Types of welfare state regimes

Using these three principles – de-commodification, social stratification, and private-public mix – Esping-Andersen (Reference Esping-Andersen1990) classified 18 OECD countries into three ideal regime types: social democratic, liberal, and conservative. Each represents a qualitatively different form of social protection that orders social relations across industrialised countries.

Social democratic welfare state regimes comprise the smallest regime cluster. The Nordic countries (Denmark, Norway, Sweden, Finland) belong in this category. Welfare provisions typically offer universal, generous benefits, treat all residents equally, and provide higher standards than minimal need. In other words, the state sets a high equity standard. The social democratic regime type thus reflects the values of solidarity and equality. The object of social democratic welfare policy is to guarantee that individuals and families “can uphold a socially acceptable standard of living independently of market participation” (Esping-Andersen, Reference Esping-Andersen1990, p. 37) but without discouraging individual aspirations or denying the opportunity for socioeconomic achievement.

In the liberal welfare regime type (Canada, USA, Australia, New Zealand, UK, Ireland), welfare coverage is typically minimal and benefits are modest, recipients are usually means-tested and often stigmatised. Characteristically, liberal welfare states have a low degree of de-commodification and a strong emphasis on market solutions in social provision. Market and private insurance plans may be encouraged by the state as an alternative to public protection. Specifically, the liberal welfare type orients economic welfare toward dependency on the market. Hence, this reflects the values of individual responsibility, efficiency, and the view that the state should rely primarily on market forces and work with these forces to prevent destitution and provide essential social services (Huber & Stephens, Reference Huber and Stephens2014).

Protection of hierarchy, preservation of social structures, and striving for harmony is the focus of conservative welfare regimes (France, Japan, Italy, Switzerland, Germany, Netherlands, Austria). One such structure is the traditional family. Strong de-commodification is accomplished with the state as the primary provider of welfare insurance. The benefits of welfare programs are often earning-related, administered through employers, and geared towards maintaining existing social patterns. The primary provider of welfare insurance is the state, with the market and private insurance having only marginal roles. The state is charged with ensuring the elimination of poverty but not changing the social order; it performs only the functions that are not performed well by the family or civil society (van Kersbergen, Reference van Kersbergen1995).

Criticisms of welfare regime theory

Since its publication in 1990, The Three Worlds of Welfare Capitalism has become one of the most central and most cited contributions to recent comparative social policy (Powell, Yörük, & Bargu, Reference Powell, Yörük and Bargu2020). Nevertheless, this typology is not without criticism. Four criticisms have drawn the most attention in the literature: nations are misclassified, his approach is “gender blind,” his original typology does not consider the redistributive effect of social policies but only expenditures, and classifications are too few or incomplete. Mitchell (Reference Mitchell1992) argued that typologies like Esping-Andersen’s are “beset by boundary problems which result in the misclassification of some countries” (p. 73). However, there is now a growing consensus that welfare states are rarely pure types and are usually hybrid cases (Arts & Gelissen, Reference Arts and Gelissen2002; Ferragina & Seeleib-Kaiser, Reference Ferragina and Seeleib-Kaiser2011). Addressing feminist critiques, more recent analyses attempt to identify policies associated with (de)familization and (de)genderization and incorporate them into welfare state classifications (Chybalski & Marcinkiewicz, Reference Chybalski and Marcinkiewicz2021; Saxonberg, Reference Saxonberg2013). Researchers have also integrated redistributive social services such as health care and social services (Bambra, Reference Bambra2007; Powell & Barrientos, Reference Powell and Barrientos2004). Finally, numerous additional welfare state types have been proposed since the publication of Esping-Anderson’s original typology. These include a fourth “southern European” world of welfare capitalism described as having “high standards of employment protection” that results in labour markets characterised by high levels of youth unemployment and low participation by women (Manow, Palier, & Schwander, Reference Manow, Palier and Schwander2018); the radical welfare state (Castles & Mitchell, Reference Castles and Mitchell1992); the “southern” world of welfare (Bonoli, Reference Bonoli1997; Ferrera, Reference Ferrera1996; Leibfried, Reference Leibfried, Ferge and Kolberg1992); “the three worlds of post-Soviet transition” (Malinovskiy & Shibanova, Reference Malinovskiy and Shibanova2023); as well as East Asian and post-Soviet welfare state types (Aidukaite, Reference Aidukaite2009; Fenger, Reference Fenger2007; Kwiek, Reference Kwiek2014; London, Reference London2018).

Welfare regime theory and education

The welfare state complex comprises not only a nation’s social policy repertoire but also includes the varying roles of the state, family, and market in providing social protection. In the 1950s, education was incorporated as an essential part of a coherent welfare policy structure (along with health care, education, housing, and employment) in Western industrialised nations. Education has been acknowledged in the research literature as “clearly one component of a nation’s total public policy package and… not independent of other social policies” (Hokenmaier, Reference Hokenmaier1998, p. 710). However, while education is understood to play a central role in social stratification more generally, education and skill acquisition and the interaction between education systems and labour markets were largely neglected from early analyses employing welfare regime theory; it is still frequently ignored in the social policy literature (Busemeyer & Nikolai, Reference Busemeyer, Nikolai and Castles2010; Di Stasio & Solga, Reference Di Stasio and Solga2017).

Researchers began to develop welfare regime approaches focused on education in the early 2000s. In a seminal comparison of social and educational policy in advanced industrial societies, Hega and Hokenmaier (Reference Hega and Hokenmaier2002) focused on the educational system/social protection nexus by looking at the relationship between spending on education and spending on social insurance programs. Specifically, they analysed social insurance spending as a percentage of GDP and the ratio of educational expenditures as a percentage of total public spending and as a proportion of GDP. They tested a hypothesis originally proposed by Heidenheimer (Reference Heidenheimer, Flora and Heidenheimer1981), in which competing aims of national policies, such as reducing inequality and increasing social mobility, are understood to often result in “trade-offs” among programs, services, and system development. Some welfare regimes have given political and budgetary preference to social insurance programs over educational programs. In other welfare states, higher levels of public investment in policies and programs to enhance educational opportunities beyond a primary level may be viewed as an alternative policy strategy to income maintenance programs and other social insurance guarantees.

Hega and Hokenmaier (Reference Hega and Hokenmaier2002) determined that conservative nations tend to place greater emphasis on social insurance expenditures rather than on education, relative to the other welfare state types. Germany fits into this pattern where it tends to lead in social insurance spending as a percentage of total public spending but is a laggard in educational spending. Lack of support for educational expansion suggests choices by certain regime types to favour the reproduction of existing class structures and labour market conditions. In contrast, in liberal nations, expenditures on social insurance are relatively low in comparison with other OECD countries, while expenditures on and participation in education are high, due to an overt focus on human capital formation. The final group of social democratic countries does not exhibit a trade-off; rather, spending on social protection and education are both high.

Allmendinger and Leibfried (Reference Allmendinger and Leibfried2003) used the criteria of skill differentiation and overall levels of educational attainment to construct a welfare regime typology of education systems. They divided countries according to the extent of differentiation in competencies and the absolute level of competencies, using outcomes from the Programme for International Student Assessment (PISA) scores and years of schooling or credentials attained. They identify four worlds of competence production that differ in their focus on educational preventative policy and compensatory social policy. They are fairly similar to Esping-Andersen’s “Three Worlds,” with the addition of a “peripheral” southern grouping.

Peter, Edgerton, and Roberts (Reference Peter, Edgerton and Roberts2010) also examined secondary education from a welfare regime perspective. By measuring between-school and within-school educational inequality (socio-economic gradients) in student achievement using PISA 2003 mathematics, reading, and science assessment scores, they found support for Esping-Andersen’s “three worlds” typology. The highest between-school inequalities were in the grouping of conservative welfare states and the lowest between- and within-school inequalities in the social-democratic grouping. The liberal grouping showed moderate within- and between-school inequality.

An Anglo-American, Scandinavian, and Continental welfare regime classification is employed by Ansell (Reference Ansell2008), on which to map three types of higher education arrangements: partially private, mass public, and elite. He argues that higher education policy materialises from choices made from a “trilemma” of enrolment levels, subsidisation provided, and overall public cost. The Anglo-American regime is associated with the “partially private” higher education systems characterised by high enrolment and high private but low public spending. The Scandinavian regime corresponds with a “mass public” higher education system defined by high enrolment, costs that are highly subsidised, and high public spending. The “elite” label is reserved for the Continental welfare regime, where costs are highly subsidised and levels of private financing and enrolment are low.

Examining the different national approaches to higher education involvement, funding, tuition, and student financial aid through a multivariate comparative analysis of 16 OECD countries, Pechar and Andres (Reference Pechar and Andres2011) assert that there is a convincing relationship between higher education policies and given welfare regimes “in terms of participation in tertiary education, pre-tertiary prerequisites, educational funding, and levels of tuition and levels of borrowing” (p. 41). For example, based on enrolment and graduation patterns, they revealed that conservative regimes clearly have the lowest tertiary education growth and the liberal regimes have the highest. Liberal regimes had the largest total expenditure on educational institutions as a percentage of GDP at the tertiary level, followed by social-democratic regimes, and conservative regimes in last place. As for tuition fees and student aid, the Nordic countries have “no or low tuition fees together with generous student support systems,” while the liberal countries are defined by “high levels of tuition fees and along with well-developed student support systems; and the conservative countries have low levels of tuition fees and less developed student support systems” (p. 41).

More recently, Johnson (Reference Johnson and Jungblut2023) confirms earlier findings by showing that both public expenditure and the share of private expenditure on higher education correspond with the three welfare regimes. That is, high public spending is clearly associated with social-democratic regimes, low levels with liberal regimes, and medium levels with conservative regimes. Compared with the other regimes, liberal regimes exhibit the highest share of private funding. In addition, he demonstrates that public/private investment in higher education is most closely aligned with liberal regimes, lowest in conservative regimes, with social-democratic regimes falling in the middle. In terms of higher education funding, liberal (more market-type) and social-democratic regimes (more combined blocked funding) are at opposite ends of the continuum, with conservative welfare regimes (balanced mix of blocked and competitive funding) in the middle. Regarding mechanisms for financial support of students, liberal regimes rely on loans and means-tested grants for economically disadvantaged students; support is offered through scholarships, grants, and loans in social-democratic regimes. Student financial support in conservative regimes is described as “less well developed” (p. 218) with more dependence on family support in southern European countries.

Willemse and de Beer (Reference Willemse and de Beer2012) also use a welfare state approach to analyse the field of higher education. Using both theoretical and empirical methods to measure decommodification and stratification, they tested groupings of countries by educational system characteristics and found that they mapped broadly onto original regime types. They determined that social democratic nations lead in both social insurance and educational commitments. State policy largely maintains low or no educational fees, accessible and generous grants and loans, and high public/low private expenditure, resulting in high educational enrolment rates. Liberal welfare states, such as the USA, perform strongly in all categories of educational spending and lead in general education participation at the secondary level. Conservative countries, such as Germany and Austria, with their “dual systems” of vocational education, have high enrolments and spending for vocational education and offer a large variety of vocational curriculum, but also maintain hierarchical educational outcomes that result in social differentiation.

Studies focused on education often reclassify or include countries not captured within the original framework provided by Esping Anderson. For example, in examining education expenditure and equality, West and Nikolai (Reference West and Nikolai2013) specify four country groupings: Nordic countries, with publicly funded comprehensive school systems, integrated vocational education, and high equality of opportunity; Continental countries, with highly tracked and stratified school systems, differentiated vocational education, and large social background effects on educational outcomes; Mediterranean countries, with moderate public expenditure, low stratification in early education and high stratification in later education, and standardization and vocational specificity relatively less pronounced than in the Continental countries; and English-speaking countries, with high public expenditure on all levels of education, a quasi-market model of schooling, and more inegalitarian outcomes than the Nordic cluster.

Other researchers have moved the focus from education spending to participation in education and average years of education. Education spending shows annual flows, while average years of education measure the stock of accumulated education in a society (Mosher, Reference Mosher2015). Mosher argues that this measure better captures the impact of education on social welfare within a nation because it is not shaped by “the additional education produced in a year, but [rather] by all of the accumulated education that exists in society” (p. 243). Because liberal welfare regimes focus more on education relative to welfare spending as a functional substitute (or “trade-off” as described earlier), he argues that liberal countries are more comfortable with this “equality of opportunity” approach of intervening in education rather than directly redistributing within society through social spending.

Ansell and Gingrich (Reference Ansell, Gingrich, Manow, Palier and Schwander2018) argue that the dynamics of institutional environments, specifically those of wage compression and protected professions, result in the different “fortunes” in terms of occupational outcomes of university graduates. Building on their earlier research (Ansell & Gingrich, Reference Ansell and Gingrich2013), they employ these two dimensions (protected versus unprotected professions; wage compression versus wage dispersion) to advance four ideal types – credentialism, mismatch, social investment, and winner takes all – that they demonstrate map on to the four welfare democracies defined by Manow, Pailer, and Schwander (Reference Manow, Palier and Schwander2018). These ideal types help explain the unevenness of mass upskilling in various OECD countries.

Garritzmann (Reference Garritzmann2016) depicts a “four worlds of student finance” typology defined by average annual tuition costs and the “proportion of students receiving subsidies” (p. 89). Predictably, continental European countries are associated with low tuition/low subsidies; the Nordic countries with low tuition/high-subsidies; and Anglo- Saxon countries with high tuition/high-subsidies. Also included in the analyses are Asian and Latin American countries, some of which can be classified as high tuition/low subsidies. In earlier work (Garritzmann, Reference Garritzmann2015), he observes that since these patterns were established beginning in the 1950s and extending through to the 1980s, “no country has witnessed radical change in its tuition-subsidy regime” (p.141) but rather has demonstrated an even more entrenched path dependencies of these original patterns. Johnson (Reference Johnson and Jungblut2023) employs Garritzmann’s “four worlds” typology to demonstrate the extent of generosity of student support systems defined as a combination of low fees and high levels of student support. Conservative regime countries register low on both dimensions, social-democratic regimes high on student support and low on student fees, and liberal regimes high on both dimensions.

Continuing with the student funding system motif, which he states has been “largely neglected in the comparative welfare state literature,” Czarnecki (Reference Czarnecki2022) employs the concepts of decommodification versus commodification, individualisation versus familialisation, and universal versus targeted distribution of fees and benefits based on family income to offer a three-dimensional conceptualisation of student social rights. However, analyses based on this conceptualisation led him to conclude that “many countries do not fall into clearly identifiable clusters” (p. 329).

Other studies have focused on the differing impacts of education on particular life domains and quality of life, as shaped by welfare state contexts. Jongbloed and Pullman (Reference Jongbloed and Pullman2016) show that welfare regime types have an impact on the association between education and well-being, finding that decommodified institutional arrangements reduce educational inequalities while stratifying forces strengthen the education gradient in well-being. Massing and Gauly (Reference Massing and Gauly2017) ascertain that gender significantly influences participation in adult education and training and that this differs across welfare state regimes. Other studies have examined relationships between education and health (Leão et al., Reference Leão2018), health inequalities (Eikemo et al., Reference Eikemo2008), and social mobility (Beller & Hout, Reference Beller and Hout2006).

Since the early 2000s, a new wave of globalisation has had an impact on most policy dimensions that are dealt with in these theories. It can be argued that the discretion of national governments to shape social and educational policies according to their own traditions has become comparatively weaker over time. In recent decades, growing global competition and international agreements, such as the Bologna Process, have limited the power of national governments. Some argue that there has been a general convergence of social policy beyond an era of liberalisation into an era of knowledge-based growth (Hall, Reference Hall2022). Production regime theory casts light on such assumptions by illuminating these dynamics in relation to distinctive national characteristics.

Types of production regimes

Liberal market economies (LMEs) (e.g., Canada, USA, UK) are characterised by hierarchical structures and competitive market arrangements. Other characteristics include fragmented coordination and cooperation among employers, a short-term financing perspective, and a system that promotes adversarial industrial relations (Estevez-Abe, Iversen, & Soskice, Reference Estevez-Abe, Iversen, Soskice, Hall and Soskice2001; Thelen, Reference Thelen2004). Scepticism of firms in LMEs to participate in the education and training of future employees is based on concerns about whether workers will learn skills that are useful to the firm and the extent to which employees trained at one firm will be poached by firms that do not provide training. From the perspective of future workers, participation in firm-based training schemes is contingent on the belief that training will lead to paid employment. LMEs favour an approach where the demand to produce individuals who possess high skills is fuelled by intensive competition among firms (Estevez-Abe, Iversen, & Soskice, Reference Estevez-Abe, Iversen, Soskice, Hall and Soskice2001).

In contrast, coordinated market economies (CMEs) traditionally take a long-term orientation and a cooperative approach to labour management and skills training. Firms depend on and foster relationships with the formal educational system, government, unions, and other actors. These relationships result in a coordinated approach to defining core competencies and determining how to train workers. A strong vocational education sector allows for these non-market-based institutional relations between employers and educational institutions. Through the establishment of extensive networks and related exchange of information, CMEs take a “patient capital” approach to returns on training investment. CMEs are generally collaborative which is more inclined to result in labour markets that are less competitive. However, there are differences within CMEs. In countries considered conservative (e.g., Germany) from a welfare regime perspective, the goal is long term employment security and modest unemployment replacement, which, because of a tendency toward rigidity, is sometimes referred to as Eurosclerosis. Wood (Reference Wood2015) comments that such rigidity is a form of “path dependency in terms of values and culture underpinning variant forms of political settlement” (p. 139). Social democratic regimes (e.g., Sweden) have adopted a “flexicurity” approach where permitted higher levels of job turnover are redressed with retraining schemes.

Production regimes and skill development

The VoC framework brings the nexus of skills, firms, and the state to the forefront of analysis. A typology of skill that encompasses three differential forms is emphasised: firm-specific skills acquired through on-the-job training; industry-specific skills acquired through vocational and apprenticeship education; and general skills recognised across specific firms and industries (Estevez-Abe, Iversen, & Soskice, Reference Estevez-Abe, Iversen, Soskice, Hall and Soskice2001). This typology highlights the differing forms of skill investment required by employers, where and how skills are recognised, and the specificity and portability of skills. While the approach has been criticised for overlooking significant within-regime variations in skill formation (Saar & Räis, Reference Saar and Räis2017), it has become a key way of distinguishing VET systems in comparative research (Skolnik, Reference Skolnik2021).

Utilising the VoC framework to analyse skill formation, Estevez-Abe, Iversen, and Soskice (Reference Estevez-Abe, Iversen, Soskice, Hall and Soskice2001) divide the OECD countries into two opposing welfare production regimes: Anglo-Saxon and Continental European. Moreover, they subdivide the latter into a principal group and three related subgroups, which results in a total of five clusters. Anglo-Saxon countries are characterised as general skill regimes with low social protection and a general-skill biased skill formation approach. The Anglo-Saxon regime provides the largest returns to tertiary degrees and includes Australia, Canada, Ireland, New Zealand, the UK, and the USA.

They define Continental European countries as specific skill regimes, where moderate social protection is provided by the state and both general and industry-specific skills are emphasised. This regime is exemplified by Norway, France, and Finland. Close to this group, a first subgroup emerges for those with firm-specific vocational and skill training, but where employment protection is high and unemployment protection is fairly low, such as Italy and Japan. In contrast, a second subgroup, including Denmark, the Netherlands, and Switzerland, is defined by its industry-specific vocational and skill training. High unemployment protection and only moderate employment protection are typical of this group, in opposition to the previous group. The last subgroup is defined by both high unemployment and high employment protections, with unique mixes of firm- and industry-level skills, as seen in Austria, Belgium, Germany, and Sweden.

A VoC analysis of skill formation also considers the active link between models of capitalism and their corresponding welfare regimes and modes of social protection. Insecurity will result, argue Estevez-Abe, Iversen, and Soskice (Reference Estevez-Abe, Iversen, Soskice, Hall and Soskice2001), in a downturn of certain skill investment, as “in the absence of institutional interventions into workers’ payoff structure[s], general rather than asset-specific skill acquisition represents the utility-maximizing strategy” (p. 150). Employment, unemployment, and wage protection all contribute to the type of skills deemed a secure investment by an individual. The forms of social protection available influence the types of skills sought by individuals and firms and what structures of training are supported through state institutions and government policies (154). From a political economy perspective, the skill base of a population is associated with a range of outcomes. This includes linking skill to economic productivity, resources for innovation, and even economic disparity across certain forms of social segregation (Brown, Green, & Lauder, Reference Brown, Green and Lauder2001; Mayer & Solga, Reference Mayer and Solga2009).

Lauder, Brown, and Ashton (Reference Lauder, Brown and Ashton2008) critique and extend this perspective, asserting that national institutions of skill formation are undergoing fundamental changes as a result of globalisation and the growing role of multinational companies (MNCs) in skill development. However, they distinguish between skill generation and skill utilisation, highlighting that convergences and divergences in each domain may differ. They develop the concept of “skills capture” to explore the changing – and globalised – nature of how skills are determined, where these skills are developed, and what rewards are associated with these skills. They show that MNCs increasingly influence the skill formation process and use “skills capture” to build and maintain their competitiveness in a global context and posit that this leads to changes privileging diverse national movements toward LME regime arrangements in the ways that skills are constructed.

Production regime theory and education

The typology of skill formulated by the VoC perspective allows for cross-national comparisons that consider the way in which different skill regimes function. This has implications for equality, the labour market, and educational segregation, as differing skill and training systems have varying impacts on those deemed academically strong or weak and on education-to-work transitions (Estevez‐Abe, Iversen, & Soskice, Reference Estevez‐Abe, Iversen and Soskice2001). Rather than considering skill only through institutional relations, more recent work under the VoC framework attempts to untangle the social value of skill and how it is perceived to affect action. Skill, in this sense, is understood as emerging within a web of relations among firms, associations, institutions, the state, and individuals. Skills differ in their portability and atrophy rate, which affect the choices that individuals make in plotting their life courses. Thus, the VoC framework, while proposing to be a typology rather than a formulation of existing capitalism, social action, and choice, makes several inroads into what skills are and how they socially function and influence individual action.

VoC frameworks of analysis are used in educational research to understand higher education systems. For example, the work of Graf (Reference Graf2009) expands the original VoC schema in order to study education systems within liberal and coordinated market economies. Under Graf’s framework, liberal market higher education systems interact with other actors and institutions “through reactions to demand and supply conditions in competitive markets” (p. 571). Market signals have an impact on institutional change, both through direct and formal contracts and informal and arm’s length relations. Importantly, institutions and systems of higher education do not operate on a purely market model and cannot be likened to for-profit firms; rather, the way in which an institution is embedded in a particular network of relations may mirror market-like conduct.

Comparatively, higher education coordination within coordinated market economies is less reliant on market influence and other means of institutional interaction regulating systems. As Graf (Reference Graf2009) argues, “uncertainty about actors’ behaviour is reduced and the formation of credible commitments, deliberation, effective information sharing, and monitoring of behaviour is supported” (p. 571). From a VoC perspective, under mutually reinforcing forms of coordination, institutions come to complement and increase their efficiency of one another. In examining higher education systems through this lens, one can hypothesise that increased coordination can lead to a comparative institutional advantage.

Investments in human capital and firm-level staff training are central to national institutional archetypes in the VoC literature explored above. Comparative capitalism understandings of these processes have resulted in a number of country groupings (Goergen et al., Reference Goergen2012). CMEs have comparatively stronger industry-and-employer-linked vocational training systems than LMEs. Because LMEs have highly fluid labour markets and low employment protection, vocational training typically occurs in formal education institutions and focuses on generic rather than firm-specific skills. In contrast, CMEs are often typified by high job security economies where organisation-specific or firm-specific skills become more advantageous for both workers and employers. These central distinctions hold true for many countries, but some differ in important ways. For example, Sweden and Finland have weaker vocational training systems than other CMEs, and the Netherlands and Denmark have distinct “flexicurity” models, where job protection is weaker but generous social security benefits and strong continuous education compensate (Goergen et al., Reference Goergen2012). The southern European or Mediterranean countries (including Portugal, Spain, Italy, and Greece) also escape easy categorisation, as they combine characteristics of both CMEs and LMEs.

Patterns of participation in adult learning have received particular attention from the VoC perspective (Saar, Ure, & Desjardins, Reference Saar, Ure and Desjardins2013). This research highlights the concept of “institutional complementarity” to contextualise the functionality of an institution in relation to the other national institutions that coexist alongside it (Saar, Ure, & Desjardins, Reference Saar, Ure and Desjardins2013). Analyses typically begin with a two-category VoC framework to develop more detailed categories. For example, Roosmaa and Saar (Reference Roosmaa and Saar2017) delineate seven categories of countries in regards to barriers to adult education participation: social democratic, liberal, southern European, continental/corporatist, and three types of post-socialist (neoliberal, embedded neoliberal, and Balkan). Other approaches focus solely on VET systems. For example, Pilz (Reference Pilz2016) creates a new typology beginning with existing skill formation theories but adds elements specific to training systems, including characteristics related to stratification and standardisation of VET.

Recently, authors have also proposed “varieties of academic capitalism” in university higher education (Schulze-Cleven & Olson, Reference Schulze-Cleven and Olson2017). These reflect the typical “three worlds” as well as the trade-off that occurs in higher education between increasing enrolment, maintaining a high level of public subsidisation, and controlling the overall public cost (Ansell, Reference Ansell2008). University education in all groupings has experienced extensive liberalisation, which Schulze-Cleven and Olson (Reference Schulze-Cleven and Olson2017) define as “the strengthening of market principles—self-reliance, rivalry, and decentralised decision-making—in university governance” (p. 814), but that this has occurred in the form of path-dependent national trajectories.

The VoC literature also highlights the evolving nature and definitions of skills, changing patterns of higher education participation, and related roles of key stakeholders in CMEs. Over the past few decades, the transition to a knowledge society and the emergence of other confluent forces have led to institutional change in both the service and core manufacturing sectors (Diessner, Durazzi, & Hope, Reference Diessner, Durazzi and Hope2022; Durazzi, Reference Durazzi2019; Thelen, Reference Thelen2019). Depending on the CME country, these forces include some or all of the following: increasing demand by students and parents for general secondary school and university level credentials; declining demand for those trained at the vocational level to fill intermediate occupations; more managerial and professional occupations requiring university level credentials; expanding dominance of information and communications technology (ICT) in both manufacturing and knowledge-intensive sectors; increasing calls by employers for workers with higher skill levels particularly in STEM fields; and a shift in power among large employers, unions, and government regarding the specification of skill acquisition levels for workers.

The evolution of the skill/education/employment nexus has taken various forms in CME countries. In Germany, the forces above have led to a pivot away from vocational training and its signature dual system and toward higher education – that is, university education – which Durazzi and Benassi (Reference Durazzi and Benassi2020) describe as having been located “traditionally at the margins” (p. 320). They report a dramatic increase in student enrolment in the academic (Gymnasium) stream (40% in 2012) that has led to higher demand for university-level education. Expansion has occurred mainly in Fachhochschulen (universities of applied science), in conjunction with the emergence of dual study programs where employers select the students they wish to sponsor for places in specific types of degree programs (most often engineering) in institutions of the employers’ choosing (Diessner, Durazzi, & Hope, Reference Diessner, Durazzi and Hope2022; Thelen, Reference Thelen2019). Students who earn places in these “highly prized dual-study programs” (Thelen, Reference Thelen2019, p. 302) are doubly certified in that they earn both vocational certificates and bachelor’s degrees. In addition, they are paid apprenticeship wages during their studies (Thelen, Reference Thelen2019). These monetary and non-monetary efforts serve to both secure young workers to firms and to prevent poaching. As such, expansion of higher education has occurred hand-in-hand with a growing high-tech manufacturing sector (Ansell & Gingrich, Reference Ansell and Gingrich2013). However, the uptake of such programs at around 100,000 students is rather low (Diessner, Durazzi, & Hope, Reference Diessner, Durazzi and Hope2022).

A consequence of this shift away from vocational training and the dual system at the secondary level toward university education and dual study programs has also resulted in a power shift from traditional coordinating bodies toward large employers. Another consequence has been the increasing likelihood of social exclusion by those who are academically less able to gain entry to university-level studies, as the low-skilled positions they previously would have held are increasing subject to outsourcing and more precarious employment arrangements. This is in contrast to increased stabilisation of employment for employees in manufacturing with high-level skills (Diessner, Durazzi, & Hope, Reference Diessner, Durazzi and Hope2022; Durazzi & Benassi, Reference Durazzi and Benassi2020; Thelen, Reference Thelen2019).

Germany has remained closely aligned with traditional manufacturing sectors and “rather than softening the line between vocational and university tracks [it has] instead subsume[d] part of the higher education system to the firm-dominated logic of the traditional vocational system” (Thelen, Reference Thelen2019, p. 303–304). Nonetheless, the increasing presence of ICT within the German manufacturing sector has augmented demand for those with university level engineering and IT credentials (Diessner, Durazzi, & Hope, Reference Diessner, Durazzi and Hope2022). Conversely, CME countries Sweden and the Netherlands have, as Thelen (Reference Thelen2019) explains, evolved in an “apparent liberal direction” (p. 298) by developing their high-tech sectors and strongly incorporating ICT-related products into their export mix. Since the 1990s, several policy initiatives in Sweden have eliminated firm-based apprenticeships and have extended access to higher education to all secondary students. In addition, several state initiatives have been implemented to increase the digital skills of the entire population. In the Netherlands, a decline in the manufacturing sector led to a transition to the knowledge economy through the promotion of education (primarily in hogescholen or universities of applied sciences) (Durazzi, Reference Durazzi2019; Marques, Suleman, & Costa, Reference Marques, Suleman and Costa2022) that emphasises general skills conducive to employment in knowledge-intensive high-end services in business and finance. Similar to Sweden, reforms in the Netherlands have resulted in large increases in university attendance. Employment in business services, and particularly FIRE (finance, insurance, and real estate), in the Netherlands has expanded to represent over a quarter of total employment in 2015.

Several authors have called for a reconceptualisation of the traditional association of LMEs with transferable general-level skill acquisition and CMEs with the acquisition of specific skills that are less transferable. According to Ansell and Gingrich (Reference Ansell and Gingrich2013), the distinction between general and specific skills is not nuanced enough to capture the general skills held by workers employed in dynamic private sector services (e.g., FIRE services) versus those in non-dynamic and public-sector services (e.g. social services). Diessner, Durazzi, and Hope (Reference Diessner, Durazzi and Hope2022) advance the concept of skill-based liberalization. In the case of transitioning to the knowledge economy in Germany, they argue that this concept transcends the well-established approaches of dualisation and liberalisation by acknowledging both liberalisation and technological change and that institutional change takes place at both core and peripheral sectors. Durazzi (Reference Durazzi2019, Reference Durazzi2023) posits that a focus on the extent of skill breadth (narrow/broad) and combinations of skills (technical and discipline-specific to generalist) in relation to academic discipline and nature of employment sector (i.e., high-end services versus the manufacturing sector) more accurately illuminates the skill dynamics at play. Additionally, Marques, Suleman, and Costa (Reference Marques, Suleman and Costa2022) warn that while investment in higher education may run the risk of overqualification of its graduates, countries that have been successful in making the transition to a knowledge-based economy are less likely to experience this phenomenon.

By applying the VoC framework to educational systems, this “varieties of education” approach has also been used to understand social inequalities more broadly in terms of income inequality. Österman (Reference Österman2018) determined that tracking within secondary education impedes equality of educational opportunity but can also lead to higher incomes for those who graduate from vocational education, while collective wage bargaining helps generate more equal opportunities in terms of relative income in contexts without overt tracking. However, the usefulness of these distinctions when examining education systems from a comparative perspective has been questioned. Some researchers argue that “grand typologies” present important limitations when attempting to explain cross-country differences. This is due to difficulties providing a clear link between empirical results and theory, as well as the variety of specific sub-domains to which these theories have been applied (Reichart & Kaufmann-Kuchta, Reference Reichart and Kaufmann-Kuchta2021).