new economic sectors.

First, the reform strengthened the state’s role in VET governance. According to the 2002 VET act, the state is now responsible for the strategic development of VET. In this new capacity, the state now demands the regular updating of all training regulations of existing occupations. In addition, the act standardized the basic principles of VET, thereby leaving it to the state to define the formal guidelines to ensure quality assurance. In turn, the state increased its financial contribution (Trampusch, 2010). It now funds pilot studies on VET-related aspects, most of which are conducted by intermediary associations in charge of individual training occupations. However, the state also strengthened the role of independent research on VET, for instance, through the new Swiss Federal Institute for Vocational Education and Training. Moreover, the state engages in extensive promotional campaigns that try to highlight the attractiveness of apprenticeships for school leavers. This clearly marks a significant change to the previous “hands off” approach (Berner, 2013).

Yet, and second, this more active role of the state did not come at the cost of collective governance. Instead, the 2002 reform aimed to strengthen collective elements—wherever possible. This is already clear in the 2002 VET act’s first article, which identifies the state and the intermediary associations in charge of training occupations as jointly responsible for VET governance. As part of this setup, several new governance forums were created to help the state shape the system in an inclusive way. The most prominent example is the annual national summit meeting on VET (Nationales Spitzentreffen der Berufsbildung). Here, the two peak-level employers’ associations and the two trade union federations meet with a broad group of other relevant stakeholders (including members of the Federal Council) to set long-term strategic goals for the VET system (Emmenegger and Seitzl, 2020: 34). The 2002 act also strengthened collective governance by giving intermediary associations the possibility to create generally binding sectoral training funds (Trampusch, 2010: 199).

Third, the 2002 reform integrated all non-university training into national VET legislation. This concerned a diverse group of economic sectors, most notably service sector occupations in areas such as health, care, and arts. The reform transferred the traditional governance model of apprenticeships to these sectors, which often implied creating new intermediary associations to take over public tasks, because such associations were missing in the newly integrated economic sectors. In this process, numerous organizations other than employers’ associations and unions were integrated into the VET governance system (Gonon and Maurer, 2012; Graf et al., 2021).

In all these processes, the government and the responsible state agencies strived for an inclusive and consensus-oriented approach (Berner, 2013). The role of unions is a case in point. Although their effective role among the intermediary associations responsible for the governance of occupations is clearly limited, the trade union federations were involved in all reform processes and are represented in all new governance forums on equal terms with the peak-level representatives of employers (Emmenegger and Seitzl, 2020). Clearly, this inclusive approach reflects the strong tradition of consultation of all stakeholders and consensus-based decision-making as well as the neo-corporatist ideology of self-governance of private actors (Katzenstein, 1985; Lehmbruch, 1984), which was never challenged throughout the whole reform process (Berner, 2013).

Although it is difficult to establish a causal relationship between the 2002 VET reform and labor market outcomes, it is clearly discernable that the Swiss VET system recovered from the 1990s apprenticeship crisis, and it is likely that the 2002 reform was instrumental in this process. After years of decline, firms’ training participation has been increasing since the turn of the century (Strebel et al., 2021: 852–853). Although firms’ participation in training does not say much about training quality, VET graduates continue to have high employment rates (OECD, 2019: 18) and training has a strong effect on wage developments (Korber and Oesch, 2019).

Conclusion

Scholarship on coordinated capitalism commonly identifies dualization or embedded flexibilization as the prime policy responses to challenges associated with the socio-economic transformation of post-industrial labor markets. This paper suggests a more optimistic assessment of the potential for CMEs to adapt to the knowledge economy, based on the potential of the state to act as an orchestrator of employer coordination. Orchestration denotes the state’s use of ideational and institutional resources to mobilize intermediary associations on a voluntary basis in pursuit of joint governance goals. This less intrusive form of state intervention is, on the one hand, made necessary because of the changing balance of power in the corporatist arena due to the weakening of unions and the empowerment of employers in a globalized economy. On the other hand, it is also in many ways a revitalization and modernization of traditional corporatist practices in collective skill formation systems, where direct and hard forms of state involvement are likely to meet considerable political resistance.

Exploring the mechanics of state orchestration of employer coordination in the context of collective skill formation, we show that in both Germany and Switzerland, the state has become a more active player in VET governance. In both cases, the state attempted to balance out power asymmetries between already involved actors as well as between established and formerly more marginal actors, all the while keeping (larger) employers involved. In addition, state orchestration provides ideational resources, for instance by publicly emphasizing the VET systems’ importance and attractiveness as well as highlighting the firms’ central role in it.

Unfortunately, the relative success of state orchestration in the two cases offers no panacea for policymakers seeking to maintain strong and inclusive skill formation systems. To some extent, the success of orchestration strategies does not only depend on the state actors’ commitment, but also on various context-specific conditions. Clearly, these conditions cannot simply be copied or transferred by state actors, no matter their commitment to building dual training in the context of the knowledge economy. For instance, in both cases, state actors are widely and credibly perceived as neutral actors. At the same time, established corporatist decision-making bodies and procedures had revealed their structural weaknesses in their initial “dualized” responses to the crisis of the late 1990s and early 2000s. Thus, the case studies are also a testament to the considerable reform capacity of CMEs, evolving from “dualization” to the more inclusive and pro-active approach of “orchestration.”

In closing, we want to briefly discuss to what extent our argument travels to other policy areas. The scope of this paper does not allow any firm conclusions on that matter, but extant scholarship lends supports for such a conclusion. Possibly the best example of state orchestration outside skill formation systems can be found in wage bargaining, where states increasingly use institutional resources to shore up employer coordination by extending bargaining coverage to non-organized firms (Paster et al., 2020; Schulten, 2018). However, state orchestration is also found in policy areas that are less tightly institutionalized but where cooperation among employers is similarly important. An example is innovation policy. Ornston (2012) points to state action in building a “creative corporatism” in the provision of supply-side resources targeted toward new actors, activities, and industries, and aimed at buttressing greater peak-level, private-public, and inter-firm cooperation in innovation and research. Further conceptual and empirical work is needed to determine the depths of the shift in state agency across a broader range of policy areas and countries.