The rise of institutional analysis in comparative welfare state research
The rapid expansion of the welfare state after 1945 first drew the attention of a number of macro-comparative sociologists in the tradition of 'structural functionalism', inspired by the American sociologist Talcott Parsons (1949, 1951). What functionalist sociologists found particularly intriguing was the extent to which the different welfare systems that crystallized after World War II increasingly began to resemble each other in terms of public social spending and social insurance coverage (van der Wee, 1986: 320). The American sociologist Harold Wilensky drew a direct link between industrial development and the expansion of the national welfare state in his study Industrial Society and Social Welfare (1958), in collaboration with Charles Lebeaux, and later The Welfare State and Equality (1975; see also 2002). The proliferation of standard methods of mass production and technocratic economic management, according to Wilensky, was clearly the driving force behind welfare state convergence (see also Rimlinger, 1971). In the long run, Wilensky (1975: xiii) boldly stated, ideological and institutional factors were of little significance in explaining the origins and general development of the modern welfare state. Economic development prevailed in explaining the welfare state effort (see also Kerr et al., 1960; Bell, 1965; Amenta and Hicks, 2010). In a more nuanced variant of 'structural functionalism', the sociologist Peter Flora underscored the imprint of political mobilization, alongside economic modernization, on the development of European welfare states (Alber, 1981; Flora and Heidenheimer, 1981; Flora, 1986-90). For Flora, 'modernization' was driven by a combination of both economic development and political change. Expanding capitalist economies in conjunction with the rise of mass democracy spearheaded the introduction of elementary labour market regulation and the expansion of social insurance provision from the late nineteenth century onwards (Flora and Heidenheimer, 1981: 22-3). Taking inspiration from the social and political theorists Reinhard Bendix (1956) and Stein Rokkan (1970), Flora's historical analysis of the evolution of social insurance provision in fifteen European countries revealed how long-standing social cleavages and social policy legacies, over time, cumulated into striking institutional differences in welfare provision across countries (Flora, 1986,1987; Leibfried and Mau, 2008; Kuhnle and Sander, 2010; Nullmeier and Kaufmann, 2010).
Writing in the tradition of comparative public policy, rather than macrosociology, the seminal work, Modern Social Politics in Britain and Sweden, of Hugh Heclo took a different turn (1974). Heclo took a special interest into how non-elected policy elites, civil servants, such as William Beveridge and Gustav Moller, and expert economists, such as John Maynard Keynes and Ernst Wigforss, acted as critical policy middlemen in shaping the modern social policy in, respectively, Britain and Sweden. While Heclo did not deny the importance of socio-economic change and political mobilization, he chose to focus on how pre- and post-war social policy innovation was conceptualized within administrative settings outside the realm of immediate political contestation. By so doing, Heclo brought a critical element of policy voluntarism on the part of administrative officials and policy experts into comparative welfare state research, an element that remained absent in the renditions of 'structural functionalism' of Wilensky and Flora.
Under the fortuitous economic environment of the 1950s and 1960s, the advanced nations of Western Europe experienced, roughly in agreement with functionalist conjectures, fairly convergent patterns of welfare state expansion. After the mid 1970s oil crisis, however, a startling variety of national patterns of crisis management to the predicament of stagflation discredited the functionalist convergence thesis (Goldthorpe, 1984). Divergent policy responses to the oil shocks inspired a new generation of social policy scholars to re-examine the political variables of partisan control over government, state traditions, past policy legacies, and the character of industrial relations, as independent variables to the explanation of divergent patterns of welfare state development (Castles, 1983; Hicks and Swank, 1984; Myles, 1984; Howell, 2005, 2006; Immergut, 2010; Schmidt, 2010). Over the 1980s, welfare state research moved towards middle-range comparative political analysis, centred most critically on power resources and political institutions. Power resources theorists such as Walter Korpi and John Stephens took the democratic class struggle, evidenced in the political power of social democracy, Christian democracy, and associated trade unionism, as conclusive to the generosity of the majority of West European welfare (Stephens, 1979; Castles, 1982; Korpi, 1983; Esping-Andersen, 1985; Myles, 1989; Palme, 1990; van Kersbergen, 1995; Hicks, 1999; Huber and Stephens, 2001; Bradley etal., 2003; Kangas and Palme, 2007). Later Peter Baldwin (1990) drew attention to the critical role of the middle classes in the establishment of social protection programmes, and Isabella Mares (2003), as well as Jacob Hacker and Paul Pierson (2002) underlined employers' interests in externalizing the social costs of production on the welfare state (see also Hacker, 2002, 2004, 2006).
Following in the footsteps of Heclo, a new research in the tradition of historical institutionalism identified relatively stable features of political- administrative systems and policy legacies as important constraints and resources to welfare state development (Evans, Rueschemeyer, and Skocpol, 1985; Katzenstein, 1985; Skocpol, 1985; Skocpol and Amenta, 1986; Steinmo, Thelen, and Longstreth, 1992). Both institutionalists and power resource theorists based their research endeavours on sovereign nation-states, which they saw as the most important loci for collective action, partisan mobilization, democratic decisionmaking, public administration, and policymaking (Krasner, 1984, 1988; Crouch, 1993, 1999).
The landmark study that built on the key insights from both power resources theory and historical institutionalism that came out of this period is Gosta Esping-Andersen's The Three Worlds of Welfare Capitalism, published in 1990. As the modern post-war welfare state fundamentally recast the boundaries between politics and economics, strengthening politics against pure market forces, social policy development required the mobilization of sufficient 'power resources', especially working-class mobilization to overcome bourgeois vested interests (Esping-Andersen and Korpi, 1984). So as not to sacrifice historical nuance, alongside power resources, Esping-Andersen placed special emphasis on the critical impact of the historical legacies of religion, democratization, and political representation (see also van Kersbergen, 1995; van Kersbergen and Manow, 2009, 2010). This allowed him to conceptualize three 'ideal-type' welfare regimes: liberal, conservative-corporatist, and social democratic. The liberal welfare regime is based on the market as the primary source of welfare, with the state providing residual support for those in need. By contrast, in the social democratic regime it is the state that provides generous universal benefits and social services. In a conservative- corporatist regime, the provision and distribution of welfare is based on familialism and male-breadwinner status differentials. In terms of the scope of social protection, the Nordic social democratic welfare regime, based on state guaranteed social rights, is clearly the welfare front-runner, the Anglo- Saxon liberal regime, based upon the market as primary source of welfare, a laggard, and the conservative-corporatist regime of the European continent, based upon family status differentials, falls somewhere in between (Arts and Gelissen, 2002). The Three Worlds of Welfare Capitalism represented a paradigm breakthrough in comparative welfare state research by bringing together a number of institutional dimensions in a coherent fashion, in terms of the predominance of states, markets, and families in welfare provision across countries. Building on an earlier descriptive categorization by Richard Titmuss (1968, 1974), Esping-Andersen presented a novel account of how social policy critically shapes the nexus between labour markets and family structures, with both the liberal and social democratic welfare regimes displaying relatively high levels of female employment and the conservative-corporatist regime, with its Christian democratic roots, discouraging female homemakers from full participation in the labour market (Esping-Andersen, 1990: 26-8). The groundbreaking work of Esping-Andersen indirectly opened up a critical research tradition of gender studies and feminist theory, taking into account transformations in family demography and labour market structures (Sara- ceno, 1994, 1997). Because women assume much greater responsibility for unpaid care, their attachment to the labour market and welfare state differs markedly from that of men (Daly, 2000b; Daly and Lewis, 2000; Lewis, 2006). In general, working women have shorter working hours and lower earnings, which in turn imply that they are more dependent on welfare state support. Feminist studies further broadened the understanding of the mutual relationship between systems of social provision, family caring and gender relations, fertility, care, early childhood education and the balance between work and family life (Lewis, 1992; O'Connor, Orloff, and Shaver, 1999; Sainsbury, 1999; Mahon, 2002, 2006; Daly and Rake, 2003; Orloff, 2006, 2010; Daly, 2010b; see also Korpi, 2000).
Heeding the core insight of historical institutionalism, in terms of the temporal character of welfare state development, Esping-Andersen explicitly separated periods of social policy innovation from periods of institutional reproduction, punctuated by two path-determining epochs. He identified the end of the nineteenth century and the decades between the Great Depression of the 1930s and the end of World War II, as critical periods of transition. In the concluding chapter of Three Worlds of Welfare Capitalism, Esping-Andersen conjectured that the 'inherent logic of our three welfare state regimes seems to reproduce itself' (Esping-Andersen, 1999: 165). Contemporary social policy is not written from a tabula rasa; prior social policy design, together with the maturity of prevailing welfare programmes, severely constrains the action radius of policy actors, obliged to work within an institutional context inherited from the past.
Over the 1980s, when neoclassical economists increasingly judged the prospects for welfare state 'survival' as poor, comparative welfare state research could offer no confirmation of the 'crisis and breakdown' logics that mainstream economists conjectured and neoliberal politicians prophesied. Consequently, the central query in comparative social policy analysis turned to explaining why national welfare states were so remarkably change-resistant. To answer this empirical puzzle, research shifted to focus to concrete institutional social mechanisms and political incentives, which could be held responsible for reinforcing institutional inertia in the face of fundamental social and economic restructuring. In other words, there was another shift in the central research question: from gaining a better understanding behind welfare state diversity and nation-specific historical trajectories of social policy development, through the path-dependent interplay of political voluntarism and historically shaped policy legacies, towards explaining institutional 'lock in' and stability tout court. The theoretical argument of the welfare state as an 'unmovable object' is most closely associated with the seminal work of Paul Pierson (1994, 1996, 1998, 2001c, 2004). In his groundbreaking study Dismantling the Welfare State? (1994), Pierson found only minor changes in welfare provision in the Anglo-Saxon regimes of the UK and the USA in the 1980s. He was able to demonstrate how difficult it is to retrench prevailing social policy commitments, even under the helm of ideological political leaders such as Margaret Thatcher and Ronald Reagan, who were highly motivated to unburden the free market from the overloaded welfare state. Pierson concluded that 'the welfare state remains the most resilient aspect of the post war political economy' (Pierson, 1996:179). He based his explanation for the 'frozen' character of mature welfare states primarily on the (negative) political incentives brought on by the expansion of welfare state itself over the Golden Age of post-war prosperity. Over the 1980s, the 'old politics' of the welfare state, driven largely by 'credit claiming' class mobilization, had been displaced by a 'new politics' of the welfare state, characterized by a general fear of electoral retribution and vested interest opposition over cuts in popular social programmes. For Pierson, mature welfare states are quintessential sites of institutional self-reinforcement, making pathbreaking reform progressively more improbable (Pierson, 2004; Jacobs, 2010). The welfare state had turned into an immovable object, leaving governments with very few feasible strategies of restraining social policy commitment, under conditions of permanent austerity, other than hiding the visibility of reform, dividing electorates and clienteles, and delaying the regressive distributive effects of half-hearted reforms on traditional beneficiaries.
Since the publication of Pierson's famous book, the 'new politics' conjecture of welfare inertia has been corroborated by a great many failed reform cases on the European continent, such as the stalemated pension reforms in Italy in 1994, leading to the downfall of the first Berlusconi government, and in France in 1995, where Prime Minister Juppe had to withdraw his social insurance reform plan after massive protests. Many more comparative studies and journal articles since observe how politicians have cautiously tried to hide and disperse the costs of retrenchment while making side payments to large constituencies and insider groups in order to minimize partisan political costs (Castles, 1998; Clayton and Pontussen, 1998; Mares, 2000, 2003; Bonoli, 2001, 2007; Huber/Stephens, 2001; Swank, 2002; Tsebelis, 2002; Kittel and Obinger, 2003; Korpi and Palme, 2003; Pontusson and Kwon, 2003; Kenworthy, 2004; van Kersbergen, 2006; Rueda, 2007; Iversen and Soskice, 2009; Vis, 2010; Martin and Swank, 2011). In agreement with power resource analysis, on the other hand, a fair number of scholars continued to find empirical evidence that leftist governments by and large push for fewer cutback than conservative ones (Korpi and Palme, 2003; Allan and Scruggs, 2004).
By the late 1990s, comparative research more in line with the tradition of Esping-Andersen's regime analysis began to identify institutional variables capable of explaining patterns of more nuanced path-dependent trajectories of policy adjustment in the context of intensified economic internationalization, the expansion of the service economy, demographic ageing, and the feminization of the labour market (Esping-Andersen, 1990, 1996; George and Taylor-Gooby, 1996; Kitschelt etal., 1999; Bonoli and Palier, 2000; Gallie and Paugam, 2000; Scharpf and Schmidt, 2000a, b; Swank, 2001; Allan and Scruggs, 2004; Gallie, 2009; Arts and Gelissen, 2010). A particularly important research project in this vein found its way into the two-volume comprehensive study Welfare and Work in the Open Economy, edited by Fritz Scharpf and Vivien Schmidt (2000a, b), in collaboration with many leading comparative social policy scholars. The Scharpf and Schmidt research team observed that the twelve countries under study varied enormously in the social reforms they undertook from the late 1970s. The challenge of intensified economic internationalization confronted each welfare regime family with a distinct constellation of regime-specific adjustment problems and reform agendas. The contribution of Martin Rhodes and Herman Schwartz highlighted how Anglo-Saxon welfare states by and large increased the scope of the free market and strengthened the selective nature of social programmes, which was accompanied by growth in employment. The flipside of the success of the Anglo-Saxon 'jobs machine', however, was a significant rise in income poverty. By contrast, the Scandinavian welfare states were best able to maintain a both generous and universally accessible system of social security through activating labour market policies. Over the same time period, the Continental welfare states got caught in a negative spiral of high gross labour costs and rising economic inactivity. The growing volume of social insurance beneficiaries led to further increases in growth labour costs, which in turn held back employment growth in services. In Southern Europe, the Continental 'inactivity trap' was exacerbated by the stringent regime of insider-biased labour market regulation, which intensified labour market exclusion of young people and especially of women. As there were virtually no provisions for combining careers with family caring, this predicament resulted in low birth rates and extremely high pension burdens. In short, comparable economic challenges and social developments across the wider European economy led to very different policy problems in the different welfare regimes, which in turn triggered diverging reform paths and reform backlashes and political opposition. In conclusion, Scharpf and Schmidt agree that institutional features shape the menu of feasible policy options across different welfare regimes, of which Reformstau is one likely outcome among many. Blame-avoiding politics and insider-biased reform opposition are not the only shows in town. Moreover, the exigencies of economic internationalization do not evoke identical policy responses, not even across comparable welfare regimes with similar institutional capacities to orchestrate, enact, and implement social reform. Over the 1980s and 1990s, Austria, the Netherlands, Sweden, Denmark, and Australia were able to steer a more balanced course in achieving rising levels of employment while preserving generous welfare provision (Hemerijck and Schludi, 2000; Scharpf and Schmidt, 2000a, b; see also Metcalfe, 1978).
Institutional inertia is an ordinary part of post-formative welfare state development. The pathbreaking studies of Pierson on institutional self-reinforcement in the welfare state remain extremely valuable contributions to our understanding of modern social politics in affluent democracies. But when taken too far, the conjecture of the 'new politics' of the welfare state literature becomes unnecessarily deterministic, offering progressively less analytical purchase on the complex processes of profound post-formative welfare state change across Europe since the late 1980s (Hausermann, 2010a; Steinmo, 2010).