Political Theories of the Welfare State
Author(s): John Myles and Jill Quadagno
Source: Social Service Review, Vol. 76, No. 1, 75th Anniversary Issue (March 2002), pp. 34-57
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/324607 .
Accessed: 23/11/2014 11:22
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Social
Service Review.
http://www.jstor.org
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
Political Theories of the
Welfare State
John Myles
University of Toronto
Jill Quadagno
Florida State University
We review the main theoretical conclusions from a quarter century of comparative studies
of welfare states in the affluent democracies. We contrast early debates over the relative
importance of industrialization, economic growth, and social classes for explaining welfare
state differences with contemporary claims about the role of globalization, postindustrialism, and gender relations in shaping their futures. We evaluate the claims against recent
empirical evidence with the aim of highlighting both important lessons from the past and
promising directions for future analysis.
Contemporary studies of the modern welfare state came of age in the
1970s. Historians, political scientists, and other social scientists had of
course long since documented the origins and development of particular national social policies. The comparative focus of this new body of
research was prefigured by a number of seminal studies written during
the postwar decades (Wilensky and Lebeaux 1958; Kerr et al. 1960; Pryor
1968; Rimlinger 1971; Heclo 1974). And during the 1970s, several classic
theoretical statements (Polanyi 1944; Marshall [1949] 1964) were rediscovered. But it was only in the mid-1970s, following more than 2
decades of accelerated expansion, that both the sheer scale of, and
striking diversity among, the welfare states of the industrialized nations
made them the focus of concentrated attention as a defining feature
of so-called modernity. In 1960, average expenditures on social transfers
were 7.5 percent of gross domestic product (GDP) in the affluent deSocial Service Review (March 2002).
᭧ 2002 by The University of Chicago. All rights reserved.
0037-7961/2002/7601-0003$10.00
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
Theories of the Welfare State
35
mocracies, only slightly more than the 6 percent being spent by the
United States.1 By 1980, average expenditures on social transfers had
doubled to 14 percent of GDP, and the diversity among countries was
much more in evidence. Swedish expenditures had risen from 7 percent
to over 17 percent while the United States was spending 9.75 percent.
How had this happened? Although the institutional foundations had
often been established decades earlier, it was the wave of reforms from
the late 1950s through the early 1970s that account for the dramatic
increases in spending over the period. Like the United States, many
countries established new social insurance schemes and national social
service programs in health and welfare virtually ex nihilo. If one adds
the rapid expansion of education and health care to the brew, the
change that had occurred in both the scale and scope of public provision
in such a brief time span was heady stuff indeed.
From the outset, two questions were central to the research agenda:
How to explain these developments? And what were their consequences?
As we show below, we know a great deal more today about both issues.
Our review of a quarter century of welfare state scholarship, however,
is not primarily intended as a piece of intellectual historiography but
rather to shed light on the present. The new politics of welfare, it is
claimed, are different (Pierson 1994, 1996). Unlike the golden age of
expansion, the social policy agenda of the late twentieth century has
been shaped by the “politics of austerity.” The forces of globalization
and postindustrialism, the revolution in family forms and gender relations, and an extended period of modest economic growth have created
a very different social and political climate from that in which contemporary welfare states came to maturity between the 1950s and the 1970s.
We do not dispute all this. There are, however, striking parallels between
the old and the new scholarship on welfare states that are worth highlighting. Lessons learned from the first generation of welfare state studies, we will argue, can shed considerable light on the present.
From the mid-1970s to the early 1990s welfare state research concentrated on the long slow growth of the social programs associated with
Bismarck’s Germany in the 1880s to the postwar boom in welfare state
expansion (the period of high industrialism) that came to maturity (and
to an end) in the mid-1970s. The puzzle to be solved was less about
why welfare states developed than with why they had developed in such
different ways and apparently reached their apogee at such strikingly
different levels of spending. This body of scholarship can usefully be
bracketed by two highly influential works: Harold Wilensky’s (1975) The
Welfare State and Equality, emphasizing the determining role of impersonal economic forces, and Gosta Esping-Andersen’s (1990) The Three
Worlds of Welfare Capitalism, where politics and political institutions play
the leading role.2
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
36
Social Service Review
Welfare State Politics in Industrial Societies: The Political
Regulation of Large Economic Forces
The first generation of welfare state studies typically turned to theories
of industrialism to account for the common trajectory of rising welfare
state expenditures throughout the developed world (Wilensky and Lebeaux 1958; Kerr et al. 1960; Pryor 1968; Rimlinger 1971; Wilensky
1975). The main themes are well known. Industrialization creates new
demands for public spending as systems of social support based on
kinship and the patrimonial traditions of agrarian societies are eroded
(Kerr et al. 1960). Growing dependence on wage labor creates new
vulnerabilities among those with little or no labor to sell, such as the
old, the sick, and the young (Pampel and Weiss 1983). The result, as
Clark Kerr and colleagues (1960, pp. 22, 152) observe, is a new and
expanded role for the state to maintain the labor force and achieve
coordination and consensus in a complex, urban society, themes that
were echoed, albeit in different language, in early marxist accounts
(Offe 1972; O’Connor 1973) discussed below.
Wilensky’s 1975 study became central to future developments not only
because it was among the first to test empirically alternative theories of
welfare state expansion but, more important, because of the bold conclusions advanced in the early chapters of the work. The root cause of
the welfare state, Wilensky concluded, was economic growth mediated
by demographic change, which resulted in rising life expectancy and
population aging: “If there is one source of welfare spending that is
most powerful—a single proximate cause, it is the proportion of old
people in the population” (Wilensky 1975, p. 47). This prescient statement would foreshadow trends underlying welfare state restructuring.
It is helpful to distinguish between weaker and stronger versions of
the theory. The weak version of the theory, that industrialism and its
correlates (economic growth, population aging) are necessary to account for the common trend line in welfare state expansion, is rarely
questioned. It is difficult, moreover, to imagine the welfare state boom
of the 1960s in the absence of the long economic expansion of the
post–World War II golden age. Rapidly rising incomes generated revenue windfalls for governments. Strong wage growth meant new taxes
could be levied while still leaving workers with a larger pay package
each year. Rising productivity and changing labor force practices led to
the spread of retirement, a development that generated enormous demand for the expansion of public pensions. These correlates of a mature
industrial order clearly matter, and measures of GDP per capita and
percentage of elderly are now standard control variables in all empirical
models of welfare state spending
The contested (strong) version of the theory concerned its capacity
to explain variations around the trend line. The implicit claim of the
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
Theories of the Welfare State
37
strong version is that nations with comparable levels of economic development would converge at similar levels of welfare state development
as well. The claim is contentious since it rests on the assumption that
public policy is the product of large, impersonal, economic forces. Politics, if it matters at all, does not matter very much. Faced with comparable political demands (e.g., a growing number of old people) and
a similar resource base (economic development), social programs develop independently of party politics, the preferences of political leaders, or the balance of power among the political forces they represent.
Political Parties and Institutions
Where industrialization theorists emphasized the determining role of
the changing forces of production (the logic of industrialism), early
neomarxist accounts (Offe 1972; O’Connor 1973) emphasized changing
relations of production (the logic of capitalism). Welfare state expansion, James O’Connor (1973) argued, was driven by the dual, if contradictory, imperatives imposed on the capitalist state to create conditions for capital accumulation, on the one hand, and the social
legitimation of this mode of production, on the other. While differing
in content, both theories shared a similar functionalist logic (Giddens
1976, pp. 716–22; Gough 1979, pp. 8–9, 50 ff.). Welfare states are the
inevitable product of large economic forces beyond the control of policy
makers and publics that compel a common response.
By the end of the 1970s, however, class-analytical accounts (e.g., Stephens 1979; Korpi 1980, 1983) took a different turn, focused around
the claim that “politics matters” in explaining welfare state diversity.
What came to be known as “power resource theory” was based on a
theory of distribution in capitalist democracies in which both terms of
the couplet—capitalism and democracy—figure prominently. In his
magisterial Power and Privilege (1966), Gerhardt Lenski argued that democratic polities created the possibility for the “many” to combine against
the “few” (the elites) and use the state to claim a larger share of the
social surplus. However, the institution of private property biases this
struggle in favor of the propertied simply because electoral constituencies are large and elections costly. For those of modest means (workers), the only compensating power resource is their numbers, but this
requires mass organization in unions and parties to be effective. The
key claim is that the formal institutions of parliamentary democracy,
including universal suffrage and free and competitive elections, are necessary but not sufficient conditions to generate significant redistribution.
As Christopher Hewitt (1977, p. 451) argued, the key issue is what the
mass electorate does with the franchise. Only if the lower classes use
their votes to elect explicitly class-based (i.e., labor) parties to represent
their interests will democracy result in more equality.
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
38
Social Service Review
Walter Korpi formalized these general claims into what he identified
as “power resource theory,” and power resource theory quickly achieved
the status of the dominant paradigm (Orloff 1993) in the field, providing
the benchmark models against which other theories would be tested.
More important, Korpi and others initiated a program of both quantitative and historical-comparative research to test the thesis empirically
(Cameron 1978; Stephens 1979; Castles 1982; Hicks and Swank 1984;
Myles 1984; Esping-Andersen 1985).3 These and a plethora of later studies in this tradition (Korpi 1989; Palme 1990; Kangas 1991; among many
others) supported the conclusion that major differences in welfare state
spending and entitlements among the capitalist democracies could be
explained by the relative success of left parties, particularly Social Democratic parties, aligned with strong trade unions in shaping the democratic class struggle.4
Careful cross-national research (e.g., Huber, Ragin, and Stephens
1993; Hicks 1999) continues to document the importance of workingclass mobilization (in unions and parties) as a condition for early (e.g.,
by 1920) welfare state consolidation and for explaining national differences in their subsequent expansion. More recent research also highlights the distinctive role of social Catholicism and Christian Democratic
parties (see below) in generating high levels of social spending, particularly when faced with intense competition from left parties and unions
for the allegiance of working-class voters (Van Kersbergen 1995). Theories of political democracy that emphasize the intensity of electoral
competition among parties and voter turnout also find support in the
literature (Hicks and Swank 1992). As Frances Fox Piven and Richard
Cloward (1994, pp. 428–33) highlight, high voter turnout typically
changes the class content of elections, shifting the political center of
gravity to the left, since increases in turnout tend to reflect increases
in participation by previously excluded lower status groups (Iversen
2001).
During the 1980s, Theda Skocpol and her colleagues (Weir, Orloff,
and Skocpol 1988; Skocpol 1992; Amenta 1998) took the “politics matters” perspective in a different direction, aimed in part at accounting
for the distinctive development of U.S. welfare institutions. In contrast
to so-called society-centered approaches that emphasize the role of elections and parties aggregating interests from below, they elaborated a
polity-centered or institutionalist approach emphasizing the organization and structure of state institutions.5 In their account, for example,
it is not just the balance of class forces that determines electoral and
policy outcomes but also the institutional features of government and
the rules of electoral competition. In the United States the diffusion of
government authority—to a degree unmatched in Europe—has placed
limits on what political leaders can accomplish (Lipset 1996; Lipset and
Marks 1999). The multiplicity of “veto points” in a system characterized
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
Theories of the Welfare State
39
by a division of power between the executive, the legislative branch, and
the courts, coupled with a weak party system and an emphasis on state
autonomy, makes radical legislative innovation difficult and political
gridlock common. For example, the diffusion of veto points made it
possible for Southern politicians to make the racial divide an overdetermining factor in shaping American social policy legislation from Reconstruction to the 1960s (Quadagno 1988).
The polity-centered perspective has also influenced comparative research of the welfare state. Ellen Immergut’s (1992) study of health insurance in Sweden, France, and Switzerland shows the importance of
centralization and the insulation of the executive from parliamentary and
electoral pressures as a precondition for reform. Antonia Maioni’s (1998)
comparison of Canada and the United States shows how the Canadian
parliamentary system with its strong party discipline diminished the ability
of medical lobbies in their efforts to oppose health-care reform. Evelyne
Huber and colleagues (1993) show a strong independent effect of constitutional structures that concentrate state authority with pooled timeseries models for 18 capitalist democracies. The ability of working-class
organizations to shape policy outcomes has been unambiguously shown
to be mediated by their success in institutionalizing corporatist decisionmaking structures, that is, peak bargaining institutions of labor and business that concentrate and monopolize the representation of their constituencies and engage in state-sponsored negotiations over wages, jobs,
and other key economic transactions (Hicks 1999).
Fred Pampel’s (1994) analysis of social spending on the old and the
young illustrates the utility of this approach. He asks whether the growing political power of the aged has skewed social spending in favor of
the elderly and against the young. He demonstrates that, in general,
the answer is no: countries that are generous to old people are also
generous to children. Nevertheless, where systems of corporatist representation (and strong labor parties) are absent, as in the United States,
the trade-off between old and young does appear.
Welfare Regimes
Perhaps the most persuasive evidence that industrialization is not destiny—that politics and political institutions matter for welfare
states—came from Esping-Andersen’s highly influential The Three Worlds
of Welfare Capitalism (1990). Taking stock of the welfare states of the 18
or so rich capitalist democracies in the early 1980s, Esping-Andersen
demonstrated that they could be distinguished not only in terms of
relative generosity and spending but, more fundamentally, by their institutional logic for assigning welfare functions to the state, the market,
and the family. He identifies three distinct welfare state regimes: liberal
(or market-oriented) welfare states characteristic of the Anglo-American
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
40
Social Service Review
democracies, the conservative regime characteristic of continental Europe, and the Nordic social democratic policy model.
In liberal welfare states, citizens are constituted primarily as individual market actors. There is a reluctance to replace market relations
with social rights, and citizens are encouraged to seek their welfare
in the market, for example, through subsidies for private welfare benefits. Basic security schemes are more likely to be means-tested and
social insurance benefits modest. The main exemplars are the AngloAmerican democracies.
The second regime type goes under a variety of labels—conservative,
corporatist, and more recently, Christian Democratic—depending on the
characteristics emphasized. It is conservative in the classic European sense
of the term, highlighting its precapitalist origins in the dynastic elites of
continental Europe (Kerr et al. 1960). It was decidedly antiliberal in
origin, concerned not at all with market efficiency but deeply concerned
with maintaining an organic-hierarchical social order inherited from the
past. Social rights are extensive, and there is only a marginal role for
private welfare arrangements (the market). It is corporatist in the sense
that, in their origins at least, these rights and privileges were differentiated
on the basis of class and status, and redistribution was marginal. These
states were also strongly influenced by Christian Democratic (i.e., Catholic) doctrines rejecting the primacy of the market, on the one hand,
but insisting on principles of subsidiarity and, hence, the primacy of the
family as the locus of social welfare on the other. While social spending
in these nations is considerably higher than in the liberal welfare states,
the emphasis is on income transfers sufficient to cover the income needs
of the male breadwinner. Social services that both facilitate women’s employment (child care) and provide jobs for women are modest. The main
exemplars are found among the continental European countries.
Social democratic welfare states (found mainly in Scandinavia) represent a model of society characterized by extensive social rights and a
marginal role for private welfare provision (as in continental Europe).
But rights are universalistic rather than corporativistic, emphasizing
equality of citizenship rather than preservation of status differences (as
represented by occupational groups). Hence, unlike the Continental
model, there is considerable emphasis on redistribution and providing
all with high levels of income security. Women’s place is in the labor
market so that the welfare state is service intensive as well as transfer
intensive. Social services provide both employment for women and the
child care and other services required to allow women or parents to
participate in the labor market (Huber and Stephens 2000). Both social
democratic and conservative welfare states tend to spend a lot on social
protection but do so in fundamentally different ways.
Finally, welfare state regimes differ not only with respect to the ways
in which they respond to inequalities generated by the market but also
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
Theories of the Welfare State
41
in the ways they condition and regulate the primary distribution of
income (wages) and job insecurities that the market is allowed to generate. Both the Nordic and continental European nations have created
bargaining structures that limit wage dispersion and regulatory institutions that limit the rights of employers to hire and fire at will.
Although Esping-Andersen’s typology has been subject to considerable criticism for ignoring and oversimplifying complexities within regime types (Castles and Mitchell 1992; Leibfried 1992) and for overlooking aspects of the gendered logic of welfare regimes (Ostner and
Lewis 1995; Sainsbury 1996; O’Connor, Orloff, and Shaver 1999), the
key insight that the welfare states that came to maturity by the 1970s
differ fundamentally in the allocation of welfare functions among states,
markets, and families has proven to be remarkably robust (Orloff 1996).
Yet at the very moment when political theories of the welfare state
seem to have relegated the “logic of industrialism” thesis with its emphasis on the overdetermining role of large impersonal economic forces
to the critical list, the theory was revived in new form. Economic globalization and postindustrialism (along with its demographic correlates)
are the new forces thought to be reshaping welfare states. However
important political parties and institutions were in accounting for the
development of welfare states, these same parties and institutions, it is
now argued, are comparatively helpless in preventing the welfare state’s
demise or, at least, its dramatic restructuring.
The New Politics of Welfare: Globalization,
Postindustrialism, and Gender
Contemporary welfare states grew up, as it were, during the quarter
century following the Second World War, and by the mid-1970s the
major contours of the national social institutions familiar to us today
were in place. Almost immediately, however, observers from both the
left (O’Connor 1973) and the right (Huntington 1975) began to announce their impending demise. For some, such as Daniel Bell (1978),
welfare states had created a revolution of rising expectations that could
not be satisfied without destroying the conditions that allow markets to
function (Geiger and Geiger 1978). This pessimism was reinforced by
slow economic growth, rising unemployment of the 1970s and 1980s,
and by the impending threat of population aging. Has this dire prediction been realized?
Unlike the heady years of expansion, the period since the end of the
1970s can broadly be characterized in Paul Pierson’s (2001) apt phrase
as one of accommodation to austerity. Governments everywhere grappled with slower economic growth, higher levels of unemployment, rising deficits, and growing awareness of the fiscal implications for pension
This content downloaded from 132.205.7.55 on Sun, 23 Nov 2014 11:22:38 AM
All use subject to JSTOR Terms and Conditions
42
Social Service Review
and health-care budgets that result from population aging. Rollbacks
have been commonplace…
