The standard partisan model

I start my argument with a simple and by now fairly common model of how and why political parties differ with regard to their policy preferences. Going back to Hibbs (Reference Hibbs1977), the expectation is that political parties of the left and the right represent different electoral constituencies with different economic interests. In comparative welfare state research, this leads to the simple hypothesis that parties of the left, representing the economic interests of the lower income classes, will support more redistribution and a larger welfare state, whereas parties of the right will oppose these policies because they would require their constituencies in the upper income classes to pay higher taxes in order to finance the expansion of social policy.

Partisan theory can be and has been applied to the study of education policy (Ansell Reference Ansell2008, Reference Ansell2010; Boix Reference Boix1997, Reference Boix1998; Busemeyer Reference Busemeyer2007, Reference Busemeyer2009b; Rauh et al.Reference Rauh, Kirchner and Kappe2011; Schmidt Reference Schmidt and Castles2007; Wolf Reference Wolf2009). In general, and simplifying greatly, the findings that have emerged from this literature support the core claims of the theory (Schmidt Reference Schmidt1996): that there are significant differences in policy output between left- and right-wing governments, and that left-wing governments are more likely to increase public spending on and expand access to education. Some of this literature (in particular Ansell Reference Ansell2008, Reference Ansell2010 and Boix Reference Boix1997, Reference Boix1998) has focused on the dichotomy between left- and right-wing parties, since these two party families are clearly the political representatives of voters in the lower and upper income strata, respectively.

Rather than formulating hypotheses about specific policies, such as spending, I would like to start with the more general claim that in education policy, political parties differ in how much they support or oppose educational expansion. Parties on the left are expected to promote educational expansion by creating opportunities to participate in education and obtain certificates, especially for those previously excluded from the higher levels of the education system. In contrast, parties on the right, if not outright opposed to educational expansion, are certainly more reluctant to promote it, instead attempting to slow down the general trend of expansion that set in after the Second World War. I focus here on educational expansion instead of spending as such, and not just because Esping-Andersen (Reference Esping-Andersen1990: 23) famously argued that political actors do not care about “spending per se.” More importantly, fostering educational opportunities for those previously excluded from participation in education can be achieved by various means; increasing education spending is just one way to do this, and not necessarily the most effective.

In principle, there are two different strategies by which to expand educational opportunities to previously excluded groups: the first is to expand access to higher levels of education for a wider share of the population, the second to expand the kinds of education more suited to the educational needs of those formerly excluded. The first strategy is associated with conflicts about expanding access to university education, whereas the second centers on the role of VET in the education system as a whole.

This is not simply a matter of choosing between the expansion of VET and higher university education, however. Because of the inherent complexity of the redistributive implications of educational investments, partisan preferences are prima facie ambiguous as well. For example, right-wing parties may support the expansion of VET as an educational alternative in order to “divert” the excess demand for educational participation away from higher education and towards other types of education. Furthermore, right-wing parties should have an interest in limiting access to higher education, in order to ensure their educational privileges, but they might be in favor of increasing spending on higher education anyway – as long as access continues to be limited – since such investment will benefit their electoral constituencies (Ansell Reference Ansell2010). The ambiguities are even more pronounced for left-wing parties. It could be expected that these parties will support general and vocational education, the types of education that will directly benefit the educational and labor-market needs of their constituencies (Ansell Reference Ansell2010; Boix Reference Boix1998). On the other hand, left-wing parties may oppose the creation of separate and potentially dead-end educational tracks for their clientele and instead promote the opening up of access to academic higher education (Busemeyer Reference Busemeyer2009b). In short, and given these ambiguities, the specific policies supported or opposed by partisan actors will likely depend on the particular institutional context, delimiting the menu of feasible policy options, as I will argue in detail below.

The first obvious problem with the standard partisan approach is that it predicts the existence of two worlds of education regimes. Depending on how power is distributed between the left and right, one ideal-typical end of the underlying dimension will be a system with a high level of stratification and little educational mobility, while the opposite end will be a system with high levels of educational and social mobility and little stratification. However, the empirical evidence sketched out above and discussed in greater detail in Chapter 3 clearly confirms the existence of (at least) three different and distinct worlds. To understand the underlying reasons why this is the case is to step beyond simply observing the existence of “three worlds of human capital formation” (Iversen & Stephens Reference Iversen and Stephens2008; see also Hega & Hokenmaier Reference Hega and Hokenmaier2002; Willemse & de Beer Reference Willemse and de Beer2012) and highlighting the obvious connections to the relative power of Christian democracy and social democracy.

First extension: preferences regarding the political process

The standard model of partisan theory focuses on the preferences of political parties in terms of policy content (which policies they promote or oppose). Because this approach disregards whether parties promote policies in order to get elected or to implement particular policy programs (cf. Müller & Strom Reference Müller and Strom1999), the extension I propose is to think of parties as having not only specific policy preferences but also preferences with regard to how the political process should be organized. In particular, political parties and their ideologies may differ on two questions: Which interests should be granted access to decision-making (unions, employers, both, or neither)? And: Which forms of interest mediation between political and economic interests should prevail (pluralist, corporatist, or statist/hierarchical forms of interest mediation)? What is more, government parties can try to use policies as instruments to alter the balance of power between organized interests in the labor-market arena (“policies shape politics”). This expansion of perspective reveals stark differences in the ideological orientation of secular conservatives and Christian democrats.

The prevailing view in comparative welfare state research is that Christian democrats in government do increase levels of social spending and hence are not opposed to the welfare state on principle, as conservatives and liberals might be (Bradley et al.Reference Bradley, Huber, Moller, Nielsen and Stephens2003; Esping-Andersen Reference Esping-Andersen1990; Van Kersbergen Reference Van Kersbergen1995; Wilensky Reference Wilensky, Flora and Heidenheimer1981). Compared to social democrats, however, Christian democrats promote social policies that are less redistributive and are more transfer- than service-oriented. Thus, Christian democratic party ideology can easily be misunderstood as a middle class-oriented blend of conservatism and social democracy. Kees Van Kersbergen (Van Kersbergen Reference Van Kersbergen1995, Reference Van Kersbergen, Kitschelt, Lange, Marks and Stephens1999; Kalyvas & Van Kersbergen Reference Kalyvas and Van Kersbergen2010) has convincingly argued that Christian democracy should be regarded as an ideology distinct from both social democracy and conservatism; he holds that the crucial difference between Christian democrats and other parties is their particular approach to the political process, which he calls the “politics of mediation” (Van Kersbergen Reference Van Kersbergen, Kitschelt, Lange, Marks and Stephens1999: 356). Historically, Christian democratic parties mobilized using the religious rather than the economic cleavage, so they had to establish procedures of interest mediation between different economic and social interests within their party organizations. This is why they ended up promoting interest mediation and “[c]ross-class coalitions” (Van Kersbergen Reference Van Kersbergen, Kitschelt, Lange, Marks and Stephens1999: 356) as a general principle of policy-making. A second important principle of Christian democratic ideology is that of “subsidiarity”: delegating decision-making powers to autonomous civil-society associations to the greatest extent possible (Van Kersbergen Reference Van Kersbergen, Kitschelt, Lange, Marks and Stephens1999: 353).

In terms of the categories mentioned above, then, Christian democratic parties are expected to grant access to both employers’ and union representatives and to support corporatist forms of interest mediation (Wilensky Reference Wilensky, Flora and Heidenheimer1981) for the sake of maintaining cross-class compromise and cooperation. To a certain extent, Christian democratic government parties may pursue policies that seek to stabilize or create cooperation between unions and employers (e.g., passing laws that establish corporatist committees in charge of vocational training). But Christian democratic parties remain fundamentally on the “right” side of the partisan spectrum, meaning that despite their willingness to support cross-class cooperation they can be expected to grant business interests a certain privilege compared to unions. The influence of business interests and the connections between organized business and policy-makers in Christian democratic governments are likely to be very different from those in secular conservative governments, however. Christian democrats may be willing to delegate significant competences to corporatist bodies run by employers’ associations and unions, but they expect employers to take over certain public responsibilities as well (Streeck & Schmitter Reference Streeck, Schmitter, Streeck and Schmitter1985): in Christian democratic ideology, employers are given privileged access in return for their willingness to participate in the production of collective goods, such as vocational training.

Secular conservative parties are thought to differ significantly from Christian democrats in many respects. First of all, they do not subscribe to the principle of cross-class cooperation. Born out of the class conflict between capital and labor, secular conservative parties strive to minimize the influence of the left in general and unions in particular. Their preferred model of interest mediation tends to be pluralistic, based on the free competition of interests, but business interests can still expect a certain level of privileged access. Compared to Christian democrats, secular conservatives put less emphasis on the reciprocity between rights and obligations, perceiving themselves more as representatives of business interests in the policy-making arena than as a political force demanding that business contribute to collective goods. Secular conservatives may also design public policies with the goal of limiting and weakening the influence of unions, whom they consider their political enemy.

Finally, social democratic or socialist parties on the left can be expected to have different preferences with regard to the political process than do right-wing parties. Generalized expectations concerning leftist parties should be formulated in a very careful manner, since there is considerable variation in terms of ideologies across countries within the political family of social democracy (Merkel et al.Reference Merkel, Petring, Henkes and Egle2006), and since partisan preferences with regard to process also depend on the institutional context (see below); in particular, whether the institutional context offers the opportunity for cross-class compromise. At one extreme, parties on the left are the representatives of the working class in the overarching battle between capital and labor: they give unions privileged access to policy-making, reject cross-class compromise, and are willing to use public policies to limit the influence of business (the pre-Blair Labour Party in the United Kingdom might serve as a rough real-typical example). At the other end, left-wing parties can become active supporters of cross-class compromise, understanding that the political sustainability of the universal welfare state model depends on broad supporting coalitions between different economic classes (Esping-Andersen Reference Esping-Andersen1990; Rothstein Reference Rothstein1987). I would therefore argue that institutional context matters more in the case of social democratic parties, as a determinant of procedural preferences. The difference between Christian democrats and secular conservatives, by contrast, reflects inherent differences in terms of ideology.

Second extension: the role of institutional context

A look at the impact of institutional context will lead us to the second extension of the standard model in partisan theory. First, however, we must clarify what is meant by “institutional context.” Here the concentration is on context in terms of socioeconomic institutions; that is, the prevailing level of economic coordination (Hall & Soskice Reference Hall, Soskice, Hall and Soskice2001). As is well known from the VoC literature, high levels of coordination essentially entail a higher degree of cross-class and nonmarket-based cooperation between the organized labor-market interests. On the one hand, this implies differences in the patterns of interest mediation between LMEs and CMEs, as argued above, and shapes partisan preferences and strategies with regard to how the political process should be organized. On the other hand, institutional context also affects the menu of feasible policy options. In other words, economic coordination entails two aspects: one that is concerned with the collective organization – and ultimately also the power resources – of labor-market actors, such as unions and employers; and one that emphasizes the institutional constraints and complementarities.

In general, I posit that high levels of coordination contribute to making VET a sustainable educational choice and a worthwhile alternative to academic higher education for those making their way up the educational ladder. The conventional argument for why this is the case is that coordination allows labor-market actors to make credible long-term commitments to each other (Cusack et al.Reference Cusack, Iversen and Rehm2006; Finegold & Soskice Reference Finegold and Soskice1988; Iversen & Soskice Reference Iversen and Soskice2001). Employees, so the argument goes, demand some kind of insurance before they are willing to invest in occupation- or firm-specific skills, through either a generous welfare state or an employer's credible commitment to lifetime employment (Estévez-Abe et al.Reference Estevez-Abe, Iversen, Soskice, Hall and Soskice2001; Iversen & Soskice Reference Iversen and Soskice2001). Conversely, employers who invest in skill formation need assurance that employees will stay with the firm beyond the training period, so that the company can recoup its investment (Acemoglu & Pischke Reference Acemoglu and Pischke1998).

A slightly different argument I would like to put forward here is that high levels of coordination also have an impact on the content of VET. The participation of employers’ associations and labor unions in the design of VET curricula, both in the firm-based part and in the school-based component, ensures that the skills that are produced in the VET sector fit the needs of the employment system. When there is a high match between the skills taught in VET and the skills demanded on the labor market, choosing VET over general academic education becomes attractive, especially for those in the middle strata of the skills distribution. This is because a tight linkage between VET and the employment system enables smooth transitions from education to employment (i.e., low levels of youth unemployment) and increases the payoff on educational investment for those in the lower half of the skills distribution (Breen Reference Breen2005; Gangl Reference Gangl, Müller and Gangl2003; Wolbers Reference Wolbers2007). When VET is not regarded as an educational dead end, but rather as a gateway to relatively attractive positions at the intermediate skill level, it increases the viability and sustainability of VET as a credible alternative to higher education. The existence of a well-developed VET system also changes labor-market structures: instead of the polarization between high-skilled services and low-skilled jobs that is often found in LMEs, CMEs have seen a strong segment of employment at the intermediate skill level (Marsden & Ryan Reference Marsden and Ryan1990).

The cooperation between unions and employers’ associations is a crucial point in the design of VET curricula. At a very general level, employers are expected to demand more specific skills (or at least a broad foundation of vocational skills with a strong specific component), whereas unions tend to favor broad skills. The simple reason for these differences in interest is that unions (and employees) have an interest in maintaining or increasing the labor-market mobility of employees, since labor-market mobility is related to bargaining power (Streeck Reference Streeck and Regini1994), whereas employers have an interest in minimizing the mobility of skilled employees (while maintaining the ability to fire the less-skilled ones), since this depresses employees’ power to demand higher wages. The compromise that is likely to emerge from these interest-driven bargaining processes over the content of VET is one of occupational profiles that combine broad vocational skills with a firm-specific component. Incidentally, these types of “polyvalent” (Streeck Reference Streeck, Dore and Berger1996: 141) and yet specific skills have given a competitive advantage to firms engaging in “diversified quality production” (Streeck Reference Streeck1992), meaning high-end manufacturing and similar activities. Coalitions between unions and employers are also important in terms of political influence. Together, unions and employers can lobby policy-makers to continue supporting VET, since this type of education has traditionally been closer to the economic and organizational interests of labor-market actors than has higher education.

In the absence of cross-class coordination, by contrast, the linkage between VET and the employment system is weaker, decreasing the attractiveness of VET as an educational choice because of the less favorable labor-market outcomes. When the content of VET is mostly designed by employers, without the involvement of unions, there is a danger that training content becomes too specific. This decreases the attractiveness of VET from the perspective of employees/young people, because it does not increase labor-market mobility and the potential to invest in transferable skills. Conversely, when the state or the unions are the only ones in charge of designing the skills content, it may become too general and broad, decreasing the attractiveness of VET from the perspective of employers, who demand a certain degree of skill specificity. When unions are in charge of regulating access to training, moreover, they may abuse this position in order to limit the supply of skilled labor, so as to drive up wages (Thelen Reference Thelen2004).

It is important to remember that high levels of coordination as such are not a sufficient explanation for the variety of education and training systems in Western Europe. Although cross-class coalitions between unions, employers, and state actors are common in CMEs, the long-term balance of power between partisan actors remains an important factor. In other words, government parties can and do influence the politics of institutional change by selectively and carefully privileging certain organized interests, even when the general institutional framework may remain corporatist. This consideration of partisan power in corporatist settings implies a crucial change in perspective that distinguishes my approach from the classical power resources and VoC perspectives by building on and extending Thelen's coalitional approach (Thelen Reference Thelen2004; see also Hall & Thelen Reference Hall and Gingerich2009). Unlike power resources theory, I acknowledge the importance of cross-class coalitions. Unlike the VoC school, I posit that cross-class coalitions may still be partisan in nature.

Third extension: long-term balance of power

Finally, I would like to draw attention to the role of time and timing. A common research design in quantitative studies of the effect of partisan politics on policy output is to include the current partisan composition of governments as an independent variable in regression models of some kind of policy output (cf. Schmidt Reference Schmidt1996). Whereas early contributions to the literature were able to identify systematic partisan effects (e.g., Castles Reference Castles and Castles1982, Reference Castles1989; Schmidt Reference Schmidt1982), more recent work has observed a weakening of these effects (Kittel & Obinger Reference Kittel and Obinger2003; Kwon & Pontusson Reference Kwon and Pontusson2010), which might be explained by the constraining impact of globalization, a general convergence of party positions to the median, or other factors. A further explanation for the disappearance of partisan effects in the short term is the buildup of policy legacies and path dependency over time (Pierson Reference Pierson1994, Reference Pierson1996; Rose & Davies Reference Rose and Davies1994). Thus, even in cases where parties come to government with a clear mandate to change policies, they often find it hard to do so because they need to deal with a large and ever-increasing set of entrenched interests and policy legacies.

In short, timing matters: the intensity of the impact of partisan forces on policy and institutional change is likely to vary across time. In the early phases of path formation, when institutional arrangements are still being established and are not yet fully developed, there is greater room for political agency (Pierson Reference Pierson2000). This means that partisan politics (or the balance of power between political interests) is likely to matter more during critical junctures of historical development.

How can we distinguish periods of path formation from periods of path consolidation, when large-scale change becomes increasingly unlikely? In broad terms, societal and economic development have proceeded alongside and in interaction with welfare state development. During the Industrial Revolution, the welfare state provided insurance against basic risks of health and safety and expanded to cover more and more life risks, such as unemployment and old age. The development of economies went along with a greater demand for education and skills (Goldin & Katz Reference Goldin and Katz2008), which led to the expansion of educational opportunities at the higher levels of the education system. The postwar decades in Western Europe were a critical period of path formation, in the sense that access to formerly elitist and closed post-secondary education systems was opened up to meet the increasing demand for education, both from business and from households. We can see from this that path formation occurs when the welfare state expands to address the new social needs and demands that arise from societal and economic development. The specific response of policy-makers to these new demands then depends on the balance of power between partisan interests.

A second point relates to the role of time, but not timing as such. Modern societies are extremely complex social systems. They are not mechanistic but organic in nature, which limits the leeway for policy-makers to influence societal developments or institutional change. Like “institutional gardeners,” policy-makers can use their hedge trimmers by changing formal regulations such as laws and decrees. Formal legal change, however, does not directly translate into actual institutional change, because economic or other social actors can react to policy change in a manner that is very different from the policy-maker's original intentions (cf. Streeck & Thelen Reference Streeck, Thelen, Streeck and Thelen2005). To provide a concrete example from the study of education systems: policy-makers can change the formal structure of education systems by creating a new educational pathway, for instance an apprenticeship-training track, but whether this educational pathway is actually accepted by employers and households is an entirely different matter. In open and democratic societies, educational choices are the privilege of individuals and their families. Policy-makers can only indirectly influence these choices, by manipulating the formal structure of educational institutions. Given this state of affairs, educational reform is a long-term project. Education is a central concern to many families. Structural changes in the education system usually need to be implemented over a period of several years, and large-scale change can only be implemented in a stepwise and incremental manner, in order to give individual households time to adjust. To put it another way: attempting to effect large-scale reforms in a short span of time is akin to political suicide and will most likely not succeed.

The upshot of this discussion of time and timing is that my empirical analysis will focus on the role of partisan politics and politico-economic coalitions in the formative years of the postwar decades, when critical decisions about the future development of education and training systems were being made. Futher, instead of paying too much attention to the short-term effects of the current partisan composition of governments, I will concentrate on differences in the long-term balance of power between different countries/regimes, which are expected to be more important for the development of education regimes. Of course, political parties may adopt their policy goals to changing circumstances in the long term. Parties’ electoral constituencies can also change, reflecting changes in the social structures of societies. These problems are mediated by the fact that I focus on partisan conflict in the postwar period, which was characterized by tighter linkages between electoral constituencies and their representative parties. Partisan politics in the postindustrial age is more complex, because of the increased importance of value conflicts (Häusermann et al.Reference Häusermann, Picot and Geering2013). Still, I would argue that basic differences between leftist and rightist parties that are related to fundamental issues such as inequality and redistribution are also discernible in the contemporary period and are rooted in the material interests and normative considerations of different electoral constituencies, even though these may have changed compared to the industrial age.

Putting the pieces together: hypotheses

The core argument for the following analysis is that different types of politico-economic coalition were responsible for shaping education and training policies during the critical juncture of the postwar decades. As a reminder, the core dimensions of variation are public involvement in education and training on the one hand and institutional stratification on the other (see Figure 1.1). The long-term dominance of one particular party family had an impact on the extensiveness and partisan nature of politico-economic coalitions, since parties have preferences with regard not only to policy content but also to how the political process should be organized. The prevailing institutional context in terms of economic coordination shaped the menu of feasible policy options. These propositions lead us to the following hypotheses/expectations:

Postwar consensus and institutional consolidation (1945–1964)

The immediate postwar era in Britain was characterized by a widespread consensus between parties in matters of education policy. The legislative foundation of this consensus was the 1944 Butler Education Act, passed by the wartime coalition government after several years of consultation with societal stakeholders (Chitty Reference Chitty2004: 18). The progressive thrust of the Act was meant to achieve “secondary education for all” by, among other things, abolishing tuition fees for secondary schools. Both Labourites and Conservatives agreed on the desirability of expanding educational opportunities at the secondary level in the 1950s (Gordon et al. 1991: 62–4). This consensus was in part based on the fact that the Butler Act remained very vague with regard to the specific institutional design of secondary schools (Chitty Reference Chitty2004: 19). Although the postwar Labour government had instigated large-scale radical change in the British welfare state by establishing the National Health Service, it had remained rather timid with regard to reforming education. The implementation of the Act was largely left in the hands of the Local Education Authorities (LEAs), and Prime Minister Attlee himself did not support the comprehensive schools and favored maintaining public schools (which are, in fact, private schools in British parlance) (Lawton Reference Lawton2005: 49).

Since the Conservative government was even less eager to intervene in local affairs, the education systems had the character of a “national system, locally administered,” based on “a benign partnership between central government, local government, and individual schools and colleges” (Chitty Reference Chitty2004: 21). The decentralized and essentially voluntarist character of the education system encouraged great variety in secondary schools, which effectively constituted a tripartite system: the elitist grammar and public schools, the new modern secondary schools, and the technical and trade schools. Access to the more prestigious types of state school was based on a competitive examination at age 11 (called 11-plus) (Cheung & Egerton Reference Cheung, Egerton, Shavit, Arum, Gamoran and Menahem2007: 201).

The Labour Party in opposition was initially divided on the issue. Assigning pupils to different schools on the basis of academic merit was regarded as an advancement over the previous system, where school choice depended on the discretion of teachers and parents (Baldi Reference Baldi2012: 1006). Over the course of the 1950s, however, critics of the tripartite system asserted a growing influence within the Labour Party, particularly after the Crowther Report (1959) and similar sociological research produced evidence that, controlling for academic ability, family background determined access to grammar schools (Baldi Reference Baldi2012: 1006–7; Gordon et al.Reference Gordon, Aldrich and Dean1991: 72; Lawton Reference Lawton2005: 57).

The climate of educational expansion penetrated both secondary and higher education. As in other countries, access to higher education was limited to a small minority of about 5 percent of a typical age cohort (Ansell Reference Ansell2010: 197), but public pressure to allow more young people to enter higher education mounted over the course of the 1950s, especially within the traditionally Conservative electoral constituencies in the upper middle class (Ansell Reference Ansell2010: 197). In 1963, a committee led by Lord Robbins recommended that access to higher education be opened to all who could benefit from it (Cheung & Egerton Reference Cheung, Egerton, Shavit, Arum, Gamoran and Menahem2007: 196). The Conservative government implemented most of the recommendations from the Robbins Report and continued to support higher education institutions with public grants. According to Ansell (Reference Ansell2010: 198), this seemingly progressive policy in fact benefited the Conservatives’ electoral constituencies in the upper income classes (see also Gordon et al.Reference Gordon, Aldrich and Dean1991: 79), because even after the expansion, only a small minority of the relevant age cohort (about 15 percent) would enroll in higher education and therefore benefit from the public subsidization of universities.

VET remained disconnected from the general education system. The curricula of most secondary schools were still committed to a liberal arts education in preparation for university studies. The technical schools catered to those with low academic abilities and were oriented towards vocational skills. Unlike the German Hauptschule, however, they covered only a small share of the student population (about 4 percent of the secondary age group in 1958; see Chitty Reference Chitty2004: 25), so that there was effectively little to no vocational education in the general education system.

This made the matter of training primarily the responsibility of the economic actors themselves. As in other European countries, Britain had a tradition of apprenticeship training going back to medieval times (Perry Reference Perry1976). The institutional setup of apprenticeship in the United Kingdom remained quite traditional until well into the postwar period: advancing from an apprentice to a journeyman was based not on the passing of externally administered exams, but simply on time served (Gospel Reference Gospel1994: 508–9). British unions are organized as crafts unions; that is, as unions of skilled workers in a particular craft or occupation. This differs from the industry-based organization of unions in Germany (Thelen Reference Thelen2004: 98–100). The nature of the UK system led unions there to develop an interest in limiting the supply of skilled workers so as to drive up wages: they used apprenticeship training as a gateway to skilled labor markets and enforced a strict separation between skilled and unskilled workers (Thelen Reference Thelen2004: 105–6). This is why the traditional system of apprenticeship training came under pressure in the postwar period, as demographic developments increased the need for more training opportunities. Critics also lamented the narrow nature of training, which was tied to a particular occupation and lacked reference to common standards (Gospel Reference Gospel1994: 508–9). The unions’ hold on apprenticeship training also led employers to look for new ways to limit the role of skilled labor in production processes and effectively prevented the emergence of a lasting cross-class compromise on matters of training during the period of industrialization in the late nineteenth century (Thelen Reference Thelen2004: 107).

Despite the more contentious nature of training policies in Britain, unions and employers actually managed to come to a consensus immediately after the Second World War, when a joint report by the British Employers’ Confederation and the Trades Union Congress (TUC) recommended that National Joint Apprenticeship Councils be established (Perry Reference Perry1976: 49). Unions supported the establishment of collective institutions, but at the same time were careful to preserve their strong bargaining position in controlling access to skilled labor markets (Perry Reference Perry1976: 60). British employers (as well as policy-makers) were increasingly worried about deficiencies in competitiveness compared to continental European countries such as France and Germany, and the issue of training and “manpower” policy was believed to be crucial in dealing with these problems (Corina Reference Corina, Hayward and Watson1975: 194). Nonmarket forms of coordination among employers were never as strongly developed in the United Kingdom as in Germany, since British employers’ associations remained voluntary in nature, in contrast to the statutory character of German Chambers of Industry and Commerce. But multi-employer bargaining in matters of wage policy was nevertheless common during the postwar decades, and it is believed to have been a crucial supporting factor for apprenticeship training (Gospel Reference Gospel1995: 41).

Demographic factors and the threat of youth unemployment continued to exert pressure on economic actors and policy-makers to reform the training system in the 1950s and early 1960s. The government refrained from intervening in industrial training at first; its position had been supported and confirmed by the report of the Carr Committee in 1958, and training was believed to fall under the purview of unions and employers (Perry Reference Perry1976: 66). The ensuing debate, however, led to a transformation of the general political climate and government policy within just a couple of years.

In 1964, the Conservative government passed the Industrial Training Act, based on broad support from unions and employers, as well as the Labour Party in opposition (Perry Reference Perry1976: 108, 125), reflecting a “general economic, social and ideological climate” (Evans Reference Evans1992: 191) more favorable to expanding the role of the central state in manpower planning. The Act was considered “one of the largest-scale attempts at structural change in Britain” (Corina Reference Corina, Hayward and Watson1975: 195). It led to the establishment of a number of Industrial Training Boards (ITBs; 27 in total by 1969; see Perry Reference Perry1976: 175), which were in charge of administering and financing firm-based training. To this end, ITBs were empowered to impose a training levy on firms in their own economic sector, the revenues of which would be used to subsidize out-of-firm training courses and the Boards’ activities in devising training standards and supplying training materials. As bureaucratic organizations with a significant degree of autonomy from the government, ITBs developed their own practices over time, such that levy practices “became extremely variable across Boards, growing up in a haphazard way” (Corina Reference Corina, Hayward and Watson1975: 197).

If we disregard the implementation of the 1964 Act and look only at the formal set of institutions it created, we can see that the British training system after 1964 exhibits many characteristics of collective skill-formation systems (Busemeyer & Trampusch Reference Busemeyer2012), and in fact surpasses what Germany achieved a couple of years later with the enactment of the Federal Law on Vocational Education and Training (the BBiG). Employers in Germany successfully fought against the introduction of a training levy in the 1970s, for example, because they feared that it would bring too much state intrusion (Busemeyer Reference Busemeyer2009c: 79–106). Nevertheless, the British ITB system retained a certain voluntaristic character in the sense that the Boards could not force firms to provide training, but were restricted to supporting the training efforts of individual firms through subsidies, the provision of training material, or the organization of out-of-firm training courses. The reform also did not establish a binding national framework of qualifications or occupational profiles, as happened in Germany – or much later, at the end of the 1980s, in the United Kingdom.

The case of the 1964 Training Act remains highly interesting, because it shows that policy-makers and economic actors in the United Kingdom did attempt to build up collective institutions in skill formation (as was the case in the United States at an earlier moment in its history; see Martin Reference Martin2006). The British efforts to establish a collective skill-formation system were based on cross-class compromise between unions and employers, as well as between the Conservative government and Labour opposition. The partisan consensus proved to be short-lived, however, as will become clear in the next section. Furthermore, the pluralistic and fragmented character of industrial relations (their low degree of nonmarket coordination) prevented the short-term consensus between unions and employers from becoming sustainable in the long run. This shows the importance of the institutional context in terms of economic coordination as a constricting factor of partisan politics: as the ideological orientations of political parties rooted in class conflict became more influential in the years that followed, the partisan consensus in education policy proved to be transitory.

Increasing partisan conflict and crisis (1964–1979)

Starting in the mid-1960s, the postwar consensus on education policy, economic policy, and social policy came under increasing pressure, which was further aggravated by the massive economic crises of the 1970s and their high levels of unemployment and inflation. When Labour returned to power in 1964, one of the first actions of the new education minister, Anthony Crosland, was to issue Circular 10/65, which put more pressure on LEAs to establish comprehensive schools. Despite his reputation as a “pace-setter” (Gordon et al.Reference Gordon, Aldrich and Dean1991: 82), Crosland granted LEAs considerable leeway in implementing the governmental directive. By this time, a bottom-up movement that favored comprehensivization based on “a coalition of parents, local politicians, and grassroots opinion” (Gordon et al.Reference Gordon, Aldrich and Dean1991: 189) was gaining momentum. As comprehensive schools spread more widely, especially in Labour-controlled LEAs, teachers at comprehensive schools became more influential within the Labour Party and contributed to transforming the party's formerly ambiguous stance on the issue into a more assertive position of support (Fenwick Reference Fenwick1976: 114–15).

The issue of whether to promote comprehensive schools was now splitting the Conservatives (Knight Reference Knight1990: 72). Centrist moderates such as Edward Boyle were supportive, because of comprehensive schools’ general popularity. Margaret Thatcher, however, was radically opposed to comprehensivization. When she came into power as minister of education with the change of government in 1970, one of her first actions was to rescind Circular 10/65, but this did not stop the expansion of comprehensive schools (Fenwick Reference Fenwick1976: 147; Gordon et al.Reference Gordon, Aldrich and Dean1991: 193). One obvious indication of the breaking up of the partisan consensus on education reform was the so-called “Black Papers,” published beginning in the late 1960s by a group of Conservative educationalists representing the “Establishment Tories” (Ansell Reference Ansell2010: 199). The Black Papers criticized certain aspects of comprehensive education and the general expansion of access to higher education (Lawton Reference Lawton2005: 85).

Perhaps this is one of the reasons why Labour acted more cautiously and refrained from enacting more large-scale reforms once it was back in government in 1974. According to Lawton (Reference Lawton2005: 96), the period of Labour government between 1974 and 1979 must be regarded as an “education failure”: “[Prime Minister] Wilson seemed to be uninterested in the details of school education; [Prime Minister] Callaghan was interested, but adopted a technicist, unsocialist stance, neglecting to consider the importance of education for society as a whole rather than simply concentrating on the needs of industry.” In 1976, Prime Minister James Callaghan gave a famous speech at Ruskin College at Oxford University that is widely regarded as a “turning point” in British education policy (Lawton Reference Lawton2005: 91), because it marked “the end of the period of educational expansion” (Chitty Reference Chitty2004: 44) and contained a challenge for schools to provide more skills that were directly relevant to industry. In reality, however, the ensuing “Great Debate” on education reforms produced a large number of reports and White Papers, but little concrete policy output until the advent of the Thatcher government (Gordon et al.Reference Gordon, Aldrich and Dean1991: 95–6).

The Labour government of the 1960s also introduced a binary system of higher education institutions: “an ‘autonomous’ sector consisting mainly of old and new universities and colleges of advanced technology (which acquired university status in 1966–7); and a ‘public’ sector under local authority control and represented by the leading technical colleges and the teacher-training colleges, to be known in future as colleges of education” (Chitty Reference Chitty2004: 164). The Labour government set up “30 polytechnics intended to provide advanced vocational education and respond to local labor market and industrial research needs” (Cheung & Egerton Reference Cheung, Egerton, Shavit, Arum, Gamoran and Menahem2007: 196). Lord Robbins, who a couple of years earlier had forcefully advocated the expansion of access to higher education, expressed concern about this new Labour policy at a conference in 1965:

From a redistributive perspective, however, the Labour policy of promoting vocational education at the tertiary level made sense. Instead of expanding academic universities, which catered to children of the upper income classes, the new type of higher education institution could potentially attract new groups of students from more disadvantaged backgrounds. What is more, the new polytechnics would be placed under the direct control of LEAs, whereas the traditional universities remained largely autonomous, giving state actors less control over their admissions policies. The price for this differentiation was, of course, that polytechnics were regarded as “subordinate institutions” compared to the traditional universities (Gordon et al.Reference Gordon, Aldrich and Dean1991: 241).

The Labour government of the 1960s was also responsible for establishing and developing further the ITB system (Evans Reference Evans1992: 208). Critics of the system grew louder towards the end of the 1960s, after ITBs had been set up for most economic sectors (with the notable exception of banking and finance). Small firms in particular chafed under the levy-grant system (Perry Reference Perry1976: 273). Once the Conservatives took over, they commissioned a review of the ITBs, which together with the related Green Paper actually recommended the wholesale abolishment of the levy-grant system (Perry Reference Perry1976: 285), as could be expected from a business-friendly Conservative government. The Confederation of British Industry (CBI), on the other hand, preferred to keep the ITB structure as the “lesser of two evils,” since unions had been pressing for a more centralized approach to training policy for a long time, and with increasing success (Perry Reference Perry1976: 292). As a consequence, the 1973 Employment and Training Act retained the levy-grant system in principle but made it much easier for firms to opt out. More specifically, “the criterion for exemption [from the levy-grant scheme] was a firm's training activity measured in terms of its own needs; and only if these were deemed not to have been met would a liability for a levy arise” (Perry Reference Perry1976: 301). Given this ability to define their own needs, individual firms could more or less decide by themselves whether to participate in the scheme or not, causing the levy-grant system to be “effectively abandoned” in 1973 (Lee Reference Lee and Wood1989: 158). The Act also established the Manpower Services Commission (MSC), fulfilling long-held demands by unions for a more centralized approach to training policy (Evans Reference Evans1992: 18) and resulting from the CBI finally coming around to lending its support (Perry Reference Perry1976: 301).

Another major change in the British training system was the decline of traditional apprenticeship training. This change, at least in the beginning, happened not as a consequence of policy-making, but due to structural changes in the economy. The decline of traditional manufacturing and the rise of the service economy led to a severe downfall in the number of apprenticeship places, from 218,000 in 1970 to 73,000 in 1985 (Lee Reference Lee and Wood1989: 159). Union attempts to limit access to the skilled labor market and drive up wages triggered adversarial reactions by employers and the government (Lee Reference Lee and Wood1989: 159). As wage-bargaining institutions became increasingly fragmented in character, moreover, the wages of apprentices began to approach those of unskilled employees, making employers less willing to invest in the training of young workers (Marsden & Ryan Reference Marsden and Ryan1990, Reference Marsden, Ryan and Ryan1991). Confronted with high levels of youth unemployment, the newly founded MSC started sponsoring specific labor-market programs such as the Youth Opportunities Programme (YOP). These would be massively expanded in the 1980s.

In hindsight, the 1970s can be regarded as the onset of a critical juncture. This had less to do with Callaghan's Ruskin speech, however, than with the changes that happened in the field of training policy: the effective abandonment of the short-lived collectivist levy-grant scheme in 1973, the decline of traditional apprenticeship training, and the setup of publicly sponsored labor-market programs to deal with youth unemployment in a voluntarist framework. All of these trends would become much more pronounced in the 1980s.

Neoliberal transformation of education and training policy (1979–1997)

As far as general schooling goes, the 1988 Education Reform Act (ERA) is widely regarded as a “historically significant” milestone, “because it represents the culmination of a break with the consensus politics of education which had prevailed from 1944 to 1979” (Lawton Reference Lawton1992: 59–60). The ERA fundamentally changed the role of LEAs by shifting “power away from LEAs to the central authority, contrary to the established tradition of avoiding too much centralized control” (Lawton Reference Lawton1992: 47). In the old system, “[c]ontrol over the content of education was left, instead, in schools in the hands of head teachers, who in turn tended to leave what went on in the classroom to individual teachers” (Ainley Reference Ainley2001: 458). Based on the general mistrust of public administrators and officials common in neoliberal ideology, the ERA was aimed at centralizing control over the provision of education on the one hand and maximizing opportunities for choice by strengthening the role of consumers (in this case, parents and students) on the other (Gingrich Reference Gingrich2011). The result was the emergence of a more complex governance structure (Ainley Reference Ainley2001). The Act also created a new, privileged class of “grant-maintained” schools, which were funded directly from the central government and operated independently of the local control of LEAs. Grammar schools (and former grammar schools) were overrepresented in the group of schools that applied for this new status (Gingrich Reference Gingrich2011: 138–9).

The Conservative government also committed to deeply transforming the higher education system. Williams considers the period between the end of the Second World War and the advent of the Thatcher government the “golden age” of British universities, because they then enjoyed a considerable degree of autonomy and were generously financed by public funds (Williams Reference Williams, Teixeira, Jongbloed, Dill and Amaral2004: 243). The early years of the Thatcher government saw massive cuts in spending on higher education, and the government used the purse strings to effect changes in the universities (Williams Reference Williams, Teixeira, Jongbloed, Dill and Amaral2004: 244). As the share of core funding via maintenance grants from the central government fell from 70 percent in the 1960s to just over 30 percent in the mid-1990s (Cheung & Egerton Reference Cheung, Egerton, Shavit, Arum, Gamoran and Menahem2007: 200), the universities turned to new sources of revenue, such as research contracts, short courses, and conferences, as well as student fees. These fees, however, were initially still paid by the central government and administered through the local authorities (Cheung & Egerton Reference Cheung, Egerton, Shavit, Arum, Gamoran and Menahem2007: 200).

Changes in the funding mechanism fueled competition between universities for students and research funding. As student fees became a more important source of revenue, universities could effectively increase their funding by admitting more students, which led to “an explosive expansion of student numbers” (Williams Reference Williams, Teixeira, Jongbloed, Dill and Amaral2004: 247) in the late 1980s and early 1990s (Mayhew et al.Reference Mayhew, Deer and Dua2004: 66). After the ERA of 1988 upgraded the status of polytechnics, the 1992 Further and Higher Education Act finally abolished the binary structure of the system by putting the former polytechnics and universities under the same funding scheme, taking control over the former polytechnics away from local authorities (Cheung & Egerton Reference Cheung, Egerton, Shavit, Arum, Gamoran and Menahem2007: 198). Although the system was now less stratified in formal and legal terms, a de facto hierarchy of higher education institutions quickly developed, which was much more pronounced than before and was determined by the relative successes of institutions in attracting good students and research funding (Cheung & Egerton Reference Cheung, Egerton, Shavit, Arum, Gamoran and Menahem2007: 199). Responding to the large increase in student numbers in the early 1990s, the government introduced a cap on the maximum number of students to be admitted in 1994/95 (which was removed again in 2001/02; Mayhew et al.Reference Mayhew, Deer and Dua2004: 69). This is a nice example of partisan differences in the promotion of educational expansion: the Conservative government initially expanded access to higher education, but then restricted it once a critical threshold had been reached. The (New) Labour government, in turn, abolished that restriction.

The Conservative government fundamentally transformed vocational training policies as well, although some changes had begun in the 1970s. The Thatcher government, however, decisively and deliberately cleared away the institutional remnants of coordinated capitalism and collective approaches to skill formation in particular, and thus firmly consolidated the British development path as a liberal skill regime. Youth unemployment remained persistently high in the early 1980s, which persuaded the government to keep the MSC alive for the moment (Evans Reference Evans1992: 45). But instead of reviving the collective ITB scheme, the Thatcher government sought to abolish the training boards (Finn Reference Finn1987: 135) and expand training measures, with a strong focus on fighting youth unemployment. The general aim of Conservative training policy was to minimize the role of unions by undercutting the traditional apprenticeship system in particular, as well as by getting rid of the institutional legacy of collective skill-formation institutions in order to give individual employers more flexibility to make their own decisions about training (Dingeldey Reference Dingeldey1996; Evans Reference Evans1992: 152, 209; King Reference King1993: 215).

The MSC, which was established in 1973 as a genuinely tripartite institution, with equal participation of unions, employers, and state actors, was increasingly used by the Conservatives as a labor-market agency to sponsor programs for unemployed youth (Evans Reference Evans1992). With its New Training Initiative (NTI), the MSC initially proposed a collective approach to solving the problem of youth unemployment. The government's version of the NTI, published as a White Paper in 1981, strengthened the workfare component by threatening to withdraw benefits to youths who did not accept the training opportunities offered by the MSC's programs (Finn Reference Finn1987: 154–5). The YOP, mentioned above, had by this time become quite unpopular with youths, because it suffered from problems in the quality of training. Nevertheless, a tripartite task force at the MSC proposed to introduce a new version of the YOP in 1982 – the Youth Training Scheme (YTS) – which was accepted and supported by the government as a concession to unions and employers (Finn Reference Finn1987: 158). The YTS provided employers with funds to train people under the age of 18 for one year (this was extended to two years in 1986), guaranteeing a period of thirteen weeks of off-the-job training (seven weeks in the second year). Area Manpower Boards (AMBs) under the auspices of the MSC were responsible for administering and monitoring in-firm training, but the actual provision of training was outsourced to a number of training providers. These could be individual employers, but were more often groups of employers, further colleges of education, or commercial training companies. The number of youths participating in the YTS in the 1980s and the sum of public funds devoted to it were significant, although not as high as in the German apprenticeship system (Ryan & Unwin Reference Ryan and Unwin2001). At the peak of its power, in 1986, the MSC had a budget of more than £2 billion, of which about £850 million was spent on the YTS (Evans Reference Evans1992: 83). Indeed, by some measures it proved a successful short-term remedy for youth unemployment: an estimated 60 percent of trainees either stayed with their initial training sponsor or found employment elsewhere (Lee Reference Lee and Wood1989: 161).

But the scheme was heavily criticized by unions and the youths themselves. In 1985, about 200,000 pupils in over sixty towns protested against the YTS and the proposal to make it compulsory (Finn Reference Finn1987: 182). One widely uttered criticism of the scheme was its focus on quantity of training, instead of quality. While the MSC and its AMBs were formally in charge of monitoring, these organizations did not in fact have the necessary administrative capacities to effectively monitor the implementation of the scheme on the ground (Finn Reference Finn1987: 175). Neither were there any clearly defined standards on the content of training to be provided, because the original MSC task force had pursued a strategy tantamount to “black boxing”: “Black boxing meant that what trainees do on work placements is not directly vetted but placed largely at the discretion of the trainee's employer. It is constrained only by agreed ‘end-product’ criteria which the MSC has found difficult to enforce” (Lee Reference Lee and Wood1989: 166).

The lack of standards resulted in a huge variation in the quality of training provided. At the top of the hierarchy were manufacturing firms, which used the opportunity to replace traditional apprenticeship training schemes with new YTS-sponsored training (Lee Reference Lee and Wood1989: 164). Firms in the service economy were at the low end of the hierarchy, effectively using YTS trainees as a source of cheap labor without formally investing in skill formation (“bogus training”) (Lee Reference Lee and Wood1989: 165). All in all, the YTS marked a “watershed in the transition from the employer-led style of apprenticeship to a State-led approach” (Fuller & Unwin Reference Fuller and Unwin2009: 413; see also Unwin Reference Unwin1996) in the financing and provision of training as labor-market policy, even though the stated policy goal had been to promote employer involvement and leadership.

The government's strategy to undermine traditional apprenticeship training and replace it with employer-led, but government-sponsored, training schemes started with the YTS scheme and was carried to the next level by the reforms of the late 1980s. The continued electoral success of the Conservatives, coupled with union criticism of the YTS scheme, “ended any government willingness to maintain tripartite links with unions and employers” (King Reference King1993: 226–7) and effectively dissolved the tripartite MSC; it was transformed into a Training Agency within the Department of Employment. In fact, this step was long overdue, because “the MSC had enjoyed a quite anomalous existence during the first decade of the Thatcher government, explicable mainly because it solved problems during a difficult era of transition which other agencies were unable to address” (Evans Reference Evans1992: 127).

In 1988, the government established Training and Enterprise Councils (TECs, called Local Enterprise Councils in Scotland) as the successors to ITBs. Unlike ITBs, TECs were private companies in charge of both local economic development and vocational training, with the latter often confined to a secondary role (King Reference King1997: 398). TECs were dominated by representatives from industry; neither unions nor other stakeholders such as local authorities had a statutory right to be included. TECs received funding from the central government for training programs and then acted as managing agencies in outsourcing the provision of actual training (King Reference King1993: 232–3). They lacked the power to intervene in firms’ training practices, however. Since the voice of organized labor had already been marginalized, firms could “continue with their own training in traditional voluntarist style (advantaged by the weakening of trade unions)” (King Reference King1997: 401). Thus, even though TECs were supposed to be employer-led organizations of self-governance, they very much depended on continued government support (Keep Reference Keep2006: 51) and represented a less dirigiste and more voluntarist approach to training policy (Coffield Reference Coffield1992; Evans Reference Evans1992: 127, 138).

The government began developing a system of National Vocational Qualifications (NVQs) to make training quality and standards more uniform. Industrial training organizations (ITOs) were created by the state; these were supposed to develop NVQs, but their relationship with the TECs remained unclear, leading to conflicts and competition (Evans Reference Evans1992: 181). NVQs were awarded by a central body, the National Council for Vocational Qualifications (NCVQ), and later the Qualifications and Curriculum Authority (QCA). The new system reflected an output-oriented and modular approach to training, in the sense that trainees and training firms were free to combine training modules as they pleased and to accumulate qualifications by demonstrating the ability to perform skills as stipulated in the NVQs. This stands in contrast to the occupational principle (Berufsprinzip) at the core of the German training system, which values broad training in occupational skills (Pilz Reference Pilz2009: 59–60). The British NVQ system instead allowed employers to pursue a “cafeteria approach” to skill formation (Ryan & Unwin Reference Ryan and Unwin2001: 110), selecting only those training modules required for the immediate task at hand.

In addition to setting up labor-market programs, the Conservatives also promoted vocational education in the general schooling system, in the spirit of “new vocationalism” (Skilbeck et al.Reference Skilbeck, Connell, Lowe and Tait1994; cf. Knight Reference Knight1990: 151). Without consulting the MSC, the government announced the Technical and Vocational Education Initiative (TVEI) in November 1982, an attempt to increase funding for vocational education in schools for students between the ages of fourteen and eighteen (Finn Reference Finn1987: 166–7). Vocational skills acquired in a school-based setting were to be certified in the form of General National Vocational Qualifications (GNVQs), which were meant to prepare students to acquire a broader set of vocational skills in an occupational field than were available from the narrower NVQs. Critics interpreted the TVEI as an attempt by the Conservative government to reestablish a quasi-tripartite system: “The independent sector has been strengthened and subsidised through the Assisted Places Scheme, which by 1985 was supporting 20,000 pupils. Within the state sector, an academic elite will be separated from a middle tier doing something like TVEI; and the bottom 40 per cent will be suitably differentiated and offered a curriculum which no doubt will more adequately prepare them for employment, or more realistically for places in the two-year YTS” (Finn Reference Finn1987: 171).

In 1994, the John Major government engaged in yet another attempt to reform the training system by introducing the successor to YTS (which had briefly been renamed Youth Training, YT), known as Modern Apprenticeship (MA). MA signaled a certain change in the government's perception of the value of apprenticeship training. The goal was to increase the supply of intermediate level skills at the third level of the NVQ framework, which was supposed to correspond to the level of German apprenticeship training (Steedman Reference Steedman2011: 1) and promote employer involvement in the financing and provision of training (Gospel & Fuller Reference Gospel and Fuller1998: 5–6). Although MA was deemed to perform better than YTS in terms of completion rates and quality (Gospel & Fuller Reference Gospel and Fuller1998), the number of young people in high-quality apprenticeships remained low compared to Germany: Ryan & Unwin (Reference Ryan and Unwin2001) have estimated that 13.6 percent of a typical age cohort entered apprenticeship in the United Kingdom in 1998, compared to 62.9 percent in Germany. This means that 0.50–0.67 percent of employees were apprentices in the United Kingdom, compared to 4.42 percent in Germany (Ryan & Unwin Reference Ryan and Unwin2001: 101).

Consolidation of the liberal skill regime (after 1997)

Despite the consistently high levels of reform, rebranding, and relaunching of training schemes in the 1990s and beyond, the politico-economic and institutional logic behind them remained essentially unchanged after the critical juncture in the mid-1980s (Finegold & Soskice Reference Finegold and Soskice1988; Gospel & Fuller Reference Gospel and Fuller1998; Keep Reference Keep1999; King Reference King1997; Ryan & Unwin Reference Ryan and Unwin2001: 111). Even the New Labour government did not fundamentally change the basic logic of the system, neither in the general education system nor in training policy (Gleeson & Keep Reference Gleeson and Keep2004: 58; Keep Reference Keep2006: 60). The key assumption underlying New Labour's social investment strategy was the idea that public investment in the supply of skills would boost international competitiveness and at the same time help low-skilled individuals to gain access to education and training, lowering social inequality (Keep & Mayhew Reference Keep and Mayhew2010; Payne & Keep Reference Payne and Keep2011: 1) and redefining the purpose of education in economic terms (Wolf Reference Wolf1998).

As for training policies, the Labour Party had abandoned its commitment to the collective levy-grant scheme shortly before coming into office and instead supported the creation of Individual Learning Accounts (ILAs) (King & Wickham-Jones Reference King and Wickham-Jones1998). Hence, the Labour Party had turned away from collective approaches to skill formation based on cooperation between the organized labor-market interests towards supporting more individualistic and voluntarist instruments such as individual learning subsidies. Although the New Labour government envisaged a somewhat stronger role for unions, it remained committed to “voluntary partnerships,” instead of genuine social partnership (Keep Reference Keep1999: 324), and kept the basic employer-dominated infrastructure of the training market in place. The TECs were replaced with the Learning and Skills Council (LSC) and Local Learning and Skills Councils (LLSCs) in 2001, but employers remained the interest group with the “highest level of representation” in these (Gleeson & Keep Reference Gleeson and Keep2004: 44).

The focus of training policy continued to be on combating youth unemployment. Shortly after coming to office, Tony Blair initiated a series of initiatives intended to integrate more young people into the labor market. Prominent examples of these were the “New Deal for Young People” program and Foundation Apprenticeships below the level of MAs, which in fact represented a recurrence of the labor-market programs of the 1980s (Ryan & Unwin Reference Ryan and Unwin2001: 111). Public subsidies were used to sponsor low-quality and lightly regulated training programs for youths; once again, increasing the quantity of training opportunities was more important than addressing quality concerns (Fuller & Unwin Reference Fuller and Unwin2011: 191). The fundamental structures of the British “training market” (Ryan & Unwin Reference Ryan and Unwin2001: 107) were left in place, since a multitude of public, private, and semi-private training providers had to compete for public subsidies. It should be mentioned, however, that in 2009 the outgoing Labour government passed the Apprenticeships, Skills, Children and Learning Act, which established the National Apprenticeship Service, an entity designed to convince employers to participate in training, and created an entitlement to apprenticeships for young school-leavers (Steedman Reference Steedman2011: 3).

With regard to general schooling policies, the Labour government continued along the lines set out by the Conservatives, fostering more competition and choice in the education marketplace (Lawton Reference Lawton2005: 134). Political parties are generally expected to differ in their motivations for privatization and decentralization: whereas parties of the right are more inclined to promote market-based reforms for ideological reasons, the left may support decentralization and privatization of education in order to maintain middle-class support and legitimacy for the continued provision of social services in the public or quasi-public domain (Gingrich Reference Gingrich2011: 38, 136). Whatever the reason, New Labour largely accepted the principles of market-based reforms, renaming the grant-maintained schools created by the Conservatives in the 1980s “foundation schools.” The new administration expanded the powers of central government to enforce the national curriculum and school inspections (Bache Reference Bache2003: 303, 305), making the role of LEAs even more circumscribed. It also established “academies,” a new type of school that was still publicly funded but could attract additional private sponsors and was free from the governance and regulation of LEAs. Academies were initially set up in poor neighborhoods to provide educational opportunities for disadvantaged children (Gingrich Reference Gingrich2011: 142). Once the coalition of Conservatives and Liberal Democrats came to power, however, a new Academies Act was passed in 2010 that significantly expanded the number of academies in order to promote the more general process of decentralization and privatization (Taylor-Gooby & Stoker Reference Taylor-Gooby and Stoker2011: 9, 11), even though public opinion, and that of teachers’ unions in particular, was generally critical of these reforms (Hatcher Reference Hatcher2011).

Significant changes also occurred in the financing of higher education, but these were entirely path-dependent; that is, they reinforced the character of the English system as a liberal skill regime. In 1998, the Blair government decided to allow universities to charge tuition fees of up to £1000 per year (in 2004, this was increased to £3000). This increase in the private share of financing was accompanied by an expansion of subsidies (grants and loans) to low-income families (Harrison Reference Harrison2011: 452), softening the blow. Most importantly, the Blair government adopted the goal of increasing the enrollment rate in higher education to 50 percent of an age cohort (Ansell Reference Ansell2010: 200–1). This indicates that VET as an educational alternative was officially written off by a Labour government that was instead eager to expand access to academic higher education for children with disadvantaged backgrounds. This dynamic continued even more forcefully after the change in government; the coalition government of Conservatives and Liberal Democrats passed a motion in 2010 allowing universities to charge tuition fees of up to £9000 per year. This brought the share of private spending on education in the United Kingdom to the same level as found in the United States (OECD 2012: 257).

The coalition government tried to resuscitate apprenticeship training. Appealing to the “beauty of craft skills” (Fuller & Unwin Reference Fuller and Unwin2011: 196), the new education minister Michael Gove set out to secure a central place in the government's agenda for the promotion of apprenticeship training (Payne & Keep Reference Payne and Keep2011: 10). As has been argued by Fuller & Unwin Reference Fuller and Unwin2011: 200), “apprenticeship training as a pathway for young people is astonishingly under-utilised in the UK despite the policy rhetoric and investment afforded to it over the past few years.” In 2008/09, about 5 percent of 16- to 18-year-olds were in apprenticeships (and most of them in less ambitious level-2 qualifications; cf. Fuller & Unwin Reference Fuller and Unwin2011: 200; Payne & Keep Reference Payne and Keep2011: 13), compared to close to 50 percent in countries such as Denmark, Germany, and Switzerland. In line with Conservative policies of the 1980s, the new coalition government has continued the “employer-led” approach to training (Fuller & Unwin Reference Fuller and Unwin2011: 202), meaning that there is no strong push to organize a cross-class compromise on training that would include a strong voice for unions. The new government allocated public funds to promote apprenticeship training, but according to Payne & Keep (Reference Payne and Keep2011: 10), the sums are “relatively small beer” compared to the investments in YTS in the 1980s. Because of this, the government has appealed to employers “mainly through exhortation and voluntary pledges” (Payne & Keep Reference Payne and Keep2011: 27) to promote investment in skills, maintaining the British voluntarist tradition in training policies. The much-debated Richard Review of Apprenticeships has adopted a similar tactic, by arguing that “[t]he solution lies in shifting the power over designing and developing apprenticeship qualifications to employers” (Richard Reference Richard2012: 7). The 2011 Education Act by the coalition government rescinded the entitlement of young people to an apprenticeship or other qualification that had been introduced in the 2009 Act (Steedman Reference Steedman2011: 3).

All of this is further evidence that after the critical juncture of the mid-1980s, the British system became firmly committed to the path of a liberal skill regime, largely because “the vast bulk of employers have to date been unwilling to support the development of a vibrant, high-quality workbased route for new entrants to the workforce” (Payne & Keep Reference Payne and Keep2011: 13; see also Steedman Reference Steedman2011). The distinguishing feature of this regime is a market-based approach to the provision of education and training, allowing individual employers a large degree of flexibility to define the content and standards of training. The system also promotes competition for public funding among a variety of training providers (Steedman Reference Steedman2011: 4). In contrast to Germany, where firms have integrated training into their regular work processes, the large majority of British training providers are companies or colleges specialized in supplying out-of-firm training. The role of unions in the governing bodies of the British training system is marginal. One consequence of this is that NVQs are narrow in scope, since each employer or group of employers can have its individual training programs certified as an NVQ (Keep Reference Keep1999: 334–5). Even though the dominant policy approach of the last thirty years in the United Kingdom has been to promote employer leadership in training policy, employers there are confined to “a subordinate, supportive role as recipients of public subsidy and delivery agents for an ever-growing list of schemes the government has designed for them with little or no real employer input” (Keep Reference Keep2006: 56). British employers take on less responsibility for the training of young people than do German employers, but they continue to have influence over government policies and the design of vocational qualifications – a regime that Gleeson & Keep (Reference Gleeson and Keep2004: 50) call “voice without responsibility.”

The construction of the comprehensive education system (1945–1975)

Immediately after the Second World War, the Swedish education system was still as elitist and segregated as most other European education systems at the time: “Most children only received a six- or seven-year education, which provided them with basic skills in counting, reading and writing. Secondary and higher education were reserved for a minority, which typically came from the wealthier classes and aimed at forming an elite” (Hudson & Lidström Reference Hudson, Lidström, Hudson and Lidström2002: 34; see also Hickox & Lyon Reference Hickox and Lyon1998: 32). Only 5 percent of a typical age cohort passed the Studentexamen required for university admission (Härnqvist Reference Härnqvist, Ball and Larsson1989: 18). Over the ensuing decades, the old elitist system was transformed into a comprehensive education system with the goal of eliminating educational inequalities and opening up access to higher levels of education. This “Nordic model of education” was considered an integral component of the “Nordic welfare state” (Antikainen Reference Antikainen2006: 240) and was very much tied to the political dominance of social democracy (Antikainen Reference Antikainen2006: 240, cf. also Aasen Reference Aasen and Apple2003; Wiborg Reference Wiborg2010).

Even though the balance of power favored the left, the educational reforms of the postwar period were based on a broad consensus between the major political stakeholders. Crucial coalitions had been formed in the 1930s in both the political and the labor-market arenas, following a long period of industrial conflict: in 1933, the Swedish Social Democrats (SAP) and the Agrarian Party formed a “red–green” coalition, which became a crucial driving force behind the universal welfare state model (Manow & Van Kersbergen Reference Manow, Van Kersbergen, Van Kersbergen and Manow2009: 27; Wiborg Reference Wiborg2010). In 1938, the leading organizations of trade unions (Landsorganisationen i Sverige, LO) and employers’ associations (Svenska Arbetsgivareföreningen, SAF) passed the historic Saltsjöbaden or Basic Agreement, which led to the institutionalization of class compromise and is considered to be a historical turning point in labor-market relations (Ball & Larsson Reference Ball, Larsson, Ball and Larsson1989: 3–4). These far-reaching agreements between the major political stakeholders set the tone for the predominant style of policy-making in the postwar period – one that was consensus-oriented, and in fact “positivist-rationalist”: “Parliamentary commissions were established to undertake a careful collection of views, evidence and research, to initiate investigations and experiments and canvass opinions. Commission reports were then subject to widespread consultation with interested groups” (Ball & Larsson Reference Ball, Larsson, Ball and Larsson1989: 10). The consensual and rationalist character of policy-making was accompanied by the emergence of a broad cross-class coalition in support of reform, under the ideological and political leadership of the Social Democrats (Aasen Reference Aasen and Apple2003: 110; Hickox & Lyon Reference Hickox and Lyon1998: 27). The coalition included the major labor-market actors (LO, SAF, and the white-collar union Tjänstemännens Centralorganisation, TCO) and most of the major political parties (Ball & Larsson Reference Ball, Larsson, Ball and Larsson1989: 7).

When the Social Democrats came to power in the 1930s, educational reform was not particularly high on their agenda; other economic and social problems were considered to be more pressing (Paulston Reference Paulston1968: 10, 60). Members of the working class did not care much about secondary schools at the time, since these were regarded as the domain of the upper and middle classes (Paulston Reference Paulston1968: 62). After the Second World War, however, education rose to the top of the reformers’ agenda. The process of rational policy-making had already started during the war, with the establishment of the Expert School Committee in 1940, whose majority recommendation was to maintain the segregated and differentiated school system (Paulston Reference Paulston1968: 95). In the ensuing political debate, the SAP came out as the strongest supporter of comprehensive schooling; the Liberals were less supportive and the Conservatives were openly opposed (Paulston Reference Paulston1968: 95). The Agrarian Party had sided with the Conservatives in this question before 1932, but now changed its position to become the SAP's main ally: “Because of the plight of rural education at the war's end and its many fundamental problems caused by the very nature of a dual system of education, the farmers in general warmly concurred with the sentiment for longer common schooling but, as usual, balked at the thought of increased costs” (Paulston Reference Paulston1968: 96).

Another committee was established in 1946: the Political School Commission. In contrast to the Expert School Committee, the membership of this group came from the major political parties (Paulston Reference Paulston1968: 109) and was thus well placed to come up with a viable political solution to the school reform debate. The Commission submitted its report in 1948, recommending that compulsory schooling be extended to nine years and that comprehensive schools be introduced at the secondary level (Paulston Reference Paulston1968: 113). This was the starting point for the decades-long process of constructing the comprehensive education system, setting in motion a period of cautious and local experimentation with the new comprehensive schools (Härnqvist Reference Härnqvist, Ball and Larsson1989: 20). Following these experiments, legislation in 1962 introduced the nine-year comprehensive basic school (grundskola) as the standard. By this time, comprehensive schools had already spread significantly, to cover about a third of the compulsory school population (Härnqvist Reference Härnqvist, Ball and Larsson1989: 20). There was isolated but sometimes vocal opposition against the introduction of comprehensives from grammar-school teachers and the Conservative Party in the 1950s, but it remained ineffective (Husén Reference Husén1965: 182, 186–7; Lundahl Reference Lundahl1990: 163; Paulston Reference Paulston1968: 128; Rothstein 1996: 160). The move towards comprehensivization was supported by a number of popular movements, such as the Workers’ Educational Association (ABF) and trade unions. Their goal was to eliminate educational inequalities both between economic classes and between urban and rural areas (Hudson & Lidström Reference Hudson, Lidström, Hudson and Lidström2002: 35). The popular support for the comprehensive education system – also expressed in surveys of public opinion (Husén Reference Husén1965: 187) – was so widespread that it is now regarded as a “self-evident consequence of society's democratization and the development of the welfare society” (Hudson & Lidström Reference Hudson, Lidström, Hudson and Lidström2002: 35).

The introduction of the nine-year grundskola was only the first step towards the full comprehensivization of the system, however. The next was to overcome the traditional tracking of students into different types of school at the upper-secondary level. Until the reform of upper-secondary education in 1969, the traditional grammar school (Gymnasium) existed alongside a number of technical, commercial, and trade schools with a more vocational orientation (Härnqvist Reference Härnqvist, Ball and Larsson1989: 25). The 1969 reform integrated the various parts of upper-secondary education into one comprehensive school, the Gymnasieskola (Härnqvist Reference Härnqvist, Ball and Larsson1989: 26). The obvious purpose of this reform was to reduce the distinction between academic and vocational education in order to promote educational mobility. But the institutional legacy of the old system can be recognized in the structure of the new one, in which three- to four-year academic programs, the “vestiges of the traditional university entrance studies formerly provided by the gymnasium” (Opper Reference Opper1989: 140), were run in parallel with a number of two-year academic and vocational programs. The upper-secondary education system conferred about 80 to 100 final qualifications in different occupational fields (Busemeyer Reference Busemeyer2009a: 392; Opper Reference Opper1989: 140–1), signaling a relatively high degree of vocational specificity. One crucial difference from the German system was that all upper-secondary programs in Sweden, in principle, allowed students to continue studying at the tertiary level (Opper Reference Opper1989: 141), a clear expression of the social democratic goal of promoting educational mobility.

The 1969 reform, which came into force in 1972, was a critical point in the history of VET in Sweden. It firmly established the priority of school-based training over other forms. But the path towards school-based VET was by no means the only one the country could have taken, even if apprenticeship training in Sweden may never have had the same relevance as it did in Germany (or the United Kingdom, for that matter). In the critical period of the 1930s, when the foundations of the welfare state were laid, the governing coalition of Social Democrats and farmers also delved into the field of VET policy; the ruling coalition wanted to expand state financial support for apprenticeship training in order to prevent youth unemployment (Nilsson Reference Nilsson1983: 33). Both employers and unions in this era were against increased state involvement in matters of apprenticeship training, albeit for different reasons. Because of their opposition, the state itself set up and financed a number of vocational schools, particularly in areas of the country where unemployment was high (Nilsson Reference Nilsson1983: 33). The government decided to expand the number of regional workshop schools following recommendations from a parliamentary commission in 1937; these “remained the backbone of the vocational school system for 30 years, especially during the 1950s” (Nilsson Reference Nilsson1983: 34), until they were replaced by integrated upper-secondary schools. By the early 1960s, 6,700 students were enrolled in 38 vocational schools (Nilsson Reference Nilsson1983: 34). According to Lundahl (Reference Lundahl1997: 93), the total number of students in state-sponsored vocational schools was actually as high as the number of apprentices in 1945 (about 10,000). As in Germany, employers and unions had successfully lobbied for the establishment of a national board for apprenticeship training and were strongly involved in the administration of the training system between 1944 and 1970, although the system was on a smaller scale than that in Germany (Crouch Reference Crouch1992: 38; Nilsson Reference Nilsson2011; Rothstein Reference Rothstein1996: 103). The number of apprentices declined significantly after 1945, however, so that by twenty years later they totalled about 2,000, while enrollment in vocational schools had by then expanded to 75,000 (Lundahl Reference Lundahl1997: 93). The strong increase in enrollment in vocational schools can be explained by the government's attempt to promote vocational education, a response to the realization that VET had become too important to be left in the hands of employers alone (Nilsson Reference Nilsson1983: 36). State subsidies for full-time courses in vocational schools were significantly increased in 1955, and paved the way for the later integration of VET into secondary education, although dual apprenticeships remained important and popular until the end of the 1960s (Nilsson Reference Nilsson2011). Social Democrats and unions, fearing that apprentices could be exploited as cheap labor by employers, advocated integrating VET into the secondary school system instead (Crouch Reference Crouch1992: 38).

Before the big reform of the late 1960s, both unions and employers were wary of too much state intrusion into the field of vocational training: “SAF and the LO argued that preparatory vocational training in the compulsory school should be the responsibility of the State, whereas substantial parts of VET proper should be planned by and take place within industry, which could provide advanced machinery, modern methods and real working conditions” (Lundahl Reference Lundahl1997: 94). A drop in enrollment in vocational education in the 1960s, however, was “pivotal” in securing the backing of labor-market actors for the 1969 reform (Opper Reference Opper1989: 153). Unions were generally more willing to support the integration of VET into upper-secondary education, because they valued the contribution of the reform towards the goal of promoting educational mobility (Lundahl Reference Lundahl1997: 95; Nilsson Reference Nilsson2011: 28). SAF was “the more reluctant party and argued that VET had become too broad and general at the expense of a focused goal” (Lundahl Reference Lundahl1997: 95). Employers supported the reform because they wanted to make sure that the general education system produced a sufficient number of graduates with a vocational education, even if it was less specific than they would have liked. The dominance of large export-oriented firms and the absence of small and medium-sized enterprises (compared to Denmark) also helped to stabilize school-based VET in Sweden. Large firms were more willing to hire school leavers with broad occupational skills, which could be complemented with on-the-job training in firm-specific contexts (Crouch Reference Crouch1992: 38). Compared to a purely statist system (e.g., France), employers retained some influence over the design of vocational curricula in schools (Crouch Reference Crouch1992: 39).

The 1969 reform was considered a success, “both as a combination of adjustment to the second phase of industrialisation and as a response to social demand for greater equality between different types of vocational education and between the vocational school system and the corresponding system for academic education” (Nilsson Reference Nilsson1983: 38). Enrollment in upper-secondary education expanded significantly between 1971 and 1981, from 75,000 to 106,000. The vocational tracks proved a very popular choice with students: enrollment increased from 24,000 to 49,000, compared to an increase from 30,000 to 41,000 in three- to four-year academic programs (Nilsson Reference Nilsson1983: 40).

Consolidation and conflict (1975–1990)

Bureaucratic obstacles hampered the implementation of the comprehensive school model for a long time. Conflicts and administrative dualism in the National School Board between proponents of the grammar schools and those of the elementary schools continued to slow the realization of the comprehensive model, for example by preventing the formulation of common teacher training standards (Rothstein Reference Rothstein1996: 97, 157). At the same time, the introduction of comprehensive education moved from upper-secondary to tertiary under the social democratic government of Olof Palme. As was the case in VET, higher education was increasingly regarded as the building block of a nationally unified, centrally administered, and rationally planned political economy and welfare state (Elzinga Reference Elzinga, Rothblatt and Wittrock1993: 197). The Social Democrats lost power to a bourgeois coalition that governed from 1976 to 1982, but many of the reforms begun in the early 1970s were taken up again later on, in part because the nonsocialist coalition did not have strong parliamentary support. The university reform of 1977, in particular, eliminated distinctions between academic and vocational tertiary education by abolishing the binary system and forming a new unified Högskola, as well as by upgrading the professional status of welfare professions (Elzinga Reference Elzinga, Rothblatt and Wittrock1993: 217). The 1977 reform was also an attempt to promote applied instead of basic research, so as to maximize the contribution of higher education to the development of the political economy (Elzinga Reference Elzinga, Rothblatt and Wittrock1993: 207–8). In spite of these structural reforms, however, participation in higher education did not expand significantly during the 1980s (Jonsson & Erikson Reference Jonsson, Erikson, Shavit, Arum, Gamoran and Menahem2007: 116).

Regarding the general schooling system, the nonsocialist government did not transform “education in a neo-liberal direction to any large extent” (Lundahl Reference Lundahl2002: 690), but it did start a process of decentralization from the central to the local level that would continue on a small scale throughout the 1980s. Examples of decentralization included granting more responsibility to local authorities in matters of financing, transforming governance modes from hierarchical steering to goal- or output-oriented governance, and increasing choice options for parents (Lindblad & Wallin Reference Lindblad and Wallin1993: 79). This early wave of decentralization was a reaction to popular criticism of the overly bureaucratic and centralized system of the 1970s, which, amplified by the Conservative party, focused on topics such as the lack of discipline and academic standards in comprehensive schools (Lundahl Reference Lundahl1990: 156). This is an interesting parallel to the debate in the United Kingdom triggered by the Black Papers. At the same time, the Swedish Conservatives display similarities to Christian democratic ideology on matters of education policy. For instance, they supported early streaming of pupils onto different tracks, a stronger separation between academic and vocational education at the secondary level, and a closer connection between VET and the labor market (Lundahl Reference Lundahl1990: 160–1). Like British Conservatives, however, Swedish Conservatives lack a strong commitment to the corporatist “politics of mediation” (Van Kersbergen Reference Van Kersbergen, Kitschelt, Lange, Marks and Stephens1999: 356).

In contrast to the radical decentralization reforms of the British Conservative government in the 1980s, which emphasized choice and competition, the driving force behind the early Swedish efforts at decentralization was an attempt to increase the participation and influence of local stakeholders, such as local policy-makers, teachers, and parents, in the governance of education. When the Social Democrats regained power in 1982, the government engaged in a process of careful decentralization in reaction to mounting evidence that citizens felt a lack of influence in public schooling (Björklund et al.Reference Björklund, Clark, Edin, Fredriksson and Krueger2005: 2; Bunar Reference Bunar2010a: 6; Klitgaard Reference Klitgaard2008: 490). Significant political momentum for the decentralization of education policy came from the political left demanding more local flexibility in the implementation of new pedagogical concepts (Aasen Reference Aasen and Apple2003: 114). As a consequence, the government delegated responsibilities for primary schools to municipalities and increased local stakeholders’ ability to influence the content of teaching. A significant reform occurred in 1989, when municipalities were put in charge of personnel matters and a new funding scheme was introduced: in the new regime, the central government would give subsidies in the form of lump-sum payments to the municipal governments, greatly increasing the autonomy of local decision-makers (Björklund et al.Reference Björklund, Clark, Edin, Fredriksson and Krueger2005: 2; Hudson & Lidström Reference Hudson, Lidström, Hudson and Lidström2002: 45; Klitgaard Reference Klitgaard2008: 490; Lundahl Reference Lundahl2002: 691). The extent of decentralization was limited overall, in part because of the intense conflict within the Social Democrat party between the reformers who supported decentralization and the “true believers” who preferred the old, centralized model (Klitgaard Reference Klitgaard2008: 489). The result was a “cooperative model” of decentralization (Mons Reference Mons2004), in which the local level gained more autonomy but the central state remained committed to, and strongly involved in, certain matters, such as national-curriculum standards and school inspections (Bunar Reference Bunar2010a: 7; Hudson & Lidström Reference Hudson, Lidström, Hudson and Lidström2002: 46).

VET in Sweden went through a period of consolidation and simmering conflict in the 1980s. In 1976, shortly before leaving office, and confronted with the rising problem of youth unemployment (as in the United Kingdom and Germany), the social democratic government had set up a parliamentary commission to reform upper-secondary education. The goal of this reform was to further erode the differentiation between the academic and vocational tracks in the newly integrated upper-secondary school system. The incoming nonsocialist government weakened this aspect of the reform, but the parliamentary commission's report in 1981 effectively remained committed to the goal, proposing a common structure for educational programs so that they would all contain both academic and practical components (Lundahl Reference Lundahl1997: 97). Not surprisingly, the SAF strongly opposed this reform, especially the proposed extension of the role of basic vocational education, because employers wanted VET to more closely reflect the needs of industry (Crouch Reference Crouch1992: 39–40). The unions opposed the reform, too; in their view, the proposal did not go far enough towards abolishing the differences between academic and vocational education (Lundahl Reference Lundahl1997: 97). Such widespread opposition to the reform led the government to abandon it. This example shows that in CMEs, such as Sweden, government attempts to promote policy change against the joint opposition of employers and unions will likely fail.

For the remainder of the 1980s, the institutional structure of upper-secondary education followed the stipulations of the 1972 reform, but the political debate on reforming the system continued. Employers and unions began criticizing the system's strong orientation towards school-based forms of VET and its lack of linkages to the labor market. They also feared that vocational education would be neglected, as an increasing number of young people opted for academic education (Lundahl Reference Lundahl, Lundahl and Sander1998: 43). The next push for reform was a joint effort by the social partners: they set up yet another commission in 1984 (working via the Joint Industrial Training Council set up in 1944), which in 1986 recommended extending the length of VET from two to three years and expanding workplace-based components of training (Lundahl Reference Lundahl1997: 98). The social democratic government implemented some of these proposals, but on a small scale (Lundahl Reference Lundahl, Lundahl and Sander1998: 43). For example, a new agreement in 1984 between the government, educational authorities, and social partners introduced apprenticeships as a potential VET route in upper-secondary education. The number of apprentices in 1984 and thereafter remained very low, however, amounting to about 10 percent of total enrollment in vocational tracks at upper-secondary level (Opper Reference Opper1989: 147); that is, roughly 4 percent of a typical age cohort at the time.

The Conservative party was a strong supporter of apprenticeships and wanted apprenticeships to become the dominant form of VET at the upper-secondary level (Lundahl Reference Lundahl1997: 98). The employers were less enthusiastic, wondering whether firms could provide the kind of broad occupational skills that would be required (Lundahl Reference Lundahl1997: 98). Unions, in turn, wanted to keep apprenticeship training in a marginal position, because they feared education would become subordinated to the needs of employers (Lundahl Reference Lundahl, Lundahl and Sander1998: 47). In sum, despite increased support among the important stakeholders for increasing workplace-based components of VET in principle, the system essentially remained stuck on its development path of a statist skill regime.

Decentralization and reform (after 1991)

The bourgeois government coalition that was in charge from 1991 until 1994 followed a more market-oriented ideology and accelerated the decentralization of the governance of education (Aasen Reference Aasen and Apple2003: 129). The nonsocialist government attempted to increase the role of private, independent schools in particular, which traditionally had a very marginal position (accounting for 0.2 percent of enrollment in the early 1980s; cf. Klitgaard Reference Klitgaard2008: 489–90). The Social Democrats, despite strong internal conflicts, had already taken the first steps in this direction in the late 1980s: the new financing mechanism mentioned above distributed public funds to both public and private schools, according to local needs (Klitgaard Reference Klitgaard2008: 490–1). The nonsocialist government, however, went beyond that by giving private schools the legal right to receive public funds amounting to 85 percent of the average costs of a public school student and allowing schools to charge tuition fees for the remaining 15 percent (Klitgaard Reference Klitgaard2008: 492). This new regime was effectively a voucher scheme for independent schools because it was complemented by the introduction of school choice, meaning that parents could now freely choose among public (and independent) schools within their municipality, although students were sill allocated to schools on the basis of residence (Björklund et al.Reference Björklund, Clark, Edin, Fredriksson and Krueger2005: 6). Other policy changes included transferring employer responsibility for teachers (eventually including wages) from the central to the municipal level, introducing a new curriculum, and establishing more flexibility to adapt the curriculum to local needs (Arreman & Holm Reference Arreman and Holm2011: 227). Interestingly, the social democratic government after 1994 by and large did not reverse these policy reforms, although it did enact some form of recentralization in terms of regulation and oversight (Gingrich Reference Gingrich2011: 154). First, it lowered the level of subsidies to private schools to 75 percent of average per-student costs in public schools. Then, it increased subsidies to 100 percent of the average, but prohibited private schools from levying additional fees (Klitgaard Reference Klitgaard2008: 492). This policy was passed against the wishes of organized interests (especially teacher unions) and can best be understood as an attempt by reformist Social Democrats to cater to the interests of the middle class, which demanded more opportunities for choice in education (Klitgaard Reference Klitgaard2008: 492).

Can the decentralization and privatization reforms of the 1990s be considered as path-changing reforms? First of all, empirical research on the reforms indicates that their effects have been rather small. “Surprisingly or not, just 17 years after the reform was enacted, it seems that it has not managed to bring decisive changes (either positive or negative) into the educational system. […M]ost researchers and evaluators still claim that the outcomes in terms of segregation, costs, and achievement at the national level are ambiguous or at best visible but small” (Bunar Reference Bunar2010a: 13, see Björklund et al.Reference Björklund, Clark, Edin, Fredriksson and Krueger2005 for a similar conclusion). Even so, the share of post-16-year-old students in independent schools has increased to 20 percent (as of 2008), and the share of upper-secondary schools that are independent has increased to 40 percent of total (2007/08; cf. Arreman & Holms Reference Arreman and Holm2011: 235). Many of these new independent schools have a vocational orientation. While they may be run by commercial companies, they are still entirely publicly financed and are not allowed to charge tuition fees (Arreman & Holms Reference Arreman and Holm2011: 226; Bunar Reference Bunar2010b: 49). Furthermore, “free” schools have to comply with the national curriculum and are subject to public inspections and regulations (Bunar Reference Bunar2010b: 49, 53–4). Most importantly, free schools can decide whether to accept students from other districts, but “no one's application can be turned down based on his or her ability, level of achievement, gender, ethnicity, religion, or social background” (Bunar Reference Bunar2010b: 53). The prohibition on selecting students based on academic ability prevents private schools from becoming overly selective organizations catering to the academically gifted. In sum, as Bunar (Reference Bunar2010b) argues, the Swedish school system has become more competitive and decentralized, but competition plays out in a highly “controlled school market” (Bunar Reference Bunar2010b: 49), pervaded by an “ideological legacy” (Bunar Reference Bunar2010b: 56) that emphasizes equity and equality in education. In a similar manner, Aasen (Reference Aasen and Apple2003: 145) argues that the reforms of the 1990s can be interpreted as a “renewal of social-democratic progressive education to meet new challenges” and “an effort to defend the Scandinavian social democratic political model.”

With regard to VET, the outgoing social democratic government passed a large reform of upper-secondary education in 1991. This was further spelled out in the 1994 reform of the national curriculum, which took up a number of proposals developed by the social partners in the reform debate of the 1980s. More specifically, the new system consolidated the number of courses or tracks at upper-secondary level into seventeen national programs, thirteen of which had a vocational orientation. All of them conferred a basic eligibility for studying at a higher education institution and all had a common core of general subjects. The length of VET courses was also increased, to three years (cf. Arreman & Holm Reference Arreman and Holm2011: 228; Lundahl Reference Lundahl1997: 98–9; Lundahl et al.Reference Lundahl, Erixon Arreman, Lundström and Rönnberg2010: 49–50). The reform further blurred the line between academic and vocational education. It also greatly expanded access to higher education: nowadays, almost all students from lower-secondary education proceed to secondary education (Lundahl Reference Lundahl2002: 692). The Conservative government fueled the expansion of higher education with the passage of the Higher Education Act in 1993, which also increased the autonomy of individual universities with regard to admissions (Ansell Reference Ansell2010: 203).

There is also evidence of a “new vocationalism” at work: the reform was intended to increase the share of instruction in the workplace to 15 percent of scheduled time in vocational courses (Lundahl Reference Lundahl1997: 98–9). The coalitional dynamic behind the 1991 reform is interesting: The reform was largely supported by LO and SAF, because its purpose was to increase the relevance and attractiveness of VET (Lundahl Reference Lundahl1997: 100). The Conservative and Liberal parties were opposed, however, because the reform further blurred the distinction between academic and vocational education, as mentioned above (Lundahl Reference Lundahl, Lundahl and Sander1998: 49; Reference Lundahl, Erixon Arreman, Lundström and Rönnberg2010: 50). More recent reforms by the current bourgeois government coalition have gone in the direction of reinstating the distinction between academic and vocational education (Lundahl et al.Reference Lundahl, Erixon Arreman, Lundström and Rönnberg2010: 56).

The Swedish case shows that the transformation of education systems from the elitist and class-based models of the past to the comprehensive and universalist social democratic variant has taken a long time, despite the long-term dominance of Social Democrats and their allies in government. The early 1970s can be regarded as a critical juncture, in which VET was fully integrated into the secondary school system, setting Sweden on the path towards a statist model of skill formation. Significant reforms happened after this period as well, in particular the decentralization reforms of the 1980s and 1990s, but the fundamental institutional structure of the Swedish system has remained largely stable since the integration of VET.

Reconstruction and restoration (1945–1965)

Without a doubt, the end of the Second World War marked a significant watershed in the political and historical development of Germany. The large-scale destruction of cities and property, millions of refugees from the East, the redesign of political institutions, and the denazification of the public sector were immediate and pressing problems. In these times of extreme uncertainty, policy-makers often turned to established institutions and policies of the prewar era, or, in some cases, left the existing institutions in place. Instead of a large-scale break with the past, there was a great deal of institutional continuity, particularly in the field of VET (Greinert Reference Greinert1998; Thelen Reference Thelen, Mahoney and Rueschemeyer2003, Reference Thelen2004).

The policy legacies that policy-makers were confronted with go back to the roots of the German education and training system in the nineteenth and early twentieth centuries. Prussia was among the first countries to enact compulsory schooling, in 1763, although this regulation was difficult to implement and enforce at the time (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 52). Access to prestigious positions in the Prussian bureaucracy was tied to school certificates, in particular the Abitur – the school-leaving certificate of the academic track of secondary education, the Gymnasium (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 34; Thränhardt Reference Thränhardt, Von Beyme and Schmidt1990: 178). As a consequence, access to higher levels of education became an important factor in class politics. One element in Bismarck's attempts to fight off the rising power of the working and middle classes was the limiting of access to universities and Gymnasien through increased tuition fees (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 76). These efforts were complemented by efforts to channel the educational aspirations of the rising classes into new alternative tracks. A new middle school below the level of Gymnasien but above the common school (Volksschule) was established (the predecessor of the modern-day Realschule), preparing the children of the bourgeoisie for intermediate-level clerk positions in the public and private sectors (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 76, 132). In 1897, the government passed the “Handwerkerschutzgesetz,” which is widely regarded as an important cornerstone of the apprenticeship system because it protected the privileged position of craft chambers to regulate, monitor, and administer apprenticeship training in firms in the crafts sector; this was entrenched by further legislation in 1908 (Greinert Reference Greinert, Arnold and Lipsmeier2006: 500; Thelen Reference Thelen2004: 43). As Thelen (Reference Thelen2004: 44–53) has pointed out, this was aimed at convincing the craft sector (the petite bourgeoisie) to align themselves with the authoritarian regime instead of the rising labor movement. Because of this legislation, German unions could not effectively use apprenticeship to control access to skilled labor markets, as British unions had. The unintended consequence of this constellation was that it triggered the rise of industrial unions, instead of craft unions, which in later periods facilitated the formation of cross-class consensus in matters of skill formation.

The Weimar period was significant in two respects. First, the enactment of the 1920 school law (Reichsgrundschulgesetz) symbolized a preliminary school compromise between Conservatives and the supporters of educational reforms promoting a more comprehensive education system (Mitter Reference Mitter1991: 156). It introduced a four-year basic school for all children, whereas before the Gymnasien and other higher schools had their own basic schools. This progressive element was counterbalanced by the more conservative three-tiered structure of the education system at post-primary level, consisting of the Volksschule, the Realschule, and the Gymnasium.

The second legacy of the Weimar period was the establishment of a collective and essentially voluntarist approach to VET, based on coordination between firms in the crafts and industry sector (cf. Greinert Reference Greinert1998: 62–86; Thelen Reference Thelen2004: 53–89). Because previous legislation had essentially given the crafts sector a monopoly over apprenticeship training, industry interests now aimed to expand their own control over the content of training and the certification of vocational skills. The Deutscher Ausschuß für Technisches Schulwesen (DATSCH), founded in 1908 by the associations of the engineering and machine-making industries, developed training curricula and exams and thus contributed to standardizing training content across the economy. In 1925, the DATSCH teamed up with the peak employers’ associations to establish the Arbeitsausschuß für Berufsbildung (Working Committee for Vocational Education). Once trade unions had been officially recognized as partners in collective wage bargaining, they too tried to expand their influence over vocational training. Employers were initially in favor of cross-class compromise, supporting the establishment of the Zentralarbeitsgemeinschaft der industriellen und gewerblichen Arbeitgeber und Arbeitnehmer Deutschlands in 1921, a committee of employer and trade union representatives concerned with enacting comprehensive legislation. This effort at cross-class coordination fell apart at the end of the 1920s, however, when the industry and crafts sectors joined forces to fend off state legislation of this domain of employer-led self-governance (Greinert Reference Greinert1998: 86).

Both of these legacies became important during the immediate postwar period. As Ellwein puts it: “When the educational establishments were reopened in the postwar period, immediate concerns and needs were foremost; getting things off the ground was the order of the day. There was little time for reflection on the problems of established institutions and future needs. People turned to existing institutions and continued along these lines” (Ellwein Reference Ellwein, Führ and Furck1998: 89, author's translation). One new actor in the education policy arena was, of course, the American military government. The US occupying power promoted educational reform in the name of democratic reeducation, because it was believed that Germany's elitist education system had contributed to the rise of the Nazis (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 160, 163). The United States was successful in promoting a new set of educational institutions in Japan that were modeled on the US system of comprehensive secondary education (Heidenheimer Reference Heidenheimer1997: 55). In Germany, however, these attempts mostly failed. A progressive, reform-oriented coalition of Social Democrats, trade unions, associations of teachers from the Volksschulen, and the US military government was pitted against a conservative coalition of Christian Democrats, universities, teachers from the Gymnasien, and churches (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 164; Thränhardt Reference Thränhardt, Von Beyme and Schmidt1990: 188). The conservative coalition gained considerable legitimacy from the support of the Catholic Church, which was less tainted by the Nazi legacy and which derided the proposed educational reforms as “unchristian” (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 164).

Some Länder governments made use of the head start gained by virtue of their constitutions having been enacted before the Federal Basic Law was passed in 1949. In particular, Länder governed by the SPD, such as Berlin, Hamburg, Bremen, and Schleswig-Holstein, passed a series of educational reforms in the years following 1945. By 1953, most of these reform attempts had been rolled back, and the three-tiered structure of the secondary school system was fully reinstated across all of West Germany (Thränhardt Reference Thränhardt, Von Beyme and Schmidt1990: 189). One reason for the rollbacks was the losses suffered by the SPD in Länder elections of the early 1950s, perceived as being a consequence of the school reforms (Schmidt Reference Schmidt1980: 93; Wiborg Reference Wiborg2010: 552–3). A more indirect, but probably more lasting reason for the unpopularity of the comprehensive school model is that it was introduced in East Germany as the “fundamental school type of the unified socialist educational system” (Mitter Reference Mitter1991: 157). The exploitation of the comprehensive school by the socialist regime in the East contributed to its delegitimization as a reform model in the West. A further factor may have been “national policy discourses” (Baldi Reference Baldi2012: 1000); that is, the dominance of conservative thinkers in educational research who emphasized the validity of the “fixed-ability approach” tied to the segmented school system.

The peculiar character of German educational federalism is evidenced by the fact that the central coordinating organization was actually established before the founding of the Federal Republic in 1949. The Ständige Konferenz der Kultusminister der Länder in der Bundesrepublik Deutschland (Kultusministerkonferenz, or KMK, for short) was established in 1948 as a voluntary association of the education ministers of the German Länder. Because the KMK is a forum for voluntary horizontal coordination between the Länder, decisions must be unanimous, which often results in gridlock and lowest-common-denominator policies (Thränhardt Reference Thränhardt, Von Beyme and Schmidt1990: 182–3). Nevertheless, the KMK has issued hundreds of decisions and recommendations over the decades, which although formally voluntary are regularly implemented by the Länder. The KMK is thus crucial in creating and sustaining a unitary framework in the federalist education system. The agreements of Düsseldorf (1955) and Hamburg (1964), for example, defined the core features of the German three-tiered education system and remain in force today.

Once the central features and institutions of the education system had been restored and consolidated, no further large-scale structural reform attempts were made for the rest of the 1950s or much of the 1960s. The Christlich Demokratische Union Deutschlands (CDU) staunchly defended the three-tiered structure of secondary education and justified it with religious references to the divine order of an organic society, based on the principle of differences between individuals (Faulstich Reference Faulstich1977: 94), which contrasted sharply with the secular and egalitarian comprehensive school model promoted by left-wing partisans. The 1950s also saw conflicts over the survival of confessional (religious) schools, which cumulated in a popular referendum in Bavaria in which the Christian Democrats (organized into the party of the CSU in Bavaria) faced a coalition of liberals and Social Democrats who advocated ending confessional schools (Thränhardt Reference Thränhardt, Von Beyme and Schmidt1990: 194–5). After this struggle, purely confessional schools were abandoned, but private schools with ties to church authorities and largely financed by the state survived.

The SPD increasingly became a supporter of the comprehensive school model, albeit probably a less enthusiastic one than its Nordic neighbors (Faulstich Reference Faulstich1977: 98–102; Schmidt Reference Schmidt1980: 92). Comparing the effects of partisan government at the Länder level, Schmidt (Reference Schmidt1980: 59, 63) found that SPD participation in government was associated with higher levels of education spending and participation, as well as more concern for primary, lower-secondary, and vocational schools. Education policy was not a particularly important topic for trade unions in the 1950s, when economic and social policy issues were more pressing (Schmidt Reference Schmidt1980: 128, 135).

In 1953, the Deutscher Ausschuß für das Erziehungs- und Bildungswesen (German Committee for the Education System) was established as an advisory committee. Half of its members were educators, the other half academics and business experts. The committee produced a significant number of reports and recommendations, but these did not have any sustainable impact (Hepp Reference Hepp2011: 96). Compared to similar committees and commissions in the United Kingdom and Sweden at the time, its recommendations were rather cautious and conservative. In its Rahmenplan zur Umgestaltung und Vereinheitlichung des allgemeinbildenden Schulwesens (Framework for the Reform and Unification of the General School System), the committee expressed continued support for the three-tiered structure of secondary schooling (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 168). Its proposals were minor: to extend the four-year primary school by two years and to ease transitions between the different tracks of the secondary school system (Herrlitz et al.Reference Herrlitz, Hopf and Titze1998: 169–70).

In VET, the overarching topic was again the restoration of prewar institutions, meaning the quasi-voluntarist and employer-dominated system of apprenticeship training (Thelen Reference Thelen2004: 240–58). Immediately after the end of the war, various Chambers of Industry and Commerce coordinated to reinstate the prewar training curricula and to accept only training in recognized training occupations (Greinert Reference Greinert1998: 68). The successor to the DATSCH, the Arbeitsstelle für Betriebliche Berufsausbildung (ABB) was established in 1953 under the aegis of the Bundesverband der Deutschen Industrie (BDI), one of the two main employers’ associations. The CDU-led government passed two smaller pieces of legislation that supported the system of employer-led self-government: the 1953 Handwerksordnung (Crafts Ordinance) and the 1956 Gesetz zur vorläufigen Regelung des Rechts der Industrie- und Handelskammern (Law on the Preliminary Regulation of Chambers of Industry and Commerce). These pieces of legislation reinstated and reinforced the Chambers’ powers to monitor, implement, and design training, without expanding the participation rights of trade unions in these semi-public and employer-dominated institutions (Lemke Reference Lemke1969: 1305). They also reaffirmed the distinction between training in the crafts and in the industry sectors, which continued to be subject to different regulatory regimes. As Bismarck had done before him, Adenauer saw the continued protection of the privileges of the crafts sector as an effective instrument by which to support small and medium-sized businesses (Mittelstandspolitik; Stratmann Reference Stratmann, Stratmann and Schlösser1990b: 80–1), which were needed to carry out postwar reconstruction and fight surging youth unemployment.

Employers’ associations were closely aligned with the Christian democratic government. Deliberately taking up a popular campaign slogan of the CDU in the 1957 campaign (“Keine Experimente!” – “No experiments!”), the German Association of the Industry and Commerce Chambers (Deutscher Industrie- und Handelstag, DIHT) staunchly defended the principle of self-governance of employer associations in matters of training policy, pointing to the high level of training standards (DIHT 1958: 1). The ideology of German employers’ associations was also deeply influenced by religious concerns, which differentiates them from secular conservative or liberal business associations in other countries. The peak employers’ association published two pamphlets in the early 1950s that highlighted the importance of the religious underpinnings of the German variety of democratic capitalism in the form of the social market economy (Baethge Reference Baethge1970: 61–2). In an echo of Catholic social doctrine, the education of young people was considered to be an important obligation for employers (Baethge Reference Baethge1970: 63).

The trade unions were less satisfied with the quality of training than were employers. They renewed their prewar demand for the establishment of a comprehensive legislative framework of firm-based training. In 1950, there were about 900 recognized training occupations (Ausbildungsberufe), which was taken by trade unions as an indication of the narrow and overly firm-specific nature of apprenticeship training (Stratmann Reference Stratmann, Stratmann and Schlösser1990b: 85). The unions called for broader vocational education with stronger theoretical foundations, more involvement in training decisions in the workplace, and a training levy to address the disparity of investment in skill formation between training and nontraining firms (Greinert Reference Greinert1998: 84).

Towards the end of the 1950s, unions prodded the SPD to become more active in the field of VET. The Deutscher Gewerkschaftsbund (DGB, the German Federation of Trade Unions) presented its own proposal for a law on VET in 1959 (Greinert Reference Greinert1998: 87). Following the initiative of the SPD, the Bundestag unanimously passed a motion (Entschließungsantrag) in 1962 to push the CDU-led government to begin work on a proper law proposal (Stratmann Reference Stratmann, Stratmann and Schlösser1990c: 148). Ludwig Erhard, the new chancellor, stated his opposition to state intervention in the field of VET and his support for the principle of self-government by associations (Stratmann Reference Stratmann, Stratmann and Schlösser1990c: 148–149). The federal government officially responded to the motion in 1964 (one year after the deadline), stating that it was not yet able to present a proposal because of the “extreme complexity of the matter” (cited in Greinert Reference Greinert1998: 87, author's translation). Serious reform of the VET institutional framework would have to wait until the first CDU-SPD grand coalition came into power in 1966.

Expansion and reform (1965–1982)

Like other European countries, Germany went through a phase of massive educational expansion in the 1960s and 1970s. Before the 1960s, higher education policy was fully in the hands of the Länder governments, without the intervention of the federal government. Universities went through a phase of denazification and turmoil immediately after the end of the war, but by and large, their governance model remained that of the prewar years and became a textbook case of academic oligarchy (Clark Reference Clark1983). Like the governance model in VET, the principle of self-government (in this case: of university professors and their faculties) is today still deeply entrenched, although the ministerial bureaucracies of the Länder have intervened quite heavily in the recruitment of professors, as well as in management and financing issues.

The involvement of the federal government was very limited in the beginning, but expanded significantly during the 1960s and 1970s, in response to the massive increase in student numbers. In 1950–55, only 3.4 percent of a typical age cohort was students. This share increased to 5.4 percent in 1960–65 and to 11.7 percent in 1970–75 (figures from Heidenheimer Reference Heidenheimer1997: 236). The Länder governments were unable to deal with this surge alone, because many of them lacked the resources to establish new universities on their own. The grand coalition therefore passed a reform of the Basic Law in 1969, which enumerated new “Joint Tasks” (Gemeinschaftsaufgaben) to the Bund and the Länder, in particular the building of new universities, educational planning, research funding, and educational subsidies, and provided a general competence for the federal government to enact a legal framework for higher education (Hepp Reference Hepp2011: 112–13).

This constitutional reform opened the door to a massive expansion of the federal government's involvement. After the grand coalition collapsed in 1969, a new government coalition between the SPD and the liberal FDP attempted to push forward an agenda of significant reforms in many policy fields and to do away with the conservative legacy of the Adenauer era (Schmidt Reference Schmidt1980). Facilitating educational opportunities for previously excluded segments of the population was a central element of this progressive, social–liberal reform agenda. By holding the purse strings, the federal government could influence the speed and specific guise of the expansion of higher education. In 1971, the SPD-FDP government passed legislation granting educational subsidies to low-income students for the first time (Bundesausbildungsförderungsgesetz, BAFöG). In 1973, 46 percent of students received at least some subsidies for their living expenses under this law (Hepp Reference Hepp2011: 164). In 1976, the government passed framework legislation for the higher education sector (Hochschulrahmengesetz) that upgraded the status of universities of applied sciences (Fachhochschulen) to the level of universities proper; this included detailed regulations that prompted resistance on the part of the Länder and universities.

In line with the prevailing zeitgeist, the government initiated a large-scale venture into educational planning. The Deutsche Bildungsrat (German Education Council) was founded in 1965 as the successor to the ineffective Deutscher Ausschuß. This time, the group consisted of both academics and government representatives, in order to achieve a tighter connection to policy-making. The Bund-Länder Kommission für Bildungsplanung und Forschungsförderung (BLK) was established in 1970, following the constitutional reform of 1969. Its institutional setup was different from that of the KMK: in the BLK, the federal government had the same number of votes as the Länder taken together. Decisions were made by qualified majority rule. This meant that on the one hand, no decision could be made against the will of the federal government, while on the other, if a majority of Länder governments supported the position of the federal government, qualified majority decisions could be passed. Until 1976, the SPD (alone or in coalition with the FDP) ruled in a majority of the Länder, in addition to being in charge of the federal government. The social–liberal coalition was able to use qualified majority decisions to move forward with its reform agenda, even with the opposition of those Länder governments in which the CDU dominated.

There were bitter ideological conflicts in the BLK from the beginning (Hepp Reference Hepp2011: 138). The SPD had by then become a strong supporter of integrated comprehensive schools, whereas the CDU continued to oppose them. The main task of the BLK and the Bildungsrat was to develop a comprehensive educational plan for the whole system (Bildungsgesamtplan), which they managed to do in 1973 despite the bitter conflicts. A central feature of this plan was the integration of VET into the general school system (more on this point below). The Bildungsrat also recommended that forty integrated comprehensive schools be established as part of a pilot program (the social–liberal coalition had a majority on this committee, just as in the BLK), about fifteen to twenty years after the United Kingdom and Sweden had started similar projects. This recommendation was supported by the KMK in 1969, and the BLK was put in charge of monitoring and evaluating these experiments. In the years that followed, the Länder governed by the SPD established comprehensive schools on a much larger scale than the Länder governed by the CDU or the CSU, where comprehensive schools remained isolated experiments. Even in those places where a significant percentage of students attended comprehensive schools, however, the old three-tiered structure, in particular the Gymnasium, was not abolished, effectively resulting in a four-tiered structure (Heidenheimer Reference Heidenheimer1997: 98–100; Hepp Reference Hepp2011: 214; Mitter Reference Mitter1991: 157). The BLK published its report on the comprehensive schools in 1982. By then, the tide had turned: the CDU Länder fully reverted to the three-tiered system and successfully pushed to dismantle the Bildungsrat in 1975 (Heidenheimer Reference Heidenheimer1997: 98–9). The FDP had become a less ardent and reliable supporter of comprehensive schools than it had been in the early 1970s (Schmidt Reference Schmidt1980: 105). The project of comprehensive educational planning was mostly abandoned after the change in the federal government in 1982 (from the social–liberal coalition to a coalition of the Christian Democrats and the FDP) and the change in the majority of the Bundesrat in 1976 (Mäding Reference Mäding and Brinkmann1985).

Unlike the other educational sectors, the field of VET experienced a critical phase of path formation in the 1960s and 1970s. Shortly before the grand coalition came to power at the end of 1966, the SPD formally submitted a proposal for a VET law to the Bundestag. The CDU/CSU/FDP government coalition followed up with its own proposal shortly thereafter. The two different proposals strongly reflected the divergent interests of trade unions and employers’ and crafts associations (Offe Reference Offe1975: 57). The various organizations clashed during the public hearing of the proposals in 1967, and the grand coalition then decided to speed up the process by merging the two proposals and passing the Law on Vocational Education and Training (Berufsbildungsgesetz, BBiG) in 1969 (Greinert Reference Greinert1998: 87).

The reactions to the adoption of the BBiG were mixed. While trade unions had lobbied for the establishment of a comprehensive legal framework for firm-based VET, the law fell short of their expectations (“Restoration instead of reform,” Lemke Reference Lemke1969: 1303, author's translation; see also Faulstich Reference Faulstich1977; Kuda & Mignon Reference Kuda and Mignon1982). The law fundamentally did not question the autonomy of training firms in their decision to participate in apprenticeship training and to influence its content. Unions had also lobbied for the establishment of publicly controlled training funds (financed by a training levy) in order to create a redistribution of funds from nontraining to training firms, none of which became law. The extensiveness of the remit of the BBiG was limited as well, because the crafts sector was allowed to keep its own regulatory scheme for apprenticeship training (Offe Reference Offe1975: 59). Finally, the law only regulated the firm-based part of apprenticeship training, over which the federal government had exclusive law-making powers; the school-based part of training fell into the domain of the Länder. Employer representatives were largely satisfied with the law, but the unions were not (Offe Reference Offe1975: 71). As Thelen (Reference Thelen2004: 262) has argued, the law “brought greater unity to the previously disparate legal framework” but did not fundamentally change the existing system of self-governance (see also Lipsmeier Reference Lipsmeier, Koch and Reuling1992 for a similar assessment).

The impetus for a “reform of the reform” started to grow after the change in the political makeup of the federal government in 1969 (Stratmann Reference Stratmann, Stratmann and Schlösser1990a: 179). That same year, the Bildungsrat published an evaluation of the existing apprenticeship system that was critical of the lack of systematic and theoretical training and cited the danger of apprentices being misused as cheap labor (Deutscher Bildungsrat Reference Pätzold1991). The SPD-FDP coalition in the federal government paid particular attention to promoting VET in its agenda for educational reform, demanding that training no longer be treated as the “stepchild of education policy” (Bundesregierung Reference Pulte and Vorbrücken1974a: 24, author's translation). In its initial ambitious proposals for reform (Bundesregierung Reference Pulte and Vorbrücken1974b), the government wanted to expand public oversight of firm-based training and strengthen the quality and theoretical underpinnings of training curricula by establishing foundational training courses. It also proposed to establish a training levy system and fully integrate VET into the new comprehensive secondary school system. Unions were generally very supportive of these proposals, but, unsurprisingly, employer representatives of all stripes and the Christian Democrats were hostile to them (Busemeyer Reference Busemeyer2009c: 80–8). The government backed down from its initial proposals when employers threatened to boycott apprenticeship training (Baethge Reference Baethge and Lipsmeier1983: 148). Faced with increasing youth unemployment after the oil crisis, changing demographics, and a shift in administration (from Brandt to Schmidt in the chancellorship and Dohnanyi to Rhode in the education ministry), the government passed a much watered-down reform in 1976. This legislation still allowed the government to impose a training levy (which it never did), but only under exceptional circumstances, such as in the case of a severe lack of apprenticeship places. More ambitious proposals to expand public oversight in training were effectively abandoned. In 1980, the Federal Constitutional Court declared even this watered-down reform unconstitutional, and therefore void – not because of the substantive policies it contained (the training levy was not the problem), but because the federal government had failed to consult with the Länder in the Bundesrat with regard to some administrative issues (Busemeyer Reference Busemeyer2009c: 92–5). The government then passed yet another diluted version of the law in 1981 (the Berufsbildungsförderungsgesetz, BerBiFG), which did not contain any provisions concerning the training levy and mostly served to establish the Federal Institute for Vocational Education and Training (Bundesinstitut für Berufsbildung, BIBB). Despite its initially ambitious plans, therefore, the social–liberal government largely failed to implement large-scale reform in VET (Baethge Reference Baethge and Lipsmeier1983).

At the same time, the institutions of the VET system continued to develop and evolve “below the radar of national legislative politics” (Culpepper Reference Culpepper2007: 183). Once unions realized that large-scale policy reforms would be impossible, they turned their attention to promoting their interests via corporatist bodies and at the firm level (Greinert Reference Greinert1998: 99; Kuda & Mignon Reference Kuda and Mignon1982: 66). A complex framework for the regular upgrade and reform of training occupations was established in the 1970s. The BIBB came to occupy a central place in this framework, managing the coordination between employers and unions on the one hand and between the federal and Länder levels on the other (Busemeyer Reference Busemeyer2009c: 97). Unions were on more equal footing with employers in this framework than in their partially subordinated position in the local Chambers, and they used their newfound power to press for broader and more theory-oriented training occupations. Common demands for a solid foundational education led to the enactment of the school-based Vocational Foundation Year (Berufsgrundbildungsjahr, BGJ). The KMK decided to implement the BGJ in 1972, and the federal government passed legislation to ensure that apprentices would get credit for it during their regular apprenticeship. Towards the end of the 1970s, it became clear that employers had found effective ways of getting around this regulation, for example by only taking apprentices who promised not to demand recognition for their prior study (Busemeyer Reference Busemeyer2009c: 102–3; Stratmann Reference Stratmann, Stratmann and Schlösser1990a: 192). It was for this reason that the BGJ (and similar instruments set up in the following years) never posed a fundamental challenge to the principle of employer autonomy in training decisions.

Consolidation and incremental adaptation (after 1982)

The 1980s and 1990s were a period of consolidation and, at best, incremental adaptation of the education system. This is most clearly the case in VET, the education sector in which the influence of the federal government is the greatest. When coming to power in 1982, the head of the new government coalition of Christian democrats and liberals, Chancellor Kohl, proclaimed a “spiritual and moral turnaround” (“geistig-moralische Wende”), signaling a break with the reform-oriented and progressive spirit of the previous government. The new federal education minister, Dorothee Wilms (CDU), blamed the SPD for having turned education policy into a “political battlefield” (Wilms Reference Wilms and Lipsmeier1983: 59, author's translation) and neglecting VET in favor of a focus on the expansion of higher education. According to Wilms, employers did not need that many university graduates and would have preferred more skilled workers instead (Wilms Reference Wilms and Lipsmeier1983: 65). There were too many unemployed academics already, she said, and the “educational illusions” (Wilms Reference Wilms and Demuth1986: 12) from the 1970s surmising that these academics would create their own demand on the labor market were wrong. Wilms insisted that the dual system of apprenticeship training and the differentiated three-tiered system of secondary schooling should be preserved as they were (Wilms Reference Wilms and Lipsmeier1983: 86; Reference Wilms and Demuth1986).

Demographics in Germany contributed to a high demand for apprenticeship places in the late 1970s and early 1980s, as also happened in the United Kingdom. Rather than threatening employers with the imposition of a training levy, however, as the previous government had done, the Kohl government refrained from direct intervention, instead asking employers’ associations and the Chambers to lean on firms to expand the number of available training slots (Busemeyer Reference Busemeyer2009c: 106–7). This appeal to the spirit of corporatism was effective in significantly increasing the number of training slots, especially in the crafts sector, during the first half of the 1980s. The government's thank-you for this engagement on the part of employers was to refrain from pushing any structural reforms in VET:

In this new political constellation, the repeated demands from unions and the SPD to introduce a training levy and other more far-reaching structural reforms fell on deaf ears (Busemeyer Reference Busemeyer2009c: 111–15). Trade unions largely continued their strategy of pushing for incremental reforms within the confines of the corporatist framework that had been established to update and reform training occupations (Hilbert et al.Reference Hilbert, Südmersen and Weber1990; Streeck et al.Reference Streeck, Hilbert, Van Kevelaer, Maier and Weber1987). Of crucial importance in this respect was the reform of training occupations in the metal and electrical industry; the bargaining between unions and employers on this lasted more than ten years and did not finish until 1988. The final agreement on reform was a great success for the unions. It reduced the total number of occupational profiles in the metal/electrical field, which guaranteed broader training instead of narrow, firm-specific skills acquisition. The reform also largely phased out so-called “staged apprenticeships,” which initially were meant to promote foundational training but in fact allowed employers to lay off apprentices after the first stage, resulting in reduced training and lower pay. The Christian democratic government was highly committed to the principles of corporatism and self-government through “private interest governments” (Streeck & Schmitter Reference Streeck, Schmitter, Streeck and Schmitter1985). It did not intervene strongly in the process of training reform and instead assumed the position of a neutral arbiter in the background (Hilbert et al.Reference Hilbert, Südmersen and Weber1990: 55), mediating conflict between the social partners.

The 1980s were a tranquil period of consolidation in general schooling and higher education as well. Individual Länder progressed on their own specific paths with regard to the expansion of comprehensive schooling. The government did pass higher education legislation in 1985 that allowed for greater differentiation in the system and granted more autonomy to individual universities, but it was not until much later, at the end of the 1990s, that significant steps were taken to move away from the oligarchic model of governance in higher education and towards a more deregulated and market-oriented one (Hepp Reference Hepp2011: 155).

The dominant topic of the 1990s is obvious: educational institutions in the new eastern Länder had to be reformed and integrated into a unified framework. As in other spheres of the welfare state, this was largely a story of institutional transfer from the West to the East; reunification was certainly not used as an opportunity to implement large-scale reforms of existing institutions. Because the comprehensive school (polytechnische Oberschule, POS) had been the central form of secondary schooling in the GDR, it suffered a steep drop in legitimacy after reunification. The new Länder all opted for a more differentiated system, but significant differences emerged between them according to the political constellations at the beginning of the 1990s (Hepp Reference Hepp2011: 214; Overesch Reference Overesch2007; Stern Reference Stern2000; Von Below Reference Von Below2002). Brandenburg (governed by the SPD) favored the establishment of comprehensive schools, but also kept the popular Gymnasien alive. Mecklenburg-Vorpommern (where the CDU was in charge for most of the 1990s) was the only Land that maintained the classical West German three-tiered system. Saxony established a two-tiered system, with the traditional Gymnasium as one pillar and a merged Realschule and Hauptschule (called Mittelschule or “middle school”) as the other (Edelstein & Nikolai Reference Edelstein and Nikolai2013).

Institutional transfer in the field of VET was not as easy as in other parts of the welfare state. The formal legal and institutional framework was transferred quite quickly and easily through such actions as establishing Chambers and creating regulations on the recognition of vocational qualifications awarded in the GDR. The dual system of apprenticeship training had evolved and developed within an institutional ecology of intermediate associations in the West, however, which could not be easily transplanted or created from scratch in the eastern Länder (Culpepper Reference Culpepper2003; Johnson Reference Johnson and Wiesenthal1995). Large state-owned firms had been the central pillars of the vocational training system in the GDR. When these firms went bankrupt after reunification, there was no viable crafts sector that could take over apprenticeship training: the federal government and Länder governments had to step in by heavily subsidizing firm-based, school-based, and out-of-firm training (Busemeyer Reference Busemeyer2009c: 132–3). Initially regarded as a short-term measure, public subsidization of training at a much higher level than in the West has become an important and lasting feature of VET in the new Länder.

The 2000s were a period of more heightened reform activity than the 1980s and 1990s, due in part to the more activist stance of the red–green government coalition (SPD and Greens) that came into power in 1998, but also to important international developments such as the Bologna Process and the PISA debate, as well as the 2006 institutional reforms of the federalist structure. According to Hepp (Reference Hepp2011: 69–70), partisan politics in education policy-making has now become less ideological and more oriented towards problem-solving, especially compared to the 1970s (although Overesch (Reference Overesch2007) reached the opposite conclusion when comparing Germany to Finland). The PISA shock in the early 2000s resurrected the debate about structural reforms of the three-tiered system. Regardless of ideological debates, increasingly practical concerns have provoked a number of Länder to merge the two lower tiers (Hauptschule and Realschule), essentially following the example of Saxony in the early 1990s (Edelstein & Nikolai Reference Edelstein and Nikolai2013). As expansion of access to higher levels of education has continued, the lowest tier (Hauptschule) has become an increasingly unpopular choice. Demographic changes (a decrease in the number of pupils) and general fiscal pressures have made the costs of maintaining three different school types oppressive, especially in rural areas. Merging the two lower tiers into semi-comprehensive schools (labeled differently in different Länder) and maintaining continued independence of the academic Gymnasien has become a general trend, and a potential solution to the decades-long conflict over school structure. Although this trend represents a cautious movement towards the comprehensive model, a full-scale comprehensivization of the system seems very unlikely – especially the abolishment of the Gymnasien – because it would face significant political opposition from middle-class voters.

The Bologna Process triggered a surprisingly far-reaching process of internationalization and deregulation of the higher education sector (Voegtle et al.Reference Voegtle, Knill and Dobbins2011). For example, the structure of a large majority of study programs was changed from the traditional four- or five-year diplomas to the internationally more common two-cycle structure of bachelor's and master's degrees. By contrast, the attempts by the red–green government to prohibit tuition fees and establish junior professorships as an additional career pathway to full professorship (instead of the traditional Habilitation) were ruled unconstitutional by the Federal Constitutional Court because they infringed on the autonomy of the Länder (Hepp Reference Hepp2011: 157). There is nevertheless a general trend towards introducing market-oriented governance mechanisms in higher education (Kamm & Köller Reference Kamm and Köller2010). Such mechanisms include strengthening the autonomy of universities, introducing cost–benefit budgeting, and establishing new university boards that feature representatives from industry and other stakeholders. For a rather brief period in the 2000s, various Länder governments introduced tuition fees at the comparatively low level of €500 per semester. Because tuition fees were always introduced by Christian democratic governments and then revoked once the Social Democrats regained power (Kauder & Potrafke Reference Kauder and Potrafke2013), by 2014 no Land will require the payment of fees.

As for VET, German governments of different stripes have promoted the flexibilization and differentiation of the training system by doing such things as allowing firms more leeway in implementing training profiles and waiving or abolishing regulatory policies (Busemeyer Reference Busemeyer2009c, Reference Busemeyer2012; Deissinger & Hellwig Reference Deissinger and Hellwig2005). These reforms have proceeded in an incremental manner, but may end up producing a significant change in the character and institutional logic of the system, making it more attuned to the needs of training firms than of unions or young people (Thelen & Busemeyer Reference Thelen2012). For the moment, however, the reforms have also helped to maintain the popularity of the system with employers (Bosch Reference Bosch, Bosch, Krone and Langer2010; Hassel Reference Hassel, Hancké, Rhodes and Thatcher2007), and the core of the training system in the export-oriented parts of the economy remains very stable. One negative side effect of this stability is the rise of the so-called “transition system,” which is a complex and uncoordinated arrangement of school- and workplace-based training measures, reminiscent of UK-style “make-work” schemes. Similar to developments in social and labor-market policy (Palier & Thelen Reference Palier2010), this differentiation between a stable core and increasingly precarious transition paths for low-skilled youths at the periphery amounts to a form of dualization (Busemeyer Reference Busemeyer2012a,Reference Busemeyerb).

Critical junctures in the postwar development of upper and post-secondary education

The notions of critical junctures and punctuated equilibria were prominent features in the early generation of scholarship in historical institutionalism (Steinmo et al.Reference Steinmo, Thelen and Longstreth1992; Thelen Reference Thelen1999). But as Streeck and & Thelen (Reference Streeck, Thelen, Streeck and Thelen2005: 1) have pointed out, drawing a “sharp distinction between long periods of institutional stasis periodically interrupted by some sort of exogenous shock that opens things up” is too simplistic a picture of the complex process of institutional change in advanced political economies. To a significant extent, both the distinction between institutional change and stability and the definition of a “critical juncture” depend on how much the observer “zooms in” on the process of institutional development. If one studies large-scale change in polities (e.g., constitutional reforms or revolutions; Skocpol Reference Skocpol1979), as the early work in historical institutionalism did, the punctuated-equilibrium model might be the appropriate one, and critical junctures are critical in the sense of being relatively short episodes of radical change. In advanced political economies, however, the institutional density and complexity is much higher than in preindustrial societies. In such a setting, large-scale policy change is always less likely than incremental change (Streeck & Thelen Reference Streeck, Thelen, Streeck and Thelen2005).

This does not necessarily imply, however, that the notion of critical junctures cannot be applied in a meaningful way in these cases. By “zooming out” of the cases studied above, it is possible to identify certain time periods in which the potential for reform was higher than in others. These windows of opportunity open up when policy-makers face different options for the institutional development of a previously underdeveloped policy field. Indeed, compared to primary and lower-secondary education, as well as other social policies, upper-secondary and post-secondary education was an underdeveloped policy field that had not yet been fully institutionalized in European countries in the postwar period. Over time, the institutional density of policy fields increases and the window of opportunity for reform closes, because, as Pierson (Reference Pierson2000: 253) has written, events that happen earlier in a chain of path-forming events have more significant implications than events that occur later.

We can thus distinguish different stages in the process of path formation, which can be called path initialization, path formation, and path consolidation. The concept of critical junctures is most meaningful in the second stage, when switching to alternative paths is still a politically feasible option.

Path initialization refers to the idea that policy legacies matter even in the relatively early stages of path formation. But policy legacies do not fully determine future events, and path switching could and did occur in the three cases studied above. While the cases had relatively similar policy legacies in the immediate postwar period (in comparison to countries like the United States and Japan), it is obvious that there were some significant differences. Even early on, the status and importance of apprenticeship was probably higher in Germany than in Sweden, although Sweden still had an apprenticeship system, whereas the ideological appeal of the universal welfare state model was stronger in Sweden. These facts can of course be taken as foreshadowing later events, but there is room for political action and conflict. The “shadow of the 19th century” (Iversen & Soskice Reference Iversen and Soskice2009) may be long, but it is not endless. The case studies in this chapter have shown that there were significant political struggles over the institutional design of education systems in the 1950s, 1960s, and 1970s, and that these had strong implications for the future development of education systems.

The second stage, path formation, occurs when decisive choices are made about the design of institutions, choices that with hindsight can be identified as critical junctures. It is important to emphasize, however, that path formation during critical junctures is still a process, which may happen over an extended period of time. Rarely does one single instance of reform change development paths completely. Large-scale reforms, by contrast, proceed through various stages before they result in significant institutional change (see, for instance, the long process of comprehensivization in the Swedish system). It follows that the long-term balance of power between partisan actors matters more than the short-term partisan composition of governments. But it is not only partisan politics that matters: the long-term political sustainability of a chosen path depends on the congruence between the policy reforms and the socioeconomic context; that is, the institutional setup of the political economy and the interests of labor-market actors. When there is a hegemonic partisan coalition and the congruence between policies and socioeconomic institutions is high, path consolidation happens earlier and development paths are more sustainable over time. Critical junctures should thus not be conceptualized as brief periods of large-scale change, but rather as windows of opportunity that can be open for a longer stretch of time. Policy-makers may need to enact several steps of policy reform during such a stretch, each of which might look incremental at first sight, but which taken together will add up to a significant degree of institutional change over time (Streeck & Thelen Reference Streeck, Thelen, Streeck and Thelen2005: 8).

Each of this chapter's cases went through a critical juncture, but the specific timing varied. In the case of Sweden, the construction of the Nordic model of education started under social democratic hegemony in the 1950s and lasted roughly until the 1970s. The crucial reform of upper-secondary education in 1972 had a lasting, path-shaping impact. The integration of VET into the general secondary school system contributed to the phasing out of workplace-based forms of apprenticeship training, leading large companies to make changes to their personnel strategies. The promotion of educational mobility within the egalitarian education system exacerbated academic drift towards higher education to a greater extent in Sweden than in Germany.

The critical juncture in Germany happened in the 1960s and 1970s. The 1969 law on VET stabilized and institutionalized the dual system of apprenticeship training and its core features, especially the dominance of firm-based training over school-based forms. The prewar three-tiered secondary education system had been restored by the 1950s, but the real window of opportunity for reforms opened up after the social–liberal coalition came into government. This coalition started out with an ambitious reform agenda that prioritized education. The prevailing zeitgeist greatly favored the creation of a comprehensive education system similar to the Swedish one; important components of the government's plans included integrating VET into the secondary school system and strengthening school-based forms of VET. The failure of most of these policy reforms indicated a consolidation of the specific path of institutional development, characterized by a strong and deeply institutionalized separation between academic and vocational education at the secondary and post-secondary levels.

The UK is a less clear-cut case, and not only because the Conservatives had a less hegemonic position than the Christian Democrats in Germany or the Social Democrats in Sweden. That being said, the Conservatives in Britain were certainly more influential than Labour during the postwar decades, particularly after 1979. The system could be considered a hybrid case until 1979, since it combined certain features of the Scandinavian and German models and added these into a more conservative and elitist model. As in Sweden, comprehensive schools spread throughout the 1960s and 1970s, but independent private schools survived as well. Like in Germany, apprenticeship training continued to be important in the manufacturing industry, and it was actually a Conservative government that passed the collectivist 1964 Industrial Training Act and expanded access to higher education in the early 1960s. The Thatcher government was crucial in setting the United Kingdom on a liberal-conservative development path, both in education and in welfare state policies more generally, making the 1980s the critical juncture for the British case. The 1988 Education Act introduced the principles of competition and choice into the general school system. The collectivist and corporatist institutional framework of the training system was dismantled; vocational training came to be regarded by both the government and employers as a social policy program for unemployed youths, instead of an investment in skilled labor. Once VET had ceased to be a viable route to secure and well-paid employment, academic drift came to be even more pronounced in the United Kingdom than in countries such as Sweden.

The third stage, path consolidation, is reached when the dominant political and economic actors adapt their preferences and strategies to reflect the institutional status quo. This does not preclude any kind of institutional change, but the fundamental structure of institutions will rarely be challenged, and change is bound to reflect existing policy legacies to a stronger extent than in the previous phase of path formation. As a consequence, partisan conflict matters less in this later stage, especially short-term changes in the partisan composition of governments. While it is true that studies comparing the policies of different governments over time within individual countries will always find differences that are related to the different electoral constituencies of political parties, the range of feasible policy options decreases because of path dependencies.

The case studies include numerous examples of path consolidation. In the United Kingdom, for example, the New Labour government of the mid-1990s and early 2000s largely continued along the lines of its Conservative predecessor. New Labour did not go back to the collective, corporatist approach to training policy of the 1960s and 1970s, but kept the voluntarist, employer-led framework in place instead. It also continued to promote choice and decentralization in the area of school policy. The coalition government likewise continued along the path set out by New Labour, particularly in higher education policy and with respect to establishing new kinds of schools independent of local authorities. Despite its newfound enthusiasm for apprenticeships, the coalition government did not return to the collective approach of the 1960s either, but stuck with the liberal voluntarist system.

The most significant educational reforms in Sweden were the decentralization and privatization reforms of the 1990s. These reforms did introduce more competition into the Swedish education system, but the case study shows that this change did not yet pose a fundamental challenge to the political and normative logic of the Nordic model of education. Most tellingly, the political conflict was about the division of power between localities and the central government in governing the education system, not about the place of VET in the education system, the reintroduction of apprenticeship training, access to higher education, or other aspects that would have represented far-reaching changes. The fundamental characteristic of the Swedish education system today remains in place: a statist and integrationist model of skill formation in which VET is fully integrated into the upper-secondary school system.

A similar conclusion can be drawn for the case of Germany. This country has gone through a period of intensified reform activity in the last decade, but its fundamental characteristics have been maintained. The red–green coalition initiated a number of significant reforms in the welfare state, and the education system – particularly higher education – went through a phase of internationalization. The privileged status and importance of dual VET (apprenticeship) has remained unchallenged, however, and is supported by most influential political actors, including unions. This is why the red–green coalition did not take up the more ambitious plans of the social–liberal coalition of the 1970s to fully integrate VET into secondary schooling. Although there is a trend towards slowly changing the structure of schools from a three-tiered to a two-tiered structure, a full-scale comprehensivization along Swedish lines is politically unfeasible for the foreseeable future.

Partisan politics in context

As documented in Chapter 1, there now exists a large and growing literature on the partisan politics of education (Ansell Reference Ansell2008, Reference Ansell2010; Boix Reference Boix1997, Reference Boix1998; Busemeyer Reference Busemeyer2007, Reference Busemeyer2008, Reference Busemeyer2009b, Reference Busemeyer2012a; Castles Reference Castles1989, Reference Castles1998; Jensen Reference Jensen2011; Nikolai Reference Nikolai2007; Rauh et al.Reference Rauh, Kirchner and Kappe2011; Schmidt Reference Schmidt1980, Reference Schmidt2002, Reference Schmidt and Castles2007; Wolf Reference Wolf2009; Wolf & Zohlnhöfer Reference Wolf2009). Partisan preferences can vary across a number of policy dimensions, such as the preferred level of public or private investment in education, the distribution of resources across educational sectors (early childhood, compulsory, academic, and vocational education), the relative importance of education vis-à-vis other kinds of social policy, and the preferred governance model, including preferences concerning which kinds of organized interest should be involved. In this chapter I have concentrated on two aspects: the division of labor in financing education (commodification) and the stratification of education in terms of institutions.

I will now apply the theoretical framework – essentially an extended partisan model – developed in Chapter 1. I will first comment briefly on similarities and differences in partisan preferences across country cases, and then discuss the extensions to the standard partisan model proposed above. I have already discussed the third extension (i.e., the focus on long-term balance of power instead of the short-term partisan composition of governments, covered in the previous section). In this section, I concentrate on the two remaining extensions: the discussion of partisan preferences about the political process in addition to policy content, and the role of the institutional context, particularly in terms of economic coordination.

The case studies reveal some similarities in partisan preferences across countries. Finding these similarities (independent of context in terms of institutions and idiosyncratic party-specific ideological and organizational legacies) can be taken as support for the general model of partisan theory. At a very general level, left-wing parties are in favor of expanding access to higher levels of education for their core electoral constituencies in the lower half of the income and skills distribution. This also implies a preference for strong state involvement in the financing and administration of education, such as increases in public spending and individual subsidies for students, and a determination to limit educational stratification. For example, in all three countries the social democratic parties have become strong supporters of the model of comprehensive schooling, although the German Social Democrats have been less effective in implementing this preference because of the disadvantageous balance of power. Right-wing parties, by contrast, care less about educational expansion and are more supportive of private involvement in the financing and administration of education; they have strong preferences for a differentiated education system with extensive tracking. In this regard, the ideology of Christian democrats is not all that different from that of the secular conservatives. In all three countries, bourgeois parties have been the defenders of the traditional segmented school system, which also implies restricted access to higher education and a strict separation between VET and general schooling. Again, the bourgeois parties have been more or less successful in defending this model, depending on the balance of power (the Swedish conservatives’ opposition to the comprehensive secondary school was ineffective, for instance).

In sum, the standard model of partisan theory assumes that different political parties of the left and of the right have similar policy preferences regardless of the context, and that differences between country cases emerge solely because of differences in the balance of power between partisan families. The standard model should work best in the early stages of path formation, before institutional legacies and contexts begin to shape the range of feasible policy options. One extension to this standard model is to argue that political parties have preferences with regard not only to policy content but also to how the political process should be organized. Whereas secular conservatives aim to minimize the influence of labor (Wolf Reference Wolf1998), Christian Democrats pursue a “politics of mediation” (Van Kersbergen Reference Van Kersbergen, Kitschelt, Lange, Marks and Stephens1999: 356). This has implications for the range of feasible policy options. As mentioned above, there are some similarities between conservative and Christian democratic ideology with regard to policy content, such as a preference for a segmented secondary school system and elite universities that is justified by the organic division of labor in society. The most important difference, however, may be the approach to apprenticeship training. British Conservatives, especially after 1979, regarded collective institutions such as the levy-grant system and the training boards as infringing on the autonomy of individual firms, and they did not hesitate to use the policy reforms of the wage-bargaining and training system in the 1980s to weaken the influence of trade unions. The German CDU, by contrast, pursued a corporatist policy approach that encouraged cross-class compromise and, to a certain extent, deliberately granted influence to union voices. The Christian democratic approach to training respected employers’ autonomy to train as they wished, but at the same time demanded that they carry out their duty to train young people. Following the principle of subsidiarity, intermediate associations have been given a central role in the management and operation of the training system, even as the state remains a neutral moderator in the background, ready to intervene if necessary.

As expected, the case studies revealed some diversity within the social democratic family with regard to procedural preferences. In Britain, political parties are inclined to coalesce with and grant privileged access to those economic interests that belong to their electoral constituencies (business in the case of the Conservatives and unions in the case of Labour). In principle, parties are less keen on engineering cross-class consensus, which may be related to the majoritarian nature of the United Kingdom's political institutions and the liberal character of the economy. However, we also saw that there was a time when Labour and the Conservatives were able to form a broad coalition supporting education reforms in the immediate postwar period, for example in the case of comprehensive secondary schools and in the establishment of the levy-grant system in training. The consensual period ended decisively when Thatcher brought back the “class warfare” spirit. New Labour then tried to appeal to business interests while still maintaining ties to the unions, which could be interpreted as a cautious attempt to create some sort of cross-class compromise. In both Sweden and Germany, social democratic parties have historically been more open to the notion of corporatist bargaining.

The comparison between Germany and Sweden on the one hand and the United Kingdom on the other leads to the next issue: even though political parties may be willing to promote cross-class compromise, their ability to do so depends on the organizational foundations of labor-market actors and the institutional setup of the economy. Partisan politics does not happen in a societal vacuum, but is embedded in and interacts with institutional contexts. The linkage between political parties and organized labor-market actors varies across countries. This goes back to the old distinction between pluralist and corporatist systems of interest mediation (e.g., Lehmbruch Reference Lehmbruch, Schmitter and Lehmbruch1979; Schmitter Reference Schmitter, Schmitter and Lehmbruch1979), which is reflected in the recent differentiation between LMEs and CMEs (Hall & Soskice Reference Hall, Soskice, Hall and Soskice2001). In LMEs, the coordination among economic interests follows the market logic; this implies a pluralistic conception of interest mediation. It is less likely that unions and employers’ associations will engage in the formation of sustainable cross-class coalitions, nor is it likely that organized economic interests will enter into close alliances with state actors or partisan policy-makers, institutionalized in corporatist bodies. In contrast, corporatism is strong in CMEs. This includes both the monopolization of functional representation via peak associations and tight linkages between these associations and the policy-making system. In sum, the concept of economic coordination entails two aspects: coordination among organized economic interests in the labor-market arena and coordination between these interests and policy-makers in the political arena.

Socioeconomic institutions are also associated with particular patterns of democratic decision-making (Lijphart Reference Lijphart1999). Where coordination is low, as is the case in majoritarian Westminster democracies, policy-makers may be able to effect more significant policy change, but given the lack of coordination, it is an open question ex ante whether this policy change will actually lead to real change. If there is no congruency between the policies and the institutional context, they will remain ineffective in triggering real institutional change. Where coordination is high, policy-makers can hardly pass policies against the joint opposition of organized interests; instead, they must engage in the formation of politico-economic alliances that span the gap between the policy-making and labor-market arenas. Once agreement on policy reform is reached, however, it is more likely to lead to real institutional change that is sustainable in the long run.

In the United Kingdom, an LME, the formation of a sustainable cross-class coalition was not feasible in the long run. The low level of coordination in both the business and labor camps led to pluralist forms of interest mediation and relatively weak peak associations, with little influence. It thus prevented the establishment of representative monopolies, which would have been necessary for the establishment of a viable system of private-interest government via associations (Streeck & Schmitter Reference Streeck, Schmitter, Streeck and Schmitter1985). The tumultuous 1970s, with their high levels of strike activity, delegitimized and put an end to earlier attempts to establish corporatist institutions in training and manpower policy more generally. The wildcat strikes during the Winter of Discontent (1978–79) are a good example of policy-makers’ failure to coordinate with organized economic interests: although Labour was in government at the time, it could not prevent unions from striking. The lack of coordination between labor-market interests and policy-makers may also explain the government's repeatedly unsuccessful attempts to promote apprenticeship training (Finegold & Soskice Reference Finegold and Soskice1988; King Reference King1997; Ryan & Unwin Reference Ryan and Unwin2001). Because policy-makers cannot credibly commit labor-market actors to long-term strategies of skill formation, employers will refrain from investing in skills beyond short-term needs and unions will be unwilling to accept lower wages for apprentices, which would be a necessary condition for skill investments by firms (Marsden & Ryan Reference Marsden, Ryan and Ryan1991).

On the other hand, cross-class coalitions are a hallmark of CMEs. A well-known argument in the literature is that coalitions between employers and unions in matters of skill formation and social policy are important historical foundations of the institutional framework of CMEs (Cusack et al.Reference Cusack, Iversen and Soskice2007; Mares Reference Mares2003; Martin & Swank Reference Martin and Swank2008, Reference Martin, Busemeyer and Trampusch2012; Swenson Reference Swenson2002; Thelen Reference Thelen2004). What is less appreciated in the literature, however, is that there can be different types of cross-class coalitions, depending on which partisan actor is the formateur during the critical juncture (Busemeyer Reference Busemeyer2012a). Cross-class coalitions can be more or less encompassing, and policy-makers as political and partisan actors can selectively grant privileged access to certain kinds of organized interest.

In Sweden, the hegemony of social democracy was a consequence not only of the strong position of the SAP in parliament and government, but also of the extraordinary power of the trade-union movement. As is well known, employers responded to the threat of organized labor by establishing extensive coordination mechanisms on their own (Martin & Swank Reference Martin, Busemeyer and Trampusch2012; Swenson Reference Swenson2002), and the 1938 Saltsjöbaden agreement laid the foundation for sustainable cross-class compromise and macro-corporatist wage bargaining. Nevertheless, employers’ associations were the weaker partners in this cross-class coalition. In the 1950s and 1960s, the Social Democrats’ preference for statist solutions and the unions’ worries about apprentices being misused as a cheap source of labor led to the dismantling of the institutional legacies of firm-based apprenticeship training from the postwar era. The 1972 reform of upper-secondary education was supported by both unions and employers, but the SAF would have preferred a much stronger linkage of VET to the workplace (Lundahl Reference Lundahl1997: 95). Once the path was selected, however, labor-market interests adopted their strategies in the same way as the political parties had done. When Swedish conservatives promoted the resuscitation of apprenticeship training in the 1980s, the SAF was initially supportive, but it then became quite reluctant, because employers had already adjusted their recruitment strategies and were not willing or able to invest in broad occupational skill formation (Lundahl Reference Lundahl1997: 98). Unlike the unions in Germany, Swedish unions have strongly supported the integrated secondary system and school-based VET.

The German case is different from the Swedish one. Here, the formateur of the cross-class coalition was the Christian Democratic Party, not the Social Democrats. In this case, business interests had a privileged position compared to unions. The passing of the BBiG in 1969 provides a good example of the specific bias of this variety of cross-class coalition. The law did not fundamentally challenge the privileged position of training firms in training matters, even though the SPD was a formal member of the government coalition. Unions were very much part of the social–liberal reform coalition of the 1970s, but this coalition eventually failed because of effective opposition from employers, who were threatening to withdraw from apprenticeship training in a time of rising youth unemployment. After this critical juncture passed, unions changed their strategies, working through corporatist institutions at the firm and industry level, as well as the complex institutions set up to reform training occupations. Unions have since become strong defenders of firm-based training over school-based alternatives (Busemeyer Reference Busemeyer2012a), even more so than the SPD.

These cases are excellent examples of how actors adopt their strategies and preferences once a critical juncture has passed. They also demonstrate how “policies shape politics” (Pierson Reference Pierson2005: 37–9). German trade unions supported the establishment of a more statist skill-formation regime in the 1970s because they hoped this would increase their political influence in the administration of the system. The prevailing dominance of firm-based apprenticeship and the continued central role of Chambers of Industry and Commerce in the governance structure of this system crushed the unions’ aspirations. In the United Kingdom, the Thatcher government's use of training policies to break the influence of trade unions was even more blatant.