Welfare State Retrenchment: Policy Feedback and Theoretical Debate

Introduction: Policy Feedback and Welfare Theory

With the analysis of policy feedback, we are able to treat state policies as both the starting points and the end points of analysis. This approach becomes crucial when examining the assertion that "a theory which seeks to explain welfare-state growth should also be able to understand its retrenchment or decline" (Esping-Andersen, 1990: 32). The question frames a fundamental scholarly debate: do welfare expansion and welfare retrenchment operate through the same political and institutional mechanisms, or do they require fundamentally different theoretical explanations?

This paper examines the major theoretical positions in this debate and evaluates the empirical evidence offered by each school of thought. The discussion begins with Esping-Andersen's foundational claim that symmetric mechanisms underlie both processes, proceeds through Paul Pierson's influential counter-argument that retrenchment involves a distinct political logic, and then considers how subsequent scholars have challenged, refined, or partially endorsed these positions. The central tension concerns whether welfare state resilience reflects institutional path dependency, shifts in political coalitions, or the persistence of underlying class interests.

Esping-Andersen and the Class-Based Approach

Esping-Andersen (1990: 32–33) argues that welfare retrenchment is characterized by the same mechanisms that have driven welfare expansion. He asserts that any theory offering explanations for welfare growth must also account for welfare decline. According to Esping-Andersen, the threat of retrenchment originates not in mounting fiscal pressures on maturing welfare states, but rather in the inherent class character of society.

In his framework, the extensive welfare regimes of social democratic and corporatist societies rest on the support of the broad middle class, whereas the residual regimes of liberal societies enjoy, in numerical terms, far less backing. Robert Cox (2001: 468) supports this general proposition, arguing that middle-class support has been determining in shaping welfare state policies during both expansion and contraction. Cox contends that the strength of the middle class acted as an anchor on right-wing retrenchment efforts in times of growth while forcing left-wing parties to adopt reform strategies in times of decline. This class-based view suggests continuity in the political forces driving welfare change, regardless of whether the state is expanding or contracting its role.

Pierson's New Politics of the Welfare State

The "new politics of the welfare state" (Pierson, 1996) stands as the principal body of arguments in opposition to the class-based approach. Pierson argues that the political environment is profoundly different in times of retrenchment compared to times of expansion, largely due to differences in popular support for welfare policies. The fear of electoral costs associated with unpopular cuts in social spending leads to strategies of blame avoidance. According to Pierson, retrenchment—whether driven by ideology or economic necessity, whether pursued through radical restructuring or gradual reform—is only possible against a backdrop of budgetary constraints. Pierson calls for two distinct spheres of explanatory theory to address welfare expansion and retrenchment separately, a position fundamentally at odds with Esping-Andersen's symmetry thesis.

Pierson argues that existing theories on welfare expansion—the logic of industrialism, institutionalism, and the power resource model—all fall short of explaining retrenchment (1996: 148–154). Notably, he contends that the power resource model fails to explain why welfare states under right-wing political leadership have escaped widespread retrenchment. Pierson offers the cases of the Thatcher and Reagan governments as evidence. Modest decline in public support for the welfare state preceded both administrations, but public support recovered at the first sign of true retrenchment (1996: 162, 165). In the United Kingdom, retrenchment was limited to housing reform, while in the United States it took the form of non-expansion. Gradual reform rather than extensive cuts characterized welfare retrenchment in both cases (1996: 161–162, 166).

Pierson introduces the concept of path dependency to explain this resilience. He argues that welfare programmes create a proliferation of client interest groups who will rise to defend the welfare state from retrenchment efforts. The fact that beneficiaries of social programmes in some welfare states comprise as much as half of the electorate helps explain the welfare state's resilience (1996: 146–151). This institutional lock-in effect means that once programmes are established and constituencies mobilize around them, political actors face severe constraints on retrenchment regardless of fiscal pressures or ideological preferences.

The virtue of path dependency, however, has been called into question by evidence of successful reform programmes in the welfare states of Denmark and the Netherlands. Robert Cox (2001: 471) suggests that the conservatism of the status quo inherent to path dependency is not insurmountable, as demonstrated by his chosen case studies. These examples show that welfare retrenchment can occur even when client interest groups are well-organized and the electorate benefits substantially from existing programmes.

Path Dependency and Its Limits

Conversely, the lack of empirical evidence prompted Brooks and Manza (2006) to undertake research claiming that causal factors exist between the resilience of welfare states and mass policy preference. By devising a method that allows for simulating a hypothetical change in the political environment of Norwegian welfare policy, they argue that had public support been more akin to that of residualist welfare regimes, the Norwegian welfare state would have been subject to severe retrenchment. Path dependency, according to Brooks and Manza (2006: 825), is indeed a significant factor—but only in interaction with public opinion. This finding suggests a middle position: institutional structures matter, but they are not deterministic if popular attitudes shift sufficiently.

Elinor Scarbrough (2000) offers a fundamental criticism of Pierson, arguing that his central thesis—that welfare expansion and welfare retrenchment can only be explained separately—is fundamentally flawed. Scarbrough proposes a neo-Marxist approach that, she argues, transcends Pierson's divide. In this framework, the need for capital to quell class conflict and ensure political stability contrasts with labour's need to limit the influence of capital. The state thereby becomes mediator, and the emergence of the welfare state signifies the compromise between these two forces.

Scarbrough's Neo-Marxist Critique

Unable to provide an equitable equilibrium, tensions over social expenditure tend to rise as the state attempts to balance spending levels to prevent both undercutting profits and inciting social unrest. The state's inability to satisfy either capital or labour results in the inevitable breakdown of the established compromise—the welfare state itself (Scarbrough, 2000: 233–234). Scarbrough articulates her central argument thus: "the constellation of societal problems and political forces that shaped the fundamentals and expansion of the welfare states also shape the politics of retrenchment" (2000: 227).

While Scarbrough recognizes existing social groups as a barrier to retrenchment, she questions the validity of path dependency by arguing that an inverse relationship exists compared to Pierson's theory. Rather than emerge because of the welfare state, these interest groups were instrumental in creating it and are not a unique phenomenon indicative of a new political reality (Scarbrough, 2000: 247). In general, Scarbrough is much less willing to accept that circumstances framing the political environment have undergone a profound transformation. Instead, she points to the continuity of political reality and supports this claim by stressing the persistent potency of trade unions despite declining membership and by disputing the notion of left-wing party decline (2000: 243–244).

Clayton and Pontusson (1998) raise different concerns aimed at Pierson's findings that welfare retrenchment has been characterized by slow, incremental regress rather than radical decline. They pose a valid question: what constitutes radical regress? Failing to define this, they argue, is a fundamental flaw in Pierson's work (1998: 69). They suggest that rising inequality and the decline in job security are completely ignored by Pierson because these developments are not products of direct welfare policy impacts. According to Clayton and Pontusson, Pierson thus fails to recognize the influence of welfare states on the job market and how it operates (1998: 72).

Labour Market Policy and Hidden Retrenchment

This observation is supported by the findings of Walter Korpi (2003), who argues that welfare retrenchment in Western Europe has been severely underestimated due to failure in identifying shifts in labour market policy. An initial central feature of European welfare states, Korpi argues, was the commitment to full employment policies, which in recent decades have been largely abandoned in response to increased worker involvement and upward pressures on wages. The oil crisis of the 1970s, according to Korpi, provided the pretext for a shift toward vested business interests and the end of full employment policy (2003: 594–596).

This claim is mirrored in the work of Myles and Quadagno (2002: 44), who stress that the state of the global economy in the 1970s—marked by stagflation—offered a favourable environment for those critical of expanding public welfare provisions. Zylan and Soule (2000) refer to developments in the United States during the 1990s, arguing that welfare retrenchment has proven linkages to the decline of union organization and may be explained, in part, by capital's backlash against traditional working-class mobilization in support of welfare expansion (2000: 641). Zylan and Soule further argue that while welfare expansion happened at least partly due to left-of-centre politics, welfare retrenchment has principally emerged from right-of-centre politics (2000: 639), which indicates their belief in the merits of the power resource model.

The "new politics" as advocated by Pierson is not entirely dismissed by either Myles and Quadagno or Zylan and Soule. The effects of globalization—the deregulation of financial markets, the increase in capital mobility, and the emergence of the competition state—are conditioned upon political institutions, Myles and Quadagno argue. The downward pressure on the welfare state is filtered through political interests, which act as a barrier to welfare convergence and the "race to the bottom" (2002: 44). Politics matter and have provided the foundation for how a transformation in the political reality has taken place.

Conclusion: Convergence and Remaining Questions

While Myles and Quadagno (2002: 51–52) move beyond Pierson's analysis in arguing that corporate interests now influence the decision-making process to a much greater extent, their conclusion grants Pierson's fundamental thesis: that the political reality facing policy makers in times of welfare expansion is profoundly different from that in times of retrenchment. Zylan and Soule find that Pierson is right to point to a change in the political environment and concede that their study indicates the emergence of new institutional characteristics. However, they stop short of accepting Pierson's full theory. Instead, they conclude that the "new politics" is not a dramatic departure from existing theories and that the mechanisms driving welfare retrenchment remain fundamentally the same as those driving welfare expansion (2000: 644). The debate thus remains open, with scholars finding selective support for competing frameworks while acknowledging the genuine complexity of welfare state change.