Understanding comprehensive school reforms: Insights from comparative-historical sociology and power resources theory

Abstract

The historical origins and development of comprehensive schooling have seldom been analyzed systematically and comparatively. However, there is a rich comparative and historically grounded literature on the development of welfare states, which focuses on many relevant policies, but ignores the education system. In particular, the power resources approach applied by many welfare state scholars has been continuously elaborated and refined in various ways. Two major comparative-historical analyses of the development of education systems, and comprehensive schooling in particular, are therefore reviewed and discussed with a view to how their insights could be enriched with knowledge drawn from welfare state literature. The article argues that, while education does constitute a separate analytical issue, scholars of comprehensive and other educational reforms could nonetheless improve their arguments by taking into account the debates and theoretical elaborations produced in the field of welfare state analysis.

Introduction

On the whole, the origins and development of Western education systems and their varying degrees of comprehensiveness ¹ have not received very much analytical attention. Historians have traced the educational history of single countries and created much interesting and useful historical narrative, but have rarely undertaken systematic comparisons, which could explain unequal development (Green, 2013: 36; but see also, Anderson, 2004; Müller et al., 1987; Ringer, 1979). While some major contributions on the origins and development of education systems have been made in the field of comparative-historical sociology (Archer, 2013; Green, 2013; Wiborg, 2009), the focus of most educational sociology has been more on the consequences of existing systems on life chances, and to a lesser degree on how these systems came into being. However, there exists a rich literature on the development of welfare states, focusing on many relevant policies such as those affecting the labor market, pensions and health care (Bowman, 2014; Huber and Stephens, 2001). A large part of this literature, is based on historically grounded, comparative analyses of power distributions, class struggles and class coalitions behind the policies in question (Esping-Andersen, 1990; Korpi, 1978, 1983). This literature has been continuously developed and the power resources approach has been elaborated and refined in many ways (Huber and Stephens, 2001; Manow and Van Kersbergen, 2007). However, the education system has been largely ignored in this literature even though, as Willemse and de Beer (2012) point out, it is a central part of the welfare state. Iversen and Stephens (2008: 602) explain this with the fact that public education predated the first social legislation, namely German sick pay legislation from 1883, and therefore has been considered to be a separate issue. The much earlier historical origins of national education systems can certainly in part explain why they have not been a “natural” topic for scholars of the welfare state. However, Iversen and Stephens (2008) hold that skills and education are closely related to issues of social protection, economic performance, distribution, and partisan politics and therefore suggest reintegrating education in comparative welfare state analysis. This certainly is a valuable aim and their article contains some important starting points for such a project, as does the work of Willemse and de Beer (2012).

However, the question in this article will not be how the education system can be integrated into welfare state analysis, but rather how power resources theory can help understand developments in the education system. Shifting focus in this way is justified with the assumption that the education system in fact does constitute a separate analytical issue, in the sense that it is an important sub-sector of the state and partly subject to its own logic. Even though education can easily be viewed as part of the welfare arrangements in a state, it clearly is more than that since it fulfills a large variety of functions. For example, education is strongly related to (un)employment, distribution, and the shape of the production system, but it is also subject to conflicting priorities of individual emancipation, ideological indoctrination, and the shaping of class, gender, and national identities. In this complex field, political development is influenced by a large number of actors and cannot be derived directly from conditions in other sectors of society (Archer, 1989). A thorough understanding of educational reforms therefore requires focusing on the struggles behind them. It also requires a thorough knowledge of the actual systems that are or have been in place. Nonetheless, while there are many justifications for the existence of a separate research field in education, it is assumed here that the comparative-historical analysis of comprehensive schooling is based on the same kind of analytical logic as much of the power resources literature, and that these two strands of literature should be in closer communication with each other. This article will therefore examine the extent to which the power resources approach applied in much welfare state analysis could inspire the comparative-historical analysis of the origins and development of comprehensive schooling.

Certainly educational politics can be analyzed with regard to many more factors than merely the degree of comprehensiveness of the school structure. However, recent data show that early tracking within schools and socioeconomic segregation between schools constitute hindrances to providing high-quality education for the population as a whole (OECD, 2010; Schleicher, 2014). As a result, seemingly outmoded debates about comprehensive school reforms have acquired renewed relevance (Wiborg, 2009: 1). The article will therefore first review two major historical-sociological, comparative works, namely Andy Green’s (2013) analysis of education and state formation and Susanne Wiborg’s (2009, 2010) study of comprehensive schooling in Europe. Green’s (2013) study does not focus exclusively on comprehensive schooling, but rather on the historical roots of national education systems and their relation to the state. Therefore, it constitutes an important theoretical backdrop not only for Wiborg’s (2009, 2010) analysis but for anyone interested in the long-term development of school systems. Both works not only represent two of the most sophisticated analyses available but also have clear overlaps with power resources literature, as will become apparent in this article. In the next step, the basic assumptions of the classic power resources approach as formulated by Korpi (1974, 1978, 1983, 1985) and others, as well as some important topics of debate in this literature, will therefore be discussed. In particular, Manow and Van Kersbergen’s (2007) criticism regarding the neglect of the role of religious cleavages for welfare state development will be presented, as well as a further theoretical elaboration of the power resources approach by Huber and Stephens (2001). The focus rests on those debates and theoretical concepts, which could be of relevance also for the analysis of comprehensive school reforms. In the last part of the paper, a possible research agenda for further comparative-historical analyses of comprehensive school reforms will be outlined. It is suggested that the analysis of the historical development of comprehensive schooling can in many ways be inspired by power resources literature, but also that it should avoid some of the pitfalls that have been identified in debates on welfare state development.

Andy Green’s theory of education and state formation

Andy Green’s (2013) study of the origins of national education systems and the role of the state in this process was first published in 1990. Green focuses on the historical development of education systems in England, France, Prussia and the United States of America, and has added an additional chapter on East Asian education systems in the 2013 edition. In the new foreword, he also makes it clear that his theory is meant to apply to the formative stage of states and states’ education systems more than to the later development in “mature democracies today” (Green, 2013: 3–4). While the latter still have to “address issues of cultural reproduction and identity formation” (Green, 2013: 3), ideological factors related to national identity are said to have decreased in importance relative to the rationale of economic competitiveness. To what extent Green’s theory is applicable to current educational reforms in Europe could be a matter of debate. Green himself limits his analysis to “early periods of nation-building”, as well as “periods of political reconstruction after crises”, for example after the Second World War (Green, 2013: 4).

Green (2013) defines national education systems as “systems of formal schooling at least partly funded and supervised by the state which provided universal education for all children of school age in a given nation” (Green, 2013: 297). Green’s main hypothesis is that the uneven development of national education systems during the 19th century was related to the different ways in which modern states were built and consolidated. State formation, for Green (2013) “includes not only the construction of the political and administrative apparatus of government […], but also the formation of ideologies and collective beliefs which legitimate state power and underpin concepts of nationhood” (Green, 2013: 83). Where state formation occurred as an “intense” process, national education systems were formed earlier and faster, because their development was linked directly to the nature of the state.

The major impetus for the creation of national education systems lay in the need to provide the state with trained administrators, engineers and military personnel; to spread dominant national cultures and inculcate popular ideologies of nationhood; and so to forge the political and cultural unity of burgeoning nation states and cement the ideological hegemony of their dominant classes. (Green, 2013: 298)

Green (2013: 299) names several factors that contributed to this intense process of state formation, such as external military threats or territorial conflicts, major internal transformations resulting from revolution or newly achieved national independence, and economic underdevelopment, which in some cases led to far-reaching state reforms. Green’s prime opposite example is England, where none of these factors was present during the 18th and 19th centuries and which therefore: “never experienced the kind of deliberate and concerted process of state formation that occurred in the other countries and which was the driving force behind their educational reforms” (Green, 2013: 301). As a result, the English national education system developed almost a century later than its counterparts in other Western states and public education has remained weak and unsystematic up to the present day (Green, 2013: 297, 301ff.).

To support his main hypothesis, Green (2013: 82ff.) builds on theories by Karl Marx and Friedrich Engels, and in particular on later elaborations by Antonio Gramsci. While Green (2013: 87) seems to agree with Marx that the state can never be “‘neutral’ or independent”, but is always involved in the reproduction of the mode of production and resulting inequalities, he also makes it clear that he does not agree with reductionist readings of Marx’s writings on the state. In particular, he rejects the idea that the nature of the state is determined exclusively by the economic base. Green (2013: 92ff.) instead points to parts of Marx’s (1907) and Engels’ (1972) works, which show that their conception of the state is not that simplistic. Here, the state is acknowledged as a “relatively autonomous force”, influenced by complex relations to and between classes, class fractions, and the economy (Green, 2013: 93). While the state still serves the long-term aim of dominant classes to secure the survival of the mode of production, it can at times be dominated by varying classes: “or, indeed, at certain exceptional moments, may appear to override all particular class interests” (Green, 2013: 93). In particular, the state can acquire a sort of mediating role, when the force of competing classes is nearly equal (Engels, 1972: 231). A final important idea, which Green adopts from Marx and Engels, is that dominant classes will try to use the state to present their class interests as universal interests, thereby achieving hegemony over subordinate groups (Green, 2013: 94).

Gramsci’s (1972) writings on the state and hegemony elaborate on these ideas. Because consent is such an important part of hegemony, schools are explicitly mentioned by Gramsci (1972) as important state organs in which the struggle for hegemony is constantly going on. In fact, Gramsci’s (1972) whole work is neatly summarized by Green (2013) as “a sustained meditation on a single, but double-sided, theme: the way in which the dominant and the subordinate classes seek to educate society into their own conceptions of the world” (Green, 2013: 103). The state, then, is the site where unstable equilibria constantly result from this struggle. And while the dominant group may for the most part keep the upper hand, compromises and alignments of interests up to a certain point will be inevitable (Gramsci, 1972: 182).

Green (2013: 35ff.) also discusses and criticizes some alternative explanations and theories of educational development. For example, functionalist theories linking economic skill requirements with educational reform are dismantled by Green’s (2013) observation that the patterns of industrialization and educational development are not strongly correlated at all. The same holds for patterns of democratization, as well as for the predominance of Protestant religion. Again, England can serve as a prime example, since it was an educational latecomer despite its Protestant religion, early industrialization and democratization. England and the urban areas of the north-eastern US states are also at the center of theories focusing on the relationship between educational development and urbanization, proletarianization and changing family life (Bowles and Gintis, 1976). Green (2013: 55ff.) does not refute that these processes were linked to educational reform, because they disrupted traditional educational processes and created a need for social control, but maintains that these theories fail to explain educational development in rural, pre-proletarian areas, such as the continental European states.

Green points out that his elaborations do not constitute “a theory of educational change capable of explaining uneven national development” (Green, 2013: 104) and that it is doubtful whether a general theory of state formation capable of explaining all national and regional variations is even possible. While this is correct, his Marxist framework could possibly be combined with some of the theoretical elaborations in power resources literature, which will be discussed below. This might help to bridge the gap between theory and empirical analysis.

Susanne Wiborg’s (2009, 2010) analysis of the history of comprehensive schooling in Europe

A comparatively recent contribution to the analysis of educational development is Susanne Wiborg’s (2009, 2010) study on the history of comprehensive schooling in Europe. Wiborg’s (2009) aim is to make use of the comparative-historical method advocated by Ragin (1987) and Skocpol and Somers (1980), in order to explain why radically comprehensive schooling with mixed-ability classes from grades 1 to 9/10 developed in the three Scandinavian countries of Denmark, Norway and Sweden, but not in (West) Germany and only partially in England. To understand these different historical outcomes, she undertakes a sophisticated comparative-historical analysis based on four main hypotheses, which are related to: “(1) the type of state formation undergone in the nineteenth century; (2) the nature of class structure that evolved in the nineteenth century; (3) the nature of liberal politics and of alliances forged between Liberals and other parties; and (4) the nature of social democracy and its political alliances” (Wiborg, 2009: 14).

The first part of her study builds heavily on Green (2013). Wiborg (2009: 19ff.) discusses the development of the Scandinavian, German and English nation states with a view to how the state-building processes in each country might have been related to the development of a national education system, and in particular might have laid the ground for a later systematization and comprehensivization of it. She finds clear evidence for Green’s (2013) theory that intense processes of state-building were related to educational reforms. In the Scandinavian countries and among the German states, especially Prussia, state involvement in secondary education was first expanded under absolute monarchies, due to the growing need for civil servants. Gradually elementary schooling also became subject to state policy. During the first half of the 19th century, the Scandinavian nation states and Prussia then shaped and consolidated their national education systems. The power struggle between Denmark, Norway and Sweden, which went on over several decades, as well as the Napoleonic wars, also seem to have given an impetus to this development. However, the Napoleonic wars can be more clearly linked to educational reforms in Prussia than in the Scandinavian countries, because the Prussian defeat was experienced as particularly traumatic by national elites and the policy responses, including educational reforms, were directly related to it. As in Green’s (2013) study, England represents the exception also for Wiborg (2009: 19ff.) in the sense that its state-building phase had already occurred at an earlier date and the state now had a liberal shape, which was a hindrance for educational reforms. Wiborg’s (2009: 46ff.) most interesting conclusion regarding the state formation hypothesis is, however, related to the fact that, while Prussia had at least as good conditions as the Scandinavian countries for creating an integrated, comprehensive ladder system in education, this did not happen. Instead, the Prussian system remained strongly vertically stratified, with different secondary school types, as well as private feeder preparatory schools alongside the public elementary schools. Wiborg (2009) therefore concludes that, while the “state formation argument correctly identifies the development of state education as a necessary and contributory condition for the linking of different schools into a coherent system” (Wiborg, 2009: 47), it cannot explain why the levels of vertical integration differ so greatly between the countries in her study.

In the next part of her study Wiborg (2009: 49ff.) therefore analyzes how different class structures might have contributed to the development of varying degrees of vertical educational integration during the 19th century. The main hypothesis here is that “the relative homogeneity of Scandinavian societies was propitious for the development of a ladder system of education” (Wiborg, 2009: 215). Wiborg (2009: 49ff., 2009: 215ff.) points out that the Scandinavian states were characterized by comparatively weak bourgeois and noble classes and by a relatively homogeneous and numerous peasantry, which managed to gain political rights and became an independent class as early as the late 18th century. In all three Scandinavian countries, the peasants acquired large parts of the land and developed into a rural middle class, which was represented by agricultural political parties and organizations. In Prussia, on the other hand, the Junker class of large landowners remained powerful well into the inter-war period of the 20th century. German peasants had long been subject to much harsher exploitation than Scandinavian peasants and Prussian agricultural reforms turned many of them into landless agricultural laborers, while only the wealthier farmers managed to hold on to their land. As a result, peasants’ land ownership decreased, peasants were divided politically, and never formed an independent political party. Instead, they supported the conservative parties of the Junkers. In England, finally, the more advanced degree of industrialization meant that the peasantry was smaller, a large urban working class existed, and the ruling class was composed not only of land owners but just as much of business entrepreneurs. On the background of this analysis of the class structure, Wiborg (2009: 64ff.) then analyzes the social composition of the students enrolled in secondary schooling. Not surprisingly, she finds that the elite grammar schools in England enrolled the smallest percentage of working class children, while the Scandinavian secondary schools admitted at least a relevant share of children from the peasantry. Also, by the late 19th century, the Scandinavian elementary schools enrolled the majority of children, including those from the bourgeoisie, while in England elementary schooling remained largely segregated on the basis of class. However, Prussia again presents a challenge, because its secondary schools enrolled almost as many children from the peasantry as the Scandinavian secondary schools, even though class-based elementary schooling persisted. In other words, Prussian conditions in the 19th century do not seem to have been completely unfavorable to the development of a ladder system of education. While it is clear that the comparatively egalitarian class structure in the Scandinavian countries contributed to some extent to the development of comprehensive schooling, Wiborg (2009: 218) therefore concludes that this in itself does not constitute a sufficient condition.

Wiborg (2009: 75ff.) then moves on to an analysis of the role of liberal politics in the creation of comprehensive education. She points out that the Scandinavian liberal parties were all involved in the creation of middle schools between 1869 and 1905, which constituted an important step toward a comprehensive ladder system. In other areas, the Scandinavian Liberals also laid the ground for later welfare state development. However, Scandinavian social liberalism differed significantly from British and German liberalism regarding its class base and ideology. In Scandinavia the liberal parties represented not mainly the urban but also the rural middle classes who had an interest in educational policies that would ameliorate their standing. While they did struggle with the urban-rural cleavage, as well as with religious cleavages, these did not hamper their politics to the same degree as the massive religious divide hampered German liberal politics. Because the peasants supported the Conservatives and the Catholics founded their own Catholic Center party, German Liberals did not manage to muster support beyond the urban, Protestant middle class. Thus, they could not contribute to significant educational reforms, even if they had wanted to. In England, on the other hand, the strong bourgeois liberalism was of a very different kind than Scandinavian social liberalism, since the manufacturers, merchants, and landed gentry represented by it had no interest in more extensive state intervention, whether in education or other areas.

Finally, Wiborg discusses her most important hypothesis, namely that “it was ultimately the nature and strength of social democracy that explains the divergent development of comprehensive education in Scandinavia, on one hand, and Germany and England, on the other” (Wiborg, 2009: 231, 2009: 127ff.). Drawing on Esping-Andersen (1985), she points out that the main reason for social democratic success in Scandinavia lay in the ability of social democratic parties to forge alliances, first with the liberal peasantry and later with the emerging white-collar middle class. These alliances were necessary, because social democracy could not have mustered mass support had it drawn exclusively on the comparatively small working class. In all three Scandinavian countries, the most important steps toward fully comprehensive schooling were taken under social democratic rule. Comprehensive schooling initially lasted seven, then nine and finally ten years. In Britain, on the other hand, the Labour Party was rooted in the segmented trade union movement, and did not endorse a clearly socialist program. Even though it implemented a few important educational reforms, the Labour Party never seriously attempted to create a fully comprehensive system, for example by abolishing private schools. Also, Labour could not build on cross-class alliances to the same degree as Scandinavian social democrats could and did not achieve the same kind of ideological hegemony.

Furthermore, Wiborg holds that “[t]he absence of a powerful social democracy in Germany, at both the Federal and Länder levels, provides the key to understanding why the tripartite system was never reorganized on comprehensive lines” (Wiborg 2010: 554). In particular, Wiborg (2009: 139ff., 2010) is of the opinion that the German Social Democratic Party (SPD) held on to Marxist ideology for too long and refused to build alliances, for example with the farmers. Wiborg (2009: 140) also mentions the great influence of August Bebel, the SPD’s early leader, on the development of the radical political programs of the party and generally portrays the SPD as a haven of revolutionary ideas, even as far as the SPD of the inter-war and post-war years is concerned. However, Wiborg (2009: 140) also discusses a few of the conditions which made alliances between social democracy and other groups much more difficult to achieve in Germany. For example, the SPD faced massive political suppression by the Socialist Laws from 1878 to 1890. Also, the German agrarian interest lobbies were strongly antisocialist. In general, the traditional support of the conservatives, the Catholics and the Junkers by the peasants, the lower middle classes and the cultural elites could not easily be broken. Despite this, the SPD managed to push through an important comprehensive school reform in 1920, which instituted the four-year comprehensive primary school. Implementation of this reform was hampered by conservatives’ resistance, but was finally achieved (Wiborg, 2009: 157). Nonetheless, Wiborg (2009) is of the opinion that “[f]ew political parties have failed so completely in the achievement of their manifest destiny as the Social Democrats of the Weimar Republic” (Wiborg, 2009: 142). Wiborg (2010: 541ff.) also discusses the role of the German educated elite (Bildungsbürgertum) and agrees with authors such as Hahn (1998) that its conservatism was an important factor. However, she points out that other countries also had powerful conservative upper classes and nevertheless instituted comprehensive education. Therefore, she considers the existence of a powerful conservative elite an important, but by itself not a sufficient, explaining factor. While this is true, it does not justify placing this much weight on the weakness of German social democracy and abstaining from a more precise comparative analysis of the role of the ruling classes and the conservative middle classes and, in general, the power balance in society. It will be discussed below to what extent insights from power resources theory could help to further develop the analysis of the German, but possibly also the Scandinavian, case.

Power resources theory…

The origins of power resources theory are mainly associated with the names of Esping-Andersen (1985, 1990), Korpi (1974, 1985, 1978, 1983), and Stephens (1979). Broadly, these scholars argue that the different forms of welfare state development, and social change in general, result from the varying distribution of power resources between socioeconomic classes in different countries. Korpi defines power resources “as the attributes (capacities or means) of actors (individuals or collectivities), which enable them to reward or to punish other actors” (Korpi, 1985: 33). This means that power resources are defined in a relational way and are relevant even when not activated. In the production system, capital and labor power are considered the two main types of power resources (Korpi, 2006: 172). Korpi (1985) distinguishes between many dimensions in which power resources may vary, but the main difference between these two central types is that economic assets can be concentrated quite easily, while labor power cannot, because it cannot be divested from individual human beings (Korpi, 2006: 172). Because of this, socioeconomic classes possessing only labor power will be dependent on collective action. But of course collective action can also turn capital into a more efficient power resource (Korpi, 2006: 172). All in all, Korpi holds that “in capitalist democracies, business interests and employers generally have greater power resources than employees” (Korpi, 1985: 37). This is also related to the different distribution of risks between socioeconomic classes. Most risks, including “those associated with aging, illness, work accidents, unemployment, poverty, and rearing of children” (Korpi, 2006: 173), are distributed unequally, as Korpi (2006) correctly points out. Resources and risks are therefore negatively correlated (Korpi, 2006: 173). This makes the distributive struggle between classes over the surplus in society all the more relevant (Korpi, 1974: 1570).

Following on from this, a central assumption of the power resources approach is that: “employers and other interest groups that control major economic resources are likely to prefer to situate distributive processes in the context of markets, where economic assets constitute strategic resources and […] tend to outflank labor power” (Korpi, 2006: 173). In response, those who have no large amounts of capital at their disposal need to organize themselves into political parties and unions, which aim at influencing the outcome of these market processes. Also, these parties and unions have tried to remove some activities and policies from the market altogether, in order to achieve a certain degree of social citizenship and decommodification (Esping-Andersen, 1990: 35ff.; Korpi, 2006: 173).

With respect to the development of welfare states, the power resources approach assumes that the first laws on social insurance, which were passed in Europe in the late 19th century, were mainly “responses by state elites to the ‘Worker Question’” (Korpi, 2006: 175), which had been put on the agenda by the rising labor movement. However, the specific form which welfare state development took is attributed first and foremost to variations in the structure of class coalitions and to the balance of organizational power between classes and class fractions in different countries (Esping-Andersen, 1990; Huber and Stephens, 2001). It is assumed that “political parties perform the crucial mediating role” (Huber and Stephens, 2001: 17) with respect to the political articulation of class interests. In this context, it is important to note that Korpi assumes that the state is not a constant, but that “the extent of bias in the functioning of the state can vary considerably as a reflection of the distribution of power resources in these societies and thus that politics can be expected to matter; e.g., for the distributive processes in society” (Korpi, 1985: 39). In other words, Korpi (1985: 39) agrees with Green (2013: 87ff.) that functionalist and reductionist Marxist analyses of the state are too simplistic. He also points out that political democracy has led to rising tensions between market and state, because power resources in the market are more unequally distributed than political power resources (Korpi, 1985: 39).

However, besides the economic cleavage, cross-cutting cleavages such as religious, urban-rural, and gender divisions exist, which complicate the analysis considerably. For example, the role of Christian democracy and Catholicism has been a longstanding subject of debate within power resources literature. Clearly, the historical development of some of the continental welfare states, such as Germany, is strongly related to the strength of Christian democratic parties and the Catholic Church (Esping-Andersen, 1990; Huber and Stephens, 2001: 16ff.; Manow and Van Kersbergen, 2007). Because Christian democratic parties usually have a cross-class base, they are partly in competition with social democratic parties, which according to Huber and Stephens (2001: 16) has pushed them to support the introduction of more generous welfare policies than they would otherwise have pursued. Stephens (1979: 100, cited in Manow and Van Kersbergen, 2007: 1) has also argued that political Catholicism has contributed to the development of generous, but less distributive, welfare policies in what Esping-Andersen (1990) later termed the corporatist welfare states, because of anti-capitalist elements in Catholic ideology.

Manow and Van Kersbergen (2007) criticize this interpretation:

because it has fairly exclusively, but mistakenly, focused on the role of political Catholicism in the development of social protection systems, wrongly interpreted or simply ignored the role of Protestantism, failed to differentiate between different strands of Protestantism, and put an undue emphasis on the impact of religious ideas on welfare state institutions. (Manow and Van Kersbergen, 2007: 34)

While they agree with Esping-Andersen (1990: 30) that the key to understanding the varying historical development lies in understanding what kind of class coalition existed in a country, they believe that a systematic explanation of why the middle classes sided with social democracy in some countries, but with conservative parties in others, is lacking. To ameliorate this, Manow and Van Kersbergen (2007: 16ff.) suggest building on Iversen and Soskice’s (2006) model, in which the key variable is the structure of the electoral system, and combining this with a Rokkanian (2000) analysis. In their view, countries with majoritarian electoral systems tend to develop a two-party structure, which is based on the economic capital-labor cleavage, while other cleavages tend not to become equally politicized. In countries with proportional representation, however, there is room for a larger number of parties and other cleavages can become apparent in the party structure. In the Scandinavian countries this applies especially to the rural-urban cleavage and the development of agrarian parties, which then become the main coalition partners of social democracy. The religious cleavage, however, was not equally relevant because of the high religious homogeneity of the population and the existence of Lutheran state-churches, which did not feel as threatened by state expansion as the Catholic Church. In the continental states, on the other hand, massive state-church conflicts about, among others, education and poor relief took place during the late 19th and early 20th centuries, which laid the ground for Catholic and Christian democratic parties. In other words, here it was the religious cleavage, which became the second major cleavage expressed in politics, while the rural-urban cleavage remained latent (Manow and Van Kersbergen, 2007: 22). This analysis also constitutes a useful starting point for further research regarding educational reforms. It should, however, be remembered that Rokkan (2000: 360) already pointed out that one should not generally treat party structures as dependent and electoral systems as independent variables, since the rules of the electoral system are also a result of earlier political alliances and cleavages. His hypothesis is that proportional representation functioned mainly as a safety measure for the older, established parties in countries where these did not manage to form strong coalitions against the rising labor movement (Rokkan, 2000: 365). In any case, the temporal development of various political cleavages in different nation states also needs to be taken into account when analyzing the historical development of education systems (Rokkan, 2000).

Finally, an elaboration of the power resources approach developed by Huber and Stephens (2001) includes five concepts, which might be useful for the analysis of educational reforms. Firstly, they draw on Immergut (1992), who has pointed out that “veto points” in the political institutions of a country, such as electoral rules, second chambers, referenda, and so on, affect the policy-making process, because they offer varying possibilities for interest groups to interfere. Secondly, they define a “policy ratchet effect”, which refers to the observation that secular conservative parties have often accepted innovations which had been introduced by social democratic or Christian democratic governments, because they were popular among the mass public. Once instituted, many reforms changed the preferences of the population and thereby shifted the political center of gravity, making it more difficult for subsequent governments to reverse the reforms (Huber and Stephens, 2001: 28). Thirdly, Huber and Stephens (2001) point to what they call “structural limitation”. They argue that policy options and preferences of societal actors at any given time are conditioned by the constellation of power in a country, because this determines the chances of any proposition being realized: “A whole range of policy alternatives is ruled out by power relations within the society; thus certain policy alternatives that societal actors might otherwise find attractive will not even be considered by them” (Huber and Stephens, 2001: 29).

Fourthly, they discuss the effect of “regime legacies” and assume that legacies of previous periods affect the universe of actors by strengthening or weakening some of them, as well as the distribution of preferences by foreclosing and opening political opportunities. Huber and Stephens (2001: 29ff.) apply this idea mainly to the relationship between the welfare state and production regimes and mention the elimination of a large low-wage sector in the Scandinavian countries as an example. Finally, Huber and Stephens (2001: 29) discuss the importance of the “ideological hegemony” enjoyed by different social classes or movements, which may in part be a result of the previously mentioned mechanisms, but “cannot be read off existing policy arrangements, societal power balances, and production regimes on a one-to-one basis” (Huber and Stephens, 2001: 29). Here, we end up again with the insights produced by Green (2013), who argues convincingly that struggles for ideological hegemony are at the center of educational politics, because education has such a direct function in producing hegemony in the first place. What is taught in school, how it is taught, and to whom, is of relevance for all societal actors struggling for political influence. In the words of Huber and Stephens: “to the extent that progressive social and political movements can change political consciousness, they are very likely to have effects on the posture of centrist and conservative parties and thus to move the political center of gravity in society” (Huber and Stephens, 2001: 30).

…and the analysis of comprehensive school reforms

How could further research on comprehensive schooling be informed or inspired by these debates and concepts? In the following, a number of issues related to the analysis of comprehensive school reforms will be discussed, which would merit further exploration. First, it will be shown how concepts from Huber and Stephens (2001) could possibly be applied to the analysis of educational politics. In a second step, some arguments made by Wiborg (2009, 2010), in particular regarding the German case, will be critically discussed from a power resources perspective. Finally, it will be shown that building on Green’s (2013) framework in addition to insights from power resources theory could be a useful starting point for further analyses of hegemonic struggles and political cleavages impacting on educational development. The general aim is to show that future comparative-historical analyses of comprehensive school reforms would profit from taking debates and concepts from the power resources literature more extensively into account.

As discussed above, Huber and Stephens (2001) argue that the pattern of welfare state development is ultimately determined by the long-term pattern of partisan government. The argument is based on the assumption that class and gender interests are socially constructed and that changes in the consciousness and preferences of the population are closely related to changes in power relations (Huber and Stephens, 2001: 28). As far as comprehensive schooling is concerned, it would be interesting to explore in more detail the question to what extent the period of social democratic political hegemony in Scandinavia has lastingly influenced the beliefs and preferences of the population, including those of conservative politicians. Today, it is hard to imagine that Scandinavian conservative parties would directly suggest the introduction of a German-type, segregated school system. At first glance, it therefore seems that comprehensive schooling in the Scandinavian countries is based on a rather broad ideological consensus among different social strata and political camps. On the other hand, Norwegian conservatives support experiments with early tracking in Oslo, while Swedish conservatives have implemented far-reaching, market-based reforms in the public school system, which have led to increased social and ethnic segregation (Hjort and Panican, 2014; Wilden and Juven, 2013). In other words, there is a possibility that the former hegemonic equilibrium in this field is becoming more unstable. This might also be related to increasing Europeanization and internationalization of educational politics, which might impact on national power distributions in various ways (Busemeyer and Trampusch, 2011: 420ff.). A related issue is the question to what extent previous educational politics, including the introduction of comprehensive schooling, might make up an essential part of “regime legacies” as well (Huber and Stephens, 2001). It can be assumed that there exists a relationship between the high-wage, social democratic welfare state-production regime and the comparatively compressed skill distribution of the Scandinavian countries, as well as their comparatively high population share of tertiary educational attainments and low share with only lower secondary educational attainments (EUROSTAT, 2015; OECD, 2010). When the average educational level of the population is already quite high, some policies, for example aimed at creating a larger low-wage sector, might be harder to put into practice. To what extent comprehensive schooling is related to a high-road production regime would therefore be an interesting research question.

Another issue to explore comparatively is how the shape of the political institutions—what Huber and Stephens (2001) and Immergut (1992) refer to as “veto points”—has impacted on comprehensive school reforms. In this context, it is interesting to note that the German federal system has been blamed for the “stickiness” of German educational institutions (Ertl and Philipps, 2000; Hahn, 1998). Wiborg, however, points to a study conducted by Erk (2003), which reveals that German federalism has strong tendencies to develop “unitary characteristics, which challenges the argument regarding the debilitating nature of federalism in education” (Wiborg, 2010: 545). Standardization in the German educational system is high, despite federalism, and federalism might even contribute to reducing conflicts. German federalism can therefore probably not account for the fact that a consensus for comprehensivization was never reached. Nonetheless, analyses of educational politics should take the existence or non-existence of veto points into account as a condition for how political cleavages and unequal power resources can be given expression.

Also, the debates within the power resources literature illustrate that the extraordinary focus on social democracy, which is present in the early writings of the protagonists of the power resources approach (Esping-Andersen, 1985; Korpi, 1978, 1983) as well as in Wiborg’s (2009, 2010) explanation of the history of comprehensive schooling in Europe, can be criticized. In order to really understand the balance of power, it is necessary to assign just as much importance to the analysis of the role of other actors and cleavages. Even though Wiborg (2009, 2010) assigns a large and important part of her analysis to the role of liberal politics, she refrains from a similarly thorough comparative analysis of conservative educational politics. Here, many issues remain to be explored. It is, for example, an open question why Scandinavian conservative parties did not fight social democracy’s comprehensive school reforms harder and did not attempt to reverse them when they came to power in the second half of the 20th century. It is possible that “policy ratchet effects” and “structural limitation” may have played a role here (Huber and Stephens, 2001). In general, it seems that power resources in Scandinavia are and have been distributed quite differently from power resources on the continent, and that the position of business and conservative interests in Scandinavia is weaker (Korpi, 2006). There seem to be other political cleavages in play, such as the rural-urban cleavage, which make alliances among the non-left parties more difficult in Scandinavia (Rokkan, 2000). Presumably, Scandinavian conservative parties’ repeated acquiescence with policies conservative parties in other European countries would have found completely unacceptable is related to this also in the area of education (Huber and Stephens, 2001: 29). However, the specific role Scandinavian conservative parties have played in educational politics, and the existence or non-existence of alliances between these parties and liberal, agrarian, or Protestant parties, should be analyzed in more detail. This would allow a better understanding of the power balance between conservatives and social democracy in the field of education.

As far as Germany’s educational history is concerned, the application of power resources theory to this case might indicate that Wiborg’s (2009, 2010) analysis is too focused on “professions of ideology” and too little on: “parties’ strategic choices” (Esping-Andersen, 1985: 6). Wiborg (2009, 2010) builds a large part of her explanation for German educational development on the claim that German social democrats made the “wrong” ideological choices. According to her, Scandinavian social democrats moderated their views much earlier and generally conducted more pragmatic politics than German social democrats. However, while it is true that the early programs of the SPD were politically radical and inspired by Marxism, the same holds for the early programs of the Scandinavian social democratic parties, which were in part inspired by the German Erfurt program (Socialdemokraterna and Arbetarrörelsens arkiv och bibliotek, 2001: 11ff.). With regard to later historical periods, it is, furthermore, not convincing that the actual politics of the SPD should have been more radical or less pragmatic than the politics of Scandinavian social democrats, even if it is true that the tradition of “scientific socialism” stood stronger in Germany (Wiborg, 2009: 142). There was no lack of reformism in the SPD in the early years and even less after the two world wars (Abendroth, 1964; Walter, 2011). If anything, the problem was rather that the SPD backed down and adapted to conservative rule by moderating its views and offering itself as a subordinated “assistant” (Abendroth, 1964: 54). The fact that the Norwegian social democratic party was a member of Comintern at a time when the German SPD had already voted for the war credits of the First World War and entered its first governing coalition should illustrate that Wiborg’s (2009, 2010) analysis of the German case is incomplete.

Doubtlessly, Esping-Andersen’s (1985) and Korpi’s (1978, 1983) work support the view that Scandinavian social democrats maneuvered intelligently within the Scandinavian political and social surroundings and integrated different interests into a hegemonic project, which also included comprehensive school reforms. But this should not be regarded as being merely a matter of choosing the right ideology, for conditions in Scandinavia were very different from German conditions. In the face of powerful adversaries, it was quite difficult to unite the German labor movement and arguably this took up so much time that there was no surplus of resources left, which could have been used on breaking conservative hegemony among other groups (Bebel, 1968; Walter, 2011: 24ff.). Later, the Christian Democratic Union (CDU) became the most powerful West German party. Only twice after 1949 did the SPD receive more votes in national elections than the Christian democrats: in 1972 and in 1998. By then, the SPD had thoroughly abandoned Marxism, but this did not help when they attempted to introduce comprehensive school reforms in the late 1960s and 1970s (Wiborg, 2009: 198ff.). Again, reforms were obstructed by the powerful conservative camp. This was also related to the anti-communism of West German conservatives, who considered comprehensive schooling “communist”, since it had been implemented in East Germany—an important factor that is missing in Wiborg’s analysis (Herrlitz et al., 2009: 197ff.). This is not to say that the SPD could not have done more to build alliances and could not have tried harder to become a formative agent (Abendroth, 1964; Walter, 2011). But it seems fair to assume that the much stronger opposition (West) German social democrats faced, as well as the lack of possible allies, is in large part responsible for their failure to build powerful alliances and to implement educational (and other) reforms (Walter, 2011: 36ff., 2011: 51ff.). Therefore, Wiborg (2009) should perhaps have added a chapter on conservative politics to her analysis. Clearly, the different strength of conservative and bourgeois organizations, the political positioning of cultural elites, as well as the role of political parties based on religious affiliation, should make up an important part of any ultimate explanation for diverging development in educational politics—not only as conditions for social democracy or liberalism but as explanatory factors in their own right.

In general, it would clearly be valuable to continue the comparative-historical analysis of the origin and development of national education systems with insights from power resources theory in mind. Green’s (2013) analysis presents a useful starting point, but does not cover many of the European countries, including the Scandinavian ones, which are particularly important for research on comprehensive education. Also, it could be developed further by analyzing in detail what kind of power resources different actors had control over and how exactly social cleavages and potential (class) coalitions played out and affected educational politics. A detailed analysis of the distribution of power resources, such as financial resources or membership numbers, would also make comparisons between countries more meaningful. Green’s (2013) focus on the period of state formation, and the varying forms of hegemony which resulted from this, is however supported by the analyses produced in power resources literature, in particular elaborations by Manow and Van Kersbergen (2007), which in turn are inspired by Rokkan (2000). It seems that the political development in the 19th and early 20th centuries in many respects set the course for later welfare state reforms, including comprehensive school reforms. In this early period burgeoning nation states had to fight over hegemony in education with the church. On the continent this state-church cleavage led to the advent of religious parties defending traditional church powers and opposing the introduction of secularized national education systems. These parties also managed to gain the loyalty of a relevant share of the working class (Rokkan, 2000: 345ff.). In Scandinavia, on the other hand, a number of Protestant religious parties exist. However, they did not oppose educational reforms per se, but were mainly concerned about defying urban secularism (Rokkan, 2000: 345ff.). It would be a worthwhile research enterprise to analyze in more detail why state-church conflicts in Scandinavia and on the continent took such different shapes and led to such different results with regard to educational politics.

As far as later periods are concerned, it would also be useful to examine how the debates about national education systems, and comprehensive schooling in particular, were intertwined not only with the capital-labor cleavage but also with other political cleavages which are discussed in power resources literature. The religious dimension was clearly still relevant, as can be seen for example from massive conflicts about separate schooling based on denomination in Germany (Herrlitz et al., 2009: 126f; Wiborg, 2009: 158). Whether a nation state has a homogeneous or mixed population in terms of denomination is, in other words, a relevant factor (Rokkan, 2000). It is also probable that the urban-rural cleavage played a role in the debates on comprehensive education in Scandinavia. In the decades after the Second World War, the rural population represented by the Scandinavian agrarian, and in some cases liberal or Protestant parties, clearly voiced an interest in better educational provision in the countryside and, on the other hand, wanted to avoid too extensive centralization of schools as this was associated with long journeys to school and a weakening of rural communities. Both factors can be assumed to have facilitated comprehensive school reforms, such as the introduction of the Norwegian comprehensive middle school (ungdomsskole), since it is easier to uphold high-quality countryside schools when no parallel school types exist (Telhaug and Mediås, 2003: 167). It remains an open question to what extent comparable rural interests were expressed in countries such as Germany and how German conservatives in the period following the Second World War nonetheless seemingly not only managed to integrate the interests of the rural population, but even turned the field of comprehensive education into a much more contested battleground than it was in post-war Scandinavia (Herrlitz et al., 2009, Von Friedeburg, 1992). It should also be analyzed in more detail to what degree the cleavage within the labor movement between communist and socialist currents, and later the division of the German nation state, impacted on West German educational development (Rokkan, 2000: 374ff.). A possible hypothesis is that the urban-rural cleavage facilitated educational reform coalitions between social democracy and the middle classes in Scandinavia, while the state-church, the denominational and the national and systemic cleavages made such alliances difficult in (West) Germany and strengthened the conservative camp (Manow and Van Kersbergen, 2007).

Finally, it could be assumed that debates about comprehensive education also to some degree were related to debates about coeducation of boys and girls, as well as conflicts of interest between male and female teachers, and that a political cleavage based on gender thereby also played a role. Again, this possible relationship has not been explored in much detail in comparative-historical analyses of educational politics.

Conclusion

The development of education systems is often taken for granted as being some kind of “automatic” functional response to changes in the economic and social structure (Archer, 1989). Due to this assumption, the political and social processes and conflicts which “translate” power structures in society into school structures and curricula are not studied enough (Sakslind 2002: 128). Only a few major works in the field of comparative-historical sociology try to amend this lack of analysis, some of which have been discussed here (Archer, 2013; Green, 2013; Wiborg, 2009). The power resources approach, however, has produced a large number of comparative-historical studies and debates regarding the development of the modern welfare state, while largely ignoring the education system. In this literature, there is an emphasis on class struggles, class coalitions, and power distribution, which is lacking in functionalist accounts and which could inspire the further analysis of educational reforms. However, one major criticism of the power resources approach, namely that it has tended to base its explanations too exclusively on the role of social democracy, applies to the analysis of comprehensive school reforms conducted by Wiborg (2009, 2010) as well. Therefore, the debates within this field regarding, for example, the role of Christian democratic parties and churches, but also of agrarian and liberal parties, are relevant for scholars of educational politics. Generally, the comparative-historical analysis of how different political cleavages and power distributions possibly impacted on attempted as well as achieved educational reforms could be enhanced by more extensive communication between these fields. For example, insights from Green (2013), Korpi (2006), Manow and Van Kersbergen (2007), and Rokkan (2000) might be fruitfully combined. Also, future studies of comprehensive school reforms might profit from the theoretical elaborations which have been produced in power resources literature, such as the identification of veto points, policy ratchet effects, regime legacies, and structural limitation. Many important insights regarding, for example, the role of ideological-hegemonic struggles, have been brought forward in both strands of literature. It would be a favorable development if more scholars became aware of these overlaps and if the analysis of educational politics and, in particular, comprehensive school reforms, received an upswing as a result.