One of the most significant structural transformations in postwar capitalist democracies has been the rise of the welfare state. The theoretical intent of the traditional sociological and economic inquiry into the welfare state has focused less on trying to understand the welfare state itself and more on to what extent and under what conditions welfare provisions influence social and economic outcomes such as equality, employment, and labor market behavior. Over time, however, scholars have turned toward historical and political factors. G. Esping-Andersen identified three types of welfare state that seem incongruent with the real worlds of welfare capitalism: the “liberal,” “conservative/corporatist,” and “social democratic.” In contrast to the period until the mid-1980s that focused on welfare state expansion, the late 1980s saw the emergence of new streams of literature whose emphasis was on welfare state retrenchment. More recently, scholars have advanced the argument that the globalization of capital markets has effectively increased the power of capital over governments that seek to expand or maintain relatively high levels of social protection and taxation. Another notable trend is the increased intellectual interest in the relation between development and social policy and the growing interface between social policy and economic policy. A question that arises is whether distinctive welfare regimes have the ability to survive, particularly if their norms clash with those of the competition, or Schumpeterian workfare state.
The rise of the welfare state is perhaps the most significant structural transformation in postwar capitalist democracies. Yet the precise definition of the welfare state and exactly what constitutes it remains implicit. As Titmuss famously put it, the welfare state is an “indefinable abstraction” ( 1968 :124). Esping-Andersen also claims that a “remarkable attribute of the entire literature is its lack of much genuine interest in the welfare state as such […] the welfare state itself has generally received scant conceptual attention” (1990:18). Indeed, the theoretical intent of the traditional sociological and economic inquiry into the welfare state has remained primarily concerned with the issues that are less about arriving at an understanding of the welfare state itself and more about addressing the social question as to what extent and under what conditions welfare provisions influence social and economic outcomes (e.g., equality, employment, and labor market behavior).
Over time, however, a significant degree of intellectual progress has been made as the socioeconomic inquiry into welfare has turned toward historical and political reasoning (cf. Hicks and Esping-Andersen 2005). More explicitly, the recent contribution of political science literature has shifted our attention to the political causes of welfare state development and cross-national variation. The welfare state is no longer being treated only as an independent variable but also as a dependent variable (van Kersbergen 2002; cf. Clasen and Siegel 2007). The theoretical progress advancing our understanding of welfare has grown in tandem with the specific historical development of the welfare state itself. During the early years of welfare state evolution, it was the economic and sociological approaches that almost exclusively dominated the welfare state expansion literature. In comparison, the retrenchment literature that has occupied the past two decades or so has been the domain of the political science approach (Green-Pedersen and Haverland 2002).
Against this backdrop, the most promising and intellectually inspiring progress was made as the emphasis turned to the confluence between different disciplinary traditions not only in answering why welfare states ever emerge and continue to develop, but also in accounting for why and how they differ. It is also where primary explanatory variables typically stressed by each disciplinary tradition are exchanged that we see many of the most convincing accounts of welfare state development. Over a long period, a series of discrete, incremental theoretical developments have coalesced into new theoretical paradigms, most of which share the comparative tradition of welfare state development. The most fruitful and intellectually stimulating pieces of research have emerged in this tradition not only because comparative research often challenges our preconceptions about the nature of best practices and exposes us to alternative ideas, but also because it reminds us that there are vast differences between the welfare states of today.
The first generation of welfare state research was very much occupied with the question of why welfare states emerge, rather than why welfare states differ and how they differ. Instead of making explicit claims about what explains welfare state variations, the exercise was one of “devising laws” that could account for welfare state development and could be applied to a whole range of countries.
Central to this view, rooted in the functional or structural perspective, is that it is the “system that ‘wills’, and what happens is therefore easily interpreted as a functional requisite for the reproduction of society and economy” (Esping-Andersen 1990:13). The emergence of welfare states is then a necessary change that balances the “system” because industrialization undermines the traditional social support mechanisms and causes new social needs, while economic growth provides the material basis for social policy expansion to meet these needs (Kerr et al. 1973; Wilensky 1975; Flora and Alber 1981). The logical conclusion drawn from this premise is one of convergence or technological determinism: all industrialized countries are growing more alike, meeting social needs through broadly similar forms of welfare intervention even though they have vastly different class, political, and cultural traditions (Wilensky with Lebeaux 1965; Rimlinger 1971; Wilensky 1975). It is therefore this industrial development that indirectly leads to greater social equality through higher welfare spending from which those in need benefit most (Kerr et al. 1973). The differences between welfare states are seen merely as being those of the degree of development. Hence societies are distinguished as being “welfare state leaders” or “welfare state laggards” rather than by distinctions of a more fundamental kind.
Clearly, this line of reasoning carries some compelling elements, not least because it offers a narrative that broadly fits with the historical path of development of the nations to which it was applied. Nonetheless, the core proposition of this logic of industrialism has been subject to a number of criticisms. First, as is the case with the transnational diffusion literature that predicts the sequence within which welfare programs are adopted but not their form (e.g., Abbott and DeViney 1992), it oversimplifies the process of policy development and ignores the important structural variations in the organizational and institutional features of welfare states.
Second, most studies in the tradition focus on welfare spending levels, which may be misleading as these levels tell us very little about where and how money is being spent. Third, the perspective has a significant problem in accounting for the timing of change within societies as well as in explaining varying time lags between industrialization and welfare development. Finally, industrial determinism places little emphasis on ideological factors and partisan politics, dismissing them as being “almost useless in explaining the origins and general development of the welfare state” (Wilensky 1975:xiii) and considering the role of politics as merely a technological matter. The stability and the legitimacy of the state seem to be bound to technological success and prolonged economic growth (Carrier and Kendall 1971:201).
Unlike industrial determinism, the logic of capitalism approach or neo-Marxism regards welfare state development as either a response to threats to the interests of the ruling class (i.e., the Bonapartist interpretation), or preemptive mechanisms to protect against such threats (i.e., the Bismarckian interpretation). For neo-Marxists, social provisions emerge and progress because state action promotes the needs or requirements of capital (accumulation); because states take preemptive action to prevent working class discontent (legitimation); and because states require a response to class conflict. The extent of class struggle becomes the key determinant. The development of the welfare state is therefore a necessary response to the contradictions within capitalism.
This neo-Marxist understanding of welfare state development is not uniform, however: there are at least three distinctive theoretical propositions within it. To start with, instrumentalists heavily criticize the core claim of pluralists that the state plays the role of neutral broker in mediating conflicts between various interests. Rather, for them, the state plays an instrumental role only to serve the capitalist’s interests. Miliband argues that “the intervention of the state is always and necessarily partisan: as a class state, it always intervenes for the purpose of maintaining the existing system of domination, even where it intervenes to mitigate the harness of that system of domination” (1977:91). It is the capitalist who monopolizes economic organizations, exercises strong influence on political organizations, and determines social provision depending on the necessity for capital accumulation. Hence welfare reforms basically function as a form of social control rather than of social investment (Piven and Cloward 1971; Ginsburg 1979; Miliband 1991). Neither the basic structure of capitalist society nor the status, income, and political power of those who have already been influential are affected by social policy.
While this instrumental version of neo-Marxism comes closest to classical Marxist theory, a slight modification was made by those who stress the objective relation between the state and the capitalist. For them, it is the structure of the capitalist economy itself that enables agreement between the state function and the capitalists’ interests. Regardless of the capitalists’ influences, it is necessary for the state to preserve order and to sustain and enhance the conditions for capitalist economic activities. For this reason, the state might purse welfare provisions against the wishes of the capitalists. This will then strengthen the long-term stabilization of the economy (O’Connor 1973; 1984; Gough 1979; Offe 1984). In other words, the stability of a relatively autonomous state is a precondition for the political system to serve the interests of capitalists in the long run (Skocpol 1980). Of course, political conflict induced by the working class can be threatening to this political stability. Therefore, one of many important functions of the welfare state is to control and dismantle labor power. Similarly, welfare provisions can be expanded as part of a cooperative strategy of control (Piven and Cloward 1971).
In comparison, the third variation puts more emphasis on political class struggle. Instead of stressing the capitalist’s long-term or short-term interests, the development of the welfare state is interpreted as the outcome of a long political struggle between the working class and the capitalist and its allies. Known as the “power resources” approach, the fundamental power resources in advanced industrial societies are assumed to be divided and derive from labor control through the market mechanism and from political power through the capacity for collective action (Korpi 1989). Therefore, this working-class mobilization thesis explains that the development of the welfare state is dependent upon the strength of the labor movement and its political ability to implement collective welfare provisions through electoral control of the state. Similarly, proponents of the mass disruption thesis (see Piven and Cloward 1971) argue that social policies have improved because of the elite’s responses to protest by the poor and workers.
Strictly speaking, the class mobilization thesis should be seen not as a mere variation but as a distinctive theory that has its root in the neo-Marxist tradition because, unlike the logic of capitalism, which sees welfare spending as a means for capital to maintain its dominant position, it views welfare spending as a reflection of the political power gained by workers. It pays attention to the relative power of the “subordinate” class and regards variations in their power as being key in explaining variation in welfare state development.
Indeed, it is this thesis that has worked as a theoretical foundation for many. Nonetheless there are several valid objections. First, it places a strong emphasis on the harnessing of working-class power through its representative left-wing party. Indeed, a social democratic or labor party that represents the working class and highly centralized labor unions has been influential in driving welfare state expansion. However, socialist parties have not always played a dominant role in pushing forward welfare provision, especially in the early stage of welfare state development. In fact, many welfare programs have been introduced by liberal and/or conservative elites. Failure to account for the role and, indeed, power of non-leftist parties is its key weakness (Wilensky 1981; Borg and Castles 1981). Second, one of the assumptions that it makes is that the locus of decision-making power lies in parliaments. Thus great efforts have been made to collect data on leftist party representation in parliaments. But parliaments are not always the locus of power; indeed, extraparliamentary organizations often play a major role in policy development (George and Wilding 1994). Third, the ways in which working-class power is measured are problematic. Levels of unionization and percentages of votes gained by leftist parties in elections are typically used as proxy measures of working-class power (Korpi 1983:39). Yet such measures treat power as a linear concept that ignores the role the working class plays in practice and is insensitive to the particularities of the ways in which a nation’s political institutions are organized, impacting on policy outcomes to varying degrees. And finally, as with much of the early work, this too was hampered by treating the development of welfare in a linear manner, using aggregate welfare expenditure as a proxy for welfare development.
A significant advance has been made in theorizing welfare state development. Yet much of early work tends to focus on finding one single powerful causal force within the well-established procedures and assumptions which are based on conceptions of linearity. Instead, the second generation of welfare state research began to identify salient interaction effects of multiple factors. In his early work, Esping-Andersen (1985), one of the most prominent adherents of the “working-class mobilization theory,” presented a classic formulation that distinguished social democratic models from others. Yet later, in his masterwork (Esping-Andersen 1990), he refined and significantly changed this duality, abandoning an ideal mode of one extreme or the other, and identified three separate routes of welfare state instead.
Breaking with his own past, Esping-Andersen (1990) points out that there are three particularly important factors at work: the nature of class mobilization (especially union structure); the opportunities to form class-political coalitions (especially those which incorporate the middle class); and the historical legacy of regime institutionalization. Significantly too, his conceptualization of the welfare state differs radically since it is not the level of social expenditure that is measured but the “degree to which individuals, or families, can uphold a socially acceptable standard of living independently of market participation” (Esping-Andersen 1990:37). His measurement of social rights has led him to cluster welfare states around three different types, which he termed the “liberal,” “conservative/corporatist,” and “social democratic.”
The identification of these clusters itself was the key contribution of his work. Yet what is more compelling is his explanation for the reasons why particular nations developed their welfare provision along one particular path or another. He rejects the conception of an evolutionary development of social reform. For him, it is the interaction of three most important factors over time that produced a distinctive welfare state regime, much of which depended on the outcome of the power constellations during the formative phase of welfare states. Each of his three highly diverse regime types is organized around its own discrete logic of organization, stratification, and societal integration, and is given an “account that is much more nuanced, much more attentive to particularities and context – in a word, much more historical – than the older, unilinear, social scientific approach […] by taking a more contextualized and particularized approach to the development of social policy” (Baldwin 1992:703). For Esping-Andersen (1990:4), and Baldwin (1990:299) too for that matter, “politics not only matters, but is decisive.”
Esping-Andersen’s work provided a platform for further research to a great extent, often coming from the challenges concerning the validity of his regime typology itself and the methodological flaws or oversight (Arts and Gelissen 2002). For instance, his ideal-typical treatment of regime types sits uncomfortably with the real worlds of welfare capitalism (Leibfried 1992; Castles and Mitchell 1993; Jones 1993; Ferrera 1996; Goodman and Peng 1996; Bonoli 1997; Castles 1998; White and Goodman 1998; Goodin et al. 1999; Holliday 2000). Kasza (2002) dismisses the usefulness of welfare regimes typology altogether for it generates too many bad fits and argues that policy-specific comparisons could provide better offerings instead. And Castles (1993) and his colleagues developed a “social protection by other means” approach which goes beyond state welfare. Viewed from a “families of nations” perspectives (Castles 1993:xiii), there are more than three worlds of welfare. And this perspective, which is claimed to have a much wider policy span than regime interpretations, shows the coherence and persistence of the clustering of nations over time despite fundamental transformations of the international and domestic political economy (Castles and Obinger 2008). Room (2000) challenges Esping-Andersen’s analysis of decommodification, arguing that the Marxian notion of labor commodification is related not only to the economic consumption of labor but also to the loss of workers’ own opportunities for personal self-development as its consequence. And those who are dissatisfied by his interpretation of confining welfare state policies to the ones that are related to the labor contract argue that family policy, gendered division of labor, and social services have great importance for people’s standard of living regardless of whether or not they participate in the market (Lewis 1992; O’Connor 1993; 1996; Orloff 1993; 1996; Bambra 2005).
There is little doubt that Esping-Andersen’s major contribution to welfare state literature has been the concept of regime that denotes the ways in which welfare production is allocated among state, market, and households (see also Esping-Andersen 1999). His typology therefore does not confine itself to a specific social program or a type of welfare state. This construction of typology per se, however, is not entirely new. As Esping-Andersen himself acknowledges (1990:20), Titmuss (1974) had, much earlier, identified three broad types of welfare state that were not too dissimilar. But what is distinctive here is that by stressing the importance of the particularities, not the regularities, welfare states are seen as “parts of complex, historically generated configurations”; the three worlds are qualitatively different and result from “the cumulative effects of a number of interdependent causal factors” (P. Pierson 2000:809). In fact, Esping-Andersen is not alone in stressing historical and institutional factors. Indeed, much of the recent literature focuses on the argument that institutions play a key role in the determination of social and political outcomes. This forms the core of an important body of theory that has attracted increasing attention (see Baldwin 1992; Steinmo et al. 1992; Hall and Taylor 1996).
If the welfare state scholarship until the mid-1980s was dominated by works that examined the patterns, reasons, and ways of welfare state expansion, a significant shift began in the late 1980s. Particularly important was the upheaval of the new streams of literature, especially in relation to welfare state retrenchment that brought about a renewed emphasis on the role of institutions.
Evolved from the state-centered approach that criticizes the demand-driven approaches (e.g., social forces and conflict) involving passive and defensive actions from a government (Nordlinger 1981; Orloff and Skocpol 1984; Evans et al. 1985; Skocpol 1985; Skocpol and Amenta 1986; Ashford 1986; Almond 1988; Thelen and Steinmo 1992), scholars inspired by the institutional perspective began to stress the relationship between institutions and individual behavior and the distinctive political outcomes produced by this relationship (Steinmo et al. 1992; Hall and Taylor 1996). For them, “political life is characterized, not simply by a struggle over the allocation of resources, but also periodically by strife and uncertainty about the rules of the game within which this allocative process is carried out” (Krasner 1984:225). Institutions establish the rules of the game, have long-term effects, foster stability by resisting dramatic change, and condition the opportunities and incentives for political action or inaction. They are thereby seen as intervening or intermediate variables that shape behavior and political outcomes (Krasner 1984; Gorges 2001). Taking the cue from Heclo’s (1974) illuminating parallels between the British and Swedish systems, where he stressed the critical importance of the inheritance of past policies in determining what is feasible at any given time, those inspired by this tradition have shifted the study of social policy away from a predominantly sociological perspective and into the realm of political science.
Historically informed work by the likes of Esping-Andersen (1990), Baldwin (1990), Immergut (1992) and Skocpol (1992) could all be grouped under this tradition, particularly in their articulation of the ways in which institutions and interests interact and in their claims that different paths of welfare state development have occurred over an extended period of time. It is recent debates surrounding globalization and the crisis of the welfare state, however, that have brought with them a fresh wave of theorizing in the realm of the institutional analysis of social policy.
Led by the work of P. Pierson (1994), who examined the attempts of conservative governments in the United Kingdom and the United States to cut back entitlements and weaken the political foundations of the welfare states, the politics of welfare state resilience came into the center stage of welfare state literature. Obviously, the coming into office of Margaret Thatcher in 1979 and of Ronald Reagan in 1981 sparked the academic interests, backed by conservative academics who claimed that the welfare state had become a significant source of social and economic problems instead of solutions (Murray 1984; Mead 1986; Marsland 1996). Also, significant changes in the economic environment practically ended one of the key foundations of the Golden Age welfare state: full employment (Esping-Andersen 1996).
For many welfare states, the conventional method of pursuing full employment has been through neo-Keynesian interventions. However, substantial shifts in productive activity have led to problems of job security and deterioration in pay for the jobs most subject to competition. Social foundations in postindustrial economies are also fundamentally different from those in the immediate postwar period (Esping-Andersen 1999). The postwar welfare state settlement was particularly directed at meeting needs during interruptions of the industrial wage, while care provisions were mostly assumed to be widely available informally. But the increasing rate of female labor participation led to demands for more equal opportunities and for care services. On the one hand, the growth in jobs was confined within the service sector, which tended to produce low or unskilled jobs. Low productivity growth in this sector compared to the manufacturing sector resulted in a downward pressure on wages. On the other hand, a low-skilled service sector afflicted with an abundance of female labor placed added pressure on wages and on care industries. Both generated a dilemma between job growth in the service sector and rising inequality (Esping-Andersen 1999). At the same time, the shift away from manufacturing tended to weaken union power and labor-associated political parties which were traditionally understood to be able to mobilize political pressure for welfare state expansion. The long list of formidable challenges and pressures that would potentially threaten the viability of the welfare state or even its survival exerted a combined pressure on traditional forms of state welfare both in terms of the range of options available for governments to promote employment and finance social provision, and in terms of increasing risks, needs, and demands as a result of labor market and family changes (Baumol 1967; Iversen and Wren 1998; Esping-Andersen 1999; P. Pierson 2001a; Taylor-Gooby 2004). In short, the Golden Age of postwar welfare state expansion and its achievement were being viewed as having grown to its limits (cf. Flora 1986).
An increasingly interdependent world economy has also led many scholars to anticipate a significant degree of convergence (Scharpf 1991; Mishra 1996; Greider 1997; Martin and Schumann 1997; Gray 2002). Indeed, many countries have embraced the free market policy prescription as a solution to a range of policy problems, and some scholars predict a long-run decline – a race to the bottom – of the welfare state (Rodrik 1997; Allard and Danzir 2000) or a future of “permanent austerity” (P. Pierson 2001b:456). On the other hand, many studies of welfare state trajectories during the 1990s and the early years of the twenty-first century indicate that various welfare states respond differently to more or less similar sets of challenges, thereby negating a second coming of convergence thesis. The key to this divergence has been the politics of reform in each country, which has produced very different results and reform paths (Esping-Andersen 1999; Scharpf and Schmidt 2000a; 2000b; Huber and Stephens 2001; P. Pierson 2001a).
The gap between the theoretical prediction and empirical reality gave rise to a new body of literature with substantially different aims. The problem was that the major explanatory accounts that dominated the heyday of welfare state expansion no longer seemed to provide convincing explanations, despite the claim that, as Esping-Andersen put it, “a theory that seeks to explain welfare-state growth should also be able to understand its retrenchment or decline” (1990:32). For the scholars who advocate the welfare state “resilience” thesis, and P. Pierson (1996) in particular, theories about the “old politics” (Skocpol and Amenta 1986; Amenta 2003) that rely on socioeconomic functionalism and class-based power resources do not adequately explain developments after the end of the “Golden Age.” In other words, an increasing number of scholars began to identify that “politics” in the old sense, typified by class, parties, and unions, matters less and less for welfare state development and outcomes (Castles 1998; Stephens et al. 1999; Huber and Stephens 2001). Arguments about how to explain retrenchment or the resilience of existing welfare regimes began to focus on “new” political variables such as party systems, the logic of elections, political institutions, and political learning (Kuhnle 2000; P. Pierson 2001a; Huber and Stephens 2001; Swank 2002). And these new factors at play make the logic of welfare state retrenchment very different from welfare state expansion (P. Pierson 1996). The key questions that dominated the academic inquiry of the late twentieth century were concerned with how and why existing welfare state regimes either seem to resist change or appear to change only incrementally according to a built-in regime logic that seems to reaffirm, if not aggravate, the difficulties that are typical for the regime (Starke 2006).
Protagonists of the “new politics” of welfare state development argue that major policy reforms are very difficult for two reasons. First, welfare provisions established during the period of postwar welfare state expansion created their own program-specific constituencies that subsequently led to the political unpopularity of cutbacks (Boeri et al. 2001; Taylor-Gooby 2001). Second, deeply entrenched welfare state institutions exert path-dependent rigidity. Any radical attempts to alter them are severely constrained by institutional structures and existing policy designs which are intrinsically very difficult to change also. Put simply, history matters for institutional development (P. Pierson 2004). A number of scholars have contributed to this resilience thesis by adding to or refining the argument by stressing a specific dimension of institutional analysis (Green-Pedersen and Haverland 2002). Overall, it is not a political logic that governs welfare state adaptation but an institutional logic that explains it.
The renewed emphasis on the functioning of institutions is broadly divided into two central considerations: the role of political institutions and their effects on the politics of welfare state reform; and the existing structures of social provision itself as welfare state institutions and their enduring effects. First, the institutional fragmentation of political systems makes welfare reform either more difficult or relatively easier. Large-scale retrenchment is often less likely to be envisaged in systems with a high degree of vertical and horizontal fragmentation of power due to the higher number of veto players. In systems with federalism, a bicameral parliament and inclusive electoral system provide various points of veto power which substantially reduce the room for governments to make reforms (Immergut 1992; Bonoli 2001; Huber and Stephens 2001; Swank 2001; Tsebelis 2002; Obinger et al. 2005). Conversely, those countries with a relatively small number of veto players could potentially enjoy a high concentration of political power. Nonetheless this remains a potential, for a concentration of power is often accompanied by a concentration of accountability (Pal and Weaver 2003). In those countries, for instance, dissatisfied voters can easily identify who is responsible for unpopular cutbacks. Politicians ultimately seek to be reelected, and hence try to avoid blame. In systems with a high degree of concentration of power, this “blame avoidance” (Weaver 1986) is more difficult than in other political systems where there are multiple actors who may have played significant roles at different stages and levels of decision-making processes (P. Pierson 1994). How, then, have those countries with a high degree of concentration of power managed to cut back on welfare entitlements? P. Pierson (1994) highlights three particular strategies governments use to avoid blame: compensation, obfuscation, and division. Those countries with power fragmentation may be better positioned to make retrenchment reforms but they are also equally vulnerable to blame sharing, while those countries with power concentration seem to have facilitated various strategies of retrenchment. Importantly, in this context, the execution of these strategies is often highly dependent upon institutional structures and existing policy designs.
Second, welfare state institutions themselves generate enduring effects. Social policies themselves, either intended or unintended (cf. Baldwin 1990; 1992), tend to create their own institutions that are often difficult to reform because, as most vividly outlined by the experience of pensions reform (e.g., Myles and Pierson 2001), institutional stickiness results in policies displaying “increasing returns.” Similarly, the argument is that past decisions constrain and impact upon future decisions to create a policy feedback effect whereby policies themselves become a central part of the policy-making process itself, rather than a mere outcome. Swank, for instance, argues that “aspects of programmatic structure have substantial impacts on the representation and the relative political capacities of pro-welfare interests” (2002:52). Much earlier, Skocpol and Amenta captured this enduring effect of social policy by pointing out that “not only does politics create social policies; social policies also create politics. That is, once policies are enacted and implemented, they change the public agenda and patterns of group conflict through which subsequent policy changes occur” (1986:149). Huber and Stephens (2001) go even further by pointing to the importance of the “ratchet effect,” where mass opposition against retrenchment in provisions, especially in the areas of pensions, education and healthcare, is often much stronger than support for their introduction.
The overall notion of path dependency is also present in the welfare regimes literature in the sense that distinctive welfare regimes produce distinct policy legacies which in turn largely determine both the extent of change and the types of change that may be possible. For Esping-Andersen, it was the “class coalitions in which the three welfare-state regime types were founded” (1990:33) that generated the bearing of an existing welfare state structure on the current politics of change. For P. Pierson (1994; 2001a; 2001b), it was more of a sectoral dynamic that generated varying policy outcomes depending on the specific social policy areas in question. So, for instance, Myles and Pierson (2001) found that various trajectories followed by many nations in the Organisation for Economic Co-operation and Development in their pension reform do not neatly correspond with existing categorizations of those nations’ political cultures or historic welfare regimes. And they claim that examining preexisting pension arrangements provides the best explanation for the paths of reform chosen. In both cases, specific regimes, once consolidated, tend to produce unique policy path dependencies that in turn overdetermine solutions to new problems as well as strategies of welfare reform (cf. Scharpf and Schmidt 2000a; 2000b).
The idea of path dependency tending to lock the possibility of change into a predetermined trajectory seems somewhat overdone, however, ignoring the impact of subsequent changes in, for instance, power constellations. Indeed, preceding developments are often corrected or programs completed by complementary institutions. The accumulation of anomalies and those developments that may lead to greater internal inconsistency are all the more important, resulting in a possibly significant “paradigm shift” (P. Hall 1989; 1993). Even in the prototype Scandinavian redistributive welfare states, contribution-based schemes providing earnings-related benefits are increasingly supplementing, if not replacing, their universal flat-rate benefits systems. Occupationally separated schemes in conservative welfare regimes have approached near universalism, making the distinction between different occupational categories more or less meaningless. The fight of social democrats against the “exclusionary corporatism of narrow status-solidarity” or their “struggle for comprehensive, egalitarian, and universalist social security” (Esping-Andersen 1990:65, 109) may no longer hold, not least because they were not even the chief advocates of universalism during the pre- 1945 liberal regimes in Scandinavia (Baldwin 1990).
Thus far, much of scholarly endeavor has been either about welfare state expansion or welfare state retrenchment, and accordingly the resilience of welfare states in relation to change. More recently, though, a number of research projects have begun to examine the new risk configurations that have emerged in the transition to postindustrial societies and that in turn have challenged welfare state arrangements that were established in the context of old, traditional risk contexts of industrial societies (Taylor-Gooby 2004; Bonoli 2005; Armingeon and Bonoli 2006). Numerous risk categories peculiar to postindustrial restructuring make an entry (Esping-Andersen 1999). Yet the central driving force of this postindustrial change is the notable rise in the international mobility of capital, which has an unprecedented impact on the welfare state. Swank has argued that Marxists, neoliberals, political scientists, economists, and popular analysts utilize “nearly identical reasoning to argue that the globalization of capital markets has effectively increased the power of capital over governments that seek to expand or maintain relatively high levels of social protection and taxation” (2001:203).
Those subscribing to such views believe that the high-spending welfare regimes are coming under sustained pressure to reduce the size and cost of their welfare programs because failure to make “domestic investment conditions attractive to internationally mobile capital” (Evans and Cerny 2003:55) will lead to capital flight from their economies. Drawing from the international relations literature, Evans and Cerny (2003) argue that the era of postindustrialism is remarkably different from the period that preceded it. In the postwar boom period, social policy was a relatively autonomous field of policy, a domestic issue that was unimpeded by wider economic concerns and so favorable to continual increases in state spending on welfare state activity. Yet “globalization has undermined these conditions.” Hence they anticipate the emergence of the “competition state,” which is “the successor to the welfare state, incorporating many of its features but reshaping them, sometimes quite drastically, to fit a globalizing world” ( 2003 :20, 24).
A similar line of argument has been developed in the sociological literature too. Building on work from regulation theory and ideas embedded in the notion of “post-Fordism” (cf. Burrows and Loader 1994), Jessop argues that “relative to the earlier post-war period, social policy is becoming more closely subordinated to economic policy […] and its delivery has been subject to a partial rollback of the state in favour of market forces and civil society,” not least because “the opening of national economies makes it harder to pursue social policy in isolation from economic policy” ( 2000 :171–2, 182). He claims that we have seen the death of the old-style Keynesian national welfare state, as typified in its ideal type by Esping-Andersen’s social democratic regime, and are seeing it replaced by what he calls the “Schumpeterian Workfare Postnational Regime,” a more modest offering characterized by more limited social rights and an emphasis on encouraging flexible labor markets.
Much of this begs serious questions about the ability of distinctive welfare regimes to survive, particularly if their norms clash with those of the competition, or Schumpeterian workfare state. While many have been keen to proclaim the end of the high-spending welfare state and welcome convergence on a more limited model of welfare, others have pointed to a continuing diversity in welfare provision and, crucially, the generally incremental nature of welfare reform, and consequently the continuation of distinct welfare regimes. Of particular significance is the considerable contribution made by the “varieties of capitalism” literature (Hall and Soskice 2001). By bringing together disparate strands of thinking in political science, economics, law, and business, it draws directly on Esping-Andersen’s analyses of regimes and then empirically observes a distinction between two types of contemporary capitalism: liberal market economies (LMEs), where economic coordination takes place primarily through market mechanisms and neoclassical frameworks provide a reasonable approximation of economic reality; and coordinated market economies (CMEs), in which nonmarket modes of coordination are prominent. These two types are the product of the ways in which particular combinations of institutions cluster in order to solve coordination problems that characterize the rules governing industrial relations, vocational training and education, corporate governance, interfirm relations, and internal relations with employees. According to this approach, what matters is not just an institutional context determined by the terms of competition in markets, but also these areas that are crucial to the firm’s operation. Firms are seen as the central actors in the economy whose behavior aggregates into national economic performance. In order for them to prosper, they must engage with others in multiple spheres of the political economy, where efficient coordination between them is key to success – all of this making this “varieties of capitalism” approach essentially actor centered.
Mares (2001) and Estevez-Abe et al. (2001) apply this approach to welfare production regimes, challenging the decades-long argument of the power of left parties and labor organizations as determining the extent of protection. According to them, contrary to the view that social policy is often thought to interfere with labor markets by raising labor costs or the reservation wage, social policies can also improve the operation of labor markets from the perspective of the firm. Unemployment benefits with high replacement rates, for instance, can improve the ability of firms to attract and retain pools of labor with high or specific skills. Disability benefits and early retirement benefits can also allow firms to release labor without violating implicit contracts about long-term employment. So they argue that specific types of political economies lead toward distinctive welfare states not least because relative abundance in certain skills in a given country constitutes a comparative advantage for firms in that country. Firms’ product market choices are constrained by the availability of the necessary skills, and the availability of specific skills in turn requires appropriate forms and levels of social protection. The existence of institutional complementarities, in short, is therefore the theoretical core of this highly influential approach. Similar to P. Pierson ’s claim that the importance of unions and left parties is likely to decline over time as welfare systems themselves generate broader constituencies of stakeholders and beneficiaries, the varieties of capitalism literature challenges or at least supplements the view that private sector actors have their own distinctive social policy interests depending on their underlying production strategies, negating once again the second coming of convergence thesis.
Pressures under economic globalization have urged us to think again about the traditionally protective functions of social policy and whether this particular dimension will be robust enough to capture the highly complex nature of welfare state retrenchment and/or restructuring in the postglobalization era. As discussed earlier, this has led to a significant rethinking of the sustainability of welfare state. Increasingly, much attention has been paid to identifying “productive” elements of welfare compared to its traditional “protective” traits (Giddens 2000; Holliday 2000; Room 2000; Midgley and Tang 2001; Castells and Himanen 2002; Powell and Barrientos 2004; Vis 2007) and to test the argument that “welfare states have shifted away from traditional protective functions towards a model of ‘productive welfare,’ characterized by a greater emphasis on investment in human capital” (Hudson and Kühner 2009:34).
This widespread adoption of workfare represents a shift in focus to target more of a supporting and subjugated role of social policy vis-à-vis economic policy. This could be seen as a novel paradigm shift if we accepted the claim that “social security in the industrial countries is not intended to promote economic development, but rather to maintain income and provide support to those in need” (A. Hall and Midgley 2004:234). However, this is at odds with the history of welfare policy development in Western Europe. In fact, the presupposed incompatibility of the hierarchical relationship between social policy and economic policy has been seriously challenged. For instance, the United Nations Research Institute for Social Development has launched a series of publications that are in line with welfare state renewal literature. Much work under the heading of Social Policy in a Development Context is organized in such a way as to point to the interlocking relationship between economic policy and social policy. Social policy goals are not necessarily in conflict with economic goals. Rather, they can be merged to build up a modern industrial market economy (Mkandawire 2004; Kangas and Palme 2005; Kwon 2005a).
The experiences of the “late industrializers” in Western Europe show that social policy can be an effective instrument for economic development (C. Pierson 2005). The first social policy measure introduced by Bismarck was intended not only to undermine political support for the socialist movement in Prussia but also to facilitate industrialization. Hence these initial offerings were not designed to meet social needs and protect against social risks but to serve political and instrumental objectives (Rimlinger 1971). The Swedish invention of an active labor market policy in the 1930s was to battle against mass unemployment, not by providing income maintenance through creating social insurance but by equipping the recipients with the necessary skills to enable them to enter the labor market, thereby contributing to economic development (Dahl et al. 2001; Kangas and Palme 2005). Esping-Andersen calls this “Sweden’s celebrated ‘productivistic’ social policy; i.e. a pre-emptive policy to ensure that the labor market produces welfare and equality rather than social risks and poverty” (1997:181). Like many developing countries, Nordic countries, too, in their early years of industrialization used their pension funds as a source of capital accumulation. In Sweden, for instance, supplementary pension funds were used to provide housing in urban areas, while the national pension funds in Finland were used to build up national basic infrastructure (Kangas and Palme 2005). Both the Bismarckian tradition and the Scandinavian social democratic tradition share this productivism: while the former represents traits of selective social investment in an authoritarian context, the latter portrays universal social investment in a democratic setting (Kwon 2005b).
What often characterizes developing countries is not simply a lack of democracy but also, more importantly perhaps, the absence of a legitimate state and a pervasive formal labor market, both of which are often considered to be prerequisites for the development of social policy. The state in the South often lacks political legitimacy, hence being only a weak institutional actor. Importantly, statutory social policy institutions are weakly rooted and most people remain peripheral to the sphere of formal waged employment within a well-regulated capitalist economy. All the key ingredients that define the welfare state seem to be missing in the South. This may have been part of the reason why the literature on welfare state development has been “surprisingly insular, geographically confined to the advanced industrial states, and more particularly to an understanding of the European experience” (Haggard and Kaufman 2008:21). In recent years, however, much progress has been made in advancing our understanding of the ways in which welfare states in other regions have developed.
A notable trend is the rise of intellectual interest in the relation between development and social policy and the increasing interface between social policy and economic policy. The renewed interest in the integration of social and economic policy is not new. In fact, it has long been argued that social policy is and should be an integral part of development (Myrdal 1974; Hardiman and Midgley 1982; Midgley 1995; Hall and Midgely 2004). A similar argument, yet with a much more systematic approach and analytic rigor, has been made by Wood and Gough (2004), arguing that although the “formal security of welfare” of the West is “the most satisfactory way of meeting the universal need for security and well-being” ( 2004 :324), transforming the Global South’s “insecurity and welfare regimes” along these lines is neither feasible nor desirable, not least because they are constructed in relation to the existing structural division of societal power particular to them. Adapting the Esping-Andersen’s model of welfare state regimes (1990; 1999), Gough et al. (2004) have developed a more universal welfare regimes framework, identifying two further families of regime types: informal security regimes and insecurity regimes. In each, the relative strengths of key institutional components (state, market, community, and household) determine the dominant livelihood strategies pursued. Together with the welfare outcomes (e.g., human development index, millennium development goals, need satisfaction, subjective well-being), this in turn shapes the ways that populations are stratified and the dominant forms of social and political mobilizations by shaping the means of recuperation open to different groups through exit, voice, or loyalty. These reproduction consequences then reinforce the institutional structures, thus generating a path dependent form of development. Yet they can also undermine the macro-institutional conditions, thus setting in motion a process of change. In the context of the South, social policy can act as a means to an end, improving human welfare and contributing, both positively and negatively, to shaping and maintaining social structure and to distributing political power.
Compared to the work of Gough et al., Haggard and Kaufman present a perspective that is much more historically oriented and driven by political economy in their exploration of distinctive welfare trajectories in middle income countries of Latin America, East Asia, and Eastern Europe. Core to their explanatory account is an “extended historical consideration of how critical realignments, development strategies, and regime type interacted in altogether different regional settings” ( 2008 :19–20). The key set of analytical factors is largely drawn from the ones that identified the experiences of the advanced welfare states: the significance of distributive coalitions and economic interests; the performance of the economy and its organization; and political institutions, with particular focus on the influence of regime type. There are a number of forces at work which distinguish the developmental trajectories of welfare in one region from another. Haggard and Kaufman argue that “no single prescription – with respect to parties, interest groups or even social policies themselves – will provide a template for addressing the problems of injustice, inequality, and poverty that plague us. The opportunities to rectify past injustices are inevitably constrained by history” (2008:363).
It should be clear by now that no single determining factor explains welfare state expansion, retrenchment, and change. Over time, a number of factors have been suggested, tested, and modified to prove their analytic power in explaining welfare state change. The intellectual progress on the development of welfare owes much to the multiplicity of various factors at work, with varying degrees of influence, over an extended period of time, and sitting at the centre of this are those who take the historically informed institutional analysis seriously. Haggard and Kaufman’s inspiring work reminds us how fruitful the theoretical synthesis of previous comparative welfare state research can become. The application of the usual suspects in accounting for welfare development in advanced industries has been extended to wider contexts, generating lively debates on the possible way(s) forward for broadening our understanding of development, welfare policy, and the welfare state. This is critical not least because it has important implications for validating the positivity between the labour power and the progressivity of the welfare state, the benefits of democracy, and the possibility of a broader range of economic alternatives (Haggard and Kaufman 2008:21–4). The merits of the power-resource approach, the effects of economic structure and performance on welfare commitments, and the enduring effects of institutions, all of these powerful explanatory variables will continue to enter the new testing ground. And most likely, it will be the comparative research in a historical setting that provides the context within which future possibilities are tested, both in theoretical and in empirical terms.
Abbott, A. , and DeViney, S. (1992) The Welfare State as Transnational Event: Evidence from Sequences of Policy Adoption. Social Science History 16 (2), 245–74.