. 2
( 11)


ing, and if individual women do not have access to effective commitment
mechanisms, all women will be treated as a less valuable source of labor for
employers. As argued by Estevez-Abe (1999b, 2002), however, this disad-
vantage clearly depends on the skills involved. If employers can easily ¬nd
workers with the skills they need in the external labor market, career inter-
ruptions are of less consequence. Indeed, in a competitive neoclassical labor
market, every employed worker has a perfect substitute willing to work for
the same wage (it is labor L). Any employee is therefore readily replaceable.
When employees bring skills that cannot be easily substituted (Ls ), career
Welfare Production Regimes

interruptions become critical to the employer. In terms of providing train-
ing and offering promotions, employers in this setting have an obvious in-
centive to discriminate against women. This is reinforced by women™s own
investment decisions because the returns on speci¬c skills will be negatively
related to the prospect of interrupted careers. This implies a heavily gen-
dered structure of educational choices that is probably reinforced through
socialization (Estevez-Abe 2002).
Let us look at the outcomes. Figure 1.5 relates a measure of the special-
ization of skills in different occupations to the percentage of women in these
occupations using the International Labour Organization™s (ILO™s) Interna-
tional Standard Classi¬cation of Occupations (ISCO-88).12 The numbers
are averages for the thirteen countries where comparable ISCO-88 data are
available. Bolded occupations are those that have disproportionately large
numbers of low-skilled and low-paying jobs.
As expected, there is a strong negative relationship with men dominat-
ing occupations that, as implied by this measure, require highly specialized
skills. This basic pattern is repeated in all thirteen cases. Not surprisingly oc-
cupations with specialized skills are found in agriculture and manufacturing
rather than in services. Also note that because men on average participate
more in the labor market than women do, most occupations have an un-
derrepresentation of women. Women are also overrepresented in jobs that
require relatively low general skills (and that tend to pay low wages), but
women have at the same time achieved near-parity with men in professional
occupations (in the United States there is complete gender parity among
technicians and associate professionals).13
An important part of the gender story that is not captured by Figure 1.5
is the cross-national variance in the labor market position of women.
Numerous studies document such cross-national variance (e.g., Melkas and

12 Ignoring military personnel, ISCO-88 contains nine broad occupational groups, which
are subdivided into numerous subgroups depending on the diversity of skills represented
within each major group. At the most detailed level, there are 390 groups, each supposedly
characterized by a high degree of skill homogeneity. By dividing the share of unit groups in
a particular major group by the share of the labor force in that groups, we can get a rough
measure of the degree of speci¬city of skills represented by each major group. See Chapter 3
for details.
13 Within the professions, however, it is still the case that women tend to be in jobs and
positions with lower pay and skills “ for example, teachers in primary and secondary ed-
ucation rather than in higher education, nurses rather than doctors, and junior rather
than senior associates “ and they are notably underrepresented among senior of¬cials and

A Political Economy Approach to the Welfare State

Anker 1997; Chang 2000; Breen and Garcia-Penalosa 2002; Porter 2002).
The evidence suggests that speci¬c skill systems exhibit high levels of occu-
pational gender segregation (Estevez-Abe 2002). Comparative survey data
from the International Social Survey Program (ISSP), for example, show
that the correlation between vocational training activity and women™s share
of private sector employment is ’.71 for twelve Organization for Economic
Co-ordination and Development (OECD) countries in 1997 (ISSP 2000).
The weak position of women in the private labor market in speci¬c skills
countries is modi¬ed by public policies deliberately designed to counter
them, especially public provision of social services and hence employment
opportunities for women (Orloff 1993). These public policies enable the
Scandinavian countries to have high female participation rates despite hav-
ing a male-dominated private sector (Esping-Andersen 1990, 1999a; Iversen
and Wren 1998). At the same time, however, women in countries with large
service-oriented welfare states become “ghettoized” into the public sector
instead of competing equally with men for the best private sector jobs.


r = ’.74
Service and
sales workers
Percentage of women

50 and associate
senior officials,
and managers
Plant and
Craft and operators
20 workers
0 0.5 1 1.5 2 2.5
Skill specificity
Figure 1.5 Skill speci¬city and occupational gender segregation, 2000.
Note: Percentage of women refers to the average percentage share of women in different
major categories of ILO™s International Standard Classi¬cation of Occupations across thirteen
countries. The skill speci¬city measure is explained in footnote 12 and in Chapter 3.
Source: International Labour Organization, Labour Statistics Database. ILO 1998“2004.

Welfare Production Regimes

The implication of this logic is that con¬‚ict over public sector service
provision, parental leave policies, and education is likely to be gender re-
lated. Especially in speci¬c skills countries, women will have a strong in-
terest in an active state that provides daycare services and counters the
disadvantages of women in the private job market. Also, women depend
on generous leave policies to balance family and work; however, for such
policies not to put women in a serious disadvantage in the competition for
the best jobs, men must shoulder some of the burden of caring for infants
(essentially forms of mandatory parental leave). Finally, because women
have a comparative advantage in general skills, we should expect them to
be more supportive of public spending on general education than men. I
explore these hypotheses in Chapter 3.

1.5. Recap
This book builds on new institutionalist theory in both economics and
political science (especially Hall and Soskice™s VoC perspective) to offer a
general political economy approach to the study of welfare capitalism. The
book demonstrates that to a substantial extent social protection in a modern
economy, both inside and outside the state, can be understood as protec-
tion of speci¬c investments by both workers and ¬rms in human capital. It
then shows that such an understanding provides a systematic explanation
for popular preferences for redistributive spending, the economic role of
political parties and electoral systems, and labor market strati¬cation (in-
cluding gender inequality). In doing so, it helps resolve the debate between
power resources theory and recent work on employers and the welfare state
by systemically linking demand for redistribution and demand for social
insurance and by tying social protection to the way ¬rms compete in the
international economy. Contrary to the popular idea that global competi-
tion undermines cross-national differences in the level of social protection,
the book argues that these differences are made possible by a high inter-
national division of labor. Such a division is what allows ¬rms to specialize
in production that requires an abundant supply of workers with speci¬c (or
general) skills, and hence high (low) demand for protection.
The rise of nontraded services, however, undermines this specialization
and leads to pressure for more ¬‚exible labor markets. The reason is
that when speci¬c skills countries can no longer achieve a complete
specialization in production that uses such skills intensely, governments
concerned about employment must create regulatory and institutional
A Political Economy Approach to the Welfare State

conditions that are conducive to the expansion of general skills jobs, es-
pecially at the low end of the wage scale. In turn, this raises the critical issue
of how governments can deregulate a portion of the labor market without
undermining the comparative advantages of generous social protection, and
without causing a signi¬cant rise in inequality.

1.6. Outline of Book
The book is divided into three parts. This ¬rst part is an overview, which
includes this chapter. Chapter 2 presents a comparative-historical analysis
that traces the rise of the postwar social protection system to compromises
between interests rooted in the industrial economy. It also provides much
of the comparative data used throughout the book. Those who are familiar
with these data and economic history can jump to the subsequent chapters.
Part II focuses on the comparative statistics of the argument and explores
cross-national differences and the microfoundations of these differences.
Chapter 3 develops an asset-based theory of the sources of individual social
policy preferences and shows how skills and gender are key in explaining
these preferences. Chapter 4 considers the translation of these preferences
into policies through the political system, emphasizing the role of political
parties and electoral systems. Part III focuses on the dynamics of the argu-
ment. Chapter 5 explains the rise of the welfare state since the 1960s as a
politically mediated outcome of the shift from industry to services. Chapter
6 explores the distributive and political consequences of this shift, as well
as the recent reforms that can be attributed to it.


A Brief Analytical History of Modern
Welfare Production Regimes

The introductory chapter outlined a broad theoretical approach to the study
of social protection and hypothesized close linkages between social pro-
tection, skills, ¬rm strategies, and the political-institutional foundations
of the welfare state. This chapter ties national and international develop-
ments to the emergence of a few national institutional equilibria or ideal
types. Although the chapter will discuss causal mechanisms, the main pur-
pose is to put some empirical meat on the conceptual bones presented
in Chapter 1. I will provide the reader with a range of background in-
formation, and I will show some striking interconnections between skill
systems, social protection, and political institutions that cry out for ex-
planation. Building on a joint paper with Barry Eichengreen, I argue that
these interconnections metamorphosed into very distinct regime clusters
in response to the challenges of postwar reconstruction and international
economic integration, and each cluster is associated with distinct economic
advantages that are reinforced through the international division of labor.
Relative advantages have been shifting over time, however, and I discuss
these forces of change with a view to developing the themes that are ex-
plored in greater depth, and with a sharper analytical knife, in subsequent
Much of the political-institutional divergence that occurred during the
postwar period, I argue, re¬‚ects the structural-institutional potential for
postwar growth in particular countries, the strength of organized labor
and capital, and the inherited capacity for centralized collective action.
In countries where the rewards for solving the distributive and collec-
tive action problems were high and where employers and workers were

A Brief History of Modern Welfare Production Regimes

relatively well constituted as collective actors, solutions emerged that em-
phasized generous levels of social protection, responsible and program-
matic parties, long-term investment in speci¬c skills, and specialization in
high-quality niche markets. In countries where the potential for economic
growth was smaller and where distributive and collective action problems
among unions and employers were intense, policies came to rely increas-
ingly on market solutions. This divergent economic history is the focus of
Section 2.1.
Over time these differences were magni¬ed by economic crisis and in-
ternational trade because intensi¬ed competition encouraged governments
and economic actors to “specialize” in institutions and policies that were
complementary to the production system and therefore reinforced compar-
ative advantages. Section 2.2, which builds on a joint paper with Margarita
Estevez-Abe and David Soskice (Estevez-Abe et al. 2001), provides a com-
parison of some key structural-instititutional features of capitalist democ-
racies “ especially social protection, education and training, and position
in the international division of labor “ and how these features cluster to
form distinct welfare production regimes. Part II of the book is designed to
explain the differences identi¬ed between these regimes. Readers who are
familiar with the main institutional differences across advanced capitalist
democracies can skip this section.
Section 2.3 focuses on the tensions between the political-institutional
frameworks that emerged in the ¬rst decades after the war and changes
in technology and tastes that occurred subsequently. Because institutions
and policies had been built around the economic and political condi-
tions prevailing in the internationally integrated industrial sector, they
were not well adjusted to cope with the decline of industry and the rise
of a mostly sheltered sector of services. Much production in the ris-
ing services did not play to the strengths of the egalitarian and high-
protection systems of continental Europe. In terms of private sector em-
ployment, this caused a rather remarkable reversal of fortunes when we
compare continental Europe to the United States. Equally important, the
transformation of the occupational structure generated new insecurities
for workers with speci¬c skills, and these could not easily be addressed
through the existing employment protection system or through the col-
lective wage bargaining system. The causes of these changes, and the
responses of government to them, will be the focus of Part III of the

Welfare Production Regimes

2.1. Postwar Reconstruction and Institutional Divergence1

2.1.1. The Shadow of History
The remarkable economic success of western Europe after World War II
must be understood against the backdrop of prewar developments.2 In an
important sense, the industrial revolution spread beyond isolated pockets in
northern Europe and North America only in the ¬nal decades of the nine-
teenth century. Only then can it be said that Marx and Engels™s vision of
large-scale manufacturing animated by centralized power, housed in large
factories, and manned by an anonymous proletariat became a widespread
reality. The challenge for the twentieth century was, thus, to solve the prob-
lems of ef¬ciency and legitimacy posed by the spread of this new system, and
this required the creation of institutions and structures through which the
participation and cooperation of the rising industrial working class could
be secured.
Nineteenth-century liberal capitalism in its post-1848 form had been
predicated on limited-suffrage democracy, management control, and de-
centralized labor-market arrangements. By the end of the century, however,
the rise of heavy industry, large corporations, and mega-banks had raised
troubling questions about the prevailing distribution of economic power
and the legitimacy of existing political institutions. Organized labor move-
ments, socialist parties, and Catholic organizations (religious and political)
challenged the legitimacy of both the electoral institutions and the existing
social basis for production.
These were not problems on which Europe made much progress in
the ¬rst half of the twentieth century. The extension of the franchise de-
stroyed the inherited political equilibrium without substituting another.
The creation of new states and the adoption of new electoral systems, lead-
ing to party proliferation, did not simplify reaching mutually acceptable
decisions in a context of mass mobilization; indeed, the opposite was true.
The volatility of party systems and a rapidly changing class structure consti-
tuted an unpredictable environment where far-sighted economic planning
was dif¬cult and where polarizing ideologies could thrive at the expense of
cooperation and compromise. The sad fate of European democracy in the
1930s is testimony to this point.

1 This section builds on Eichengreen and Iversen (1999).
2 The relevant points have been made by Maier (1987) and Toniolo (1996).

A Brief History of Modern Welfare Production Regimes

However disruptive the experimentation between the wars with solutions
to the challenges for ef¬ciency and legitimation posed by the new twentieth
century industrial system, there is an important sense in which it neverthe-
less planted the seeds for success after World War II. The institutions of
economic governance elaborated after 1945 were themselves outgrowths
of these, for the most part, less-than-successful interwar experiments. The
labor codes of the National Industrial Recovery Act in the United States,
the labor policies of the Popular Front in France, and the Saltsjobaden
agreement in Sweden were all efforts to solve the problems of ef¬ciency
and legitimacy of the 1920s and the unemployment crisis of the 1930s. Such
frameworks can be thought of, in part, as concessions to the labor movement
by governments and elite interests seeking to head off more radical alter-
natives (Luebbert 1991). In part, they can be seen as cross-class alliances to
advance the common interest in effective con¬‚ict-resolution mechanisms
and macroeconomic recovery (Swenson 1991a). With the exception of
Austria, this is particularly true in the small European countries where
a divided right and hightened exposure to international competition inten-
si¬ed the search for common ground (Katzenstein 1985). Typically they
involved negotiations with the political arm of the labor movement, which
had acquired a parliamentary presence. PR electoral institutions guaranteed
that they could not be ignored.
The structure of these settlements “ one of industry-level negotiations
conducted under broad guidelines set down by government “ contrasted
with the starkly centralized agreements reached by state unions and industry
organs in Mussolini™s Italy and Hitler™s Germany, whose legacy neverthe-
less also persisted into the postwar era (in the German case, for example,
in the form of a dozen and a half national unions). Neither were the post-
war reliance on ministerial controls over wages and working conditions
and formal and informal incomes policy in, inter alia, France, Italy and the
United Kingdom unprecedented and radically new; these devices were di-
rect outgrowths of experiments with state direction in the 1930s and the
even greater state control made necessary by total war.
It is hard to exaggerate the role of World War II itself as a selection mech-
anism for which of the innovations of the 1930s and early 1940s persisted
into the postwar golden age. The creation of more hierarchical arrange-
ments designed to facilitate the efforts of governments to harness the market
economy for war bequeathed a set of more centralized structures ready to
be applied to peacetime use. Fascism, Nazism, and Bolshevism all worked
to discredit the more radical solutions of left and right. And, of course, the
Welfare Production Regimes

fact that the United States was the only capitalist superpower left stand-
ing was conducive to the adoption of institutional solutions appealing to
American foreign policy makers.

2.1.2. Labor Markets and Economic Growth
Reconstruction in Europe after World War II took place against a back-
drop of capital scarcity and labor militancy. Productive capacity had been
devastated in the war, and many of the conservative political parties and
organizations that were the traditional counterweights to organized labor
had been discredited by their acquiescence to or active participation in
the Nazi war effort. This economic and social disarray was all the more
alarming once the Soviet Union came to be seen as a threat to Western
Europe. For the United States and its European allies, economic growth
promised to solve all these problems in a stroke. It would give West-
ern Europe the economic and military capacity to withstand the Soviet
threat. It would give labor a stake in the market economy. And it would
restore the respectability of the capitalist class and of conservative political
But in order to initiate and sustain economic growth, three interrelated
problems that were highlighted by the polarized and turbulent interwar
experience had to be solved.
1. Short-termism and time inconsistency. Given the destruction of plant,
equipment, and infrastructure, investment was key to postwar recov-
ery. And even after the recovery phase was complete, investment re-
mained central to the process of transferring to Europe the technolo-
gies and mass production methods developed by American industry
in the course of previous years. Given the disorganization of inter-
national ¬nancial markets, investment had to be ¬nanced at home.
Faster growth and higher incomes in the future, thus, required sac-
ri¬ces of consumption in the present. Wages had to be moderated
to free up the pro¬ts to ¬nance capacity modernization and expan-
sion. Those pro¬ts had to be plowed back instead of being paid out
to shareholders. A mechanism had to be created, in other words, to
encourage labor and capital to trade current grati¬cation for future
gains, overcoming the problem of short-termism.
In a closely related process, governments had to commit to policies
that would encourage investments in the future, even while concern
A Brief History of Modern Welfare Production Regimes

for reelection forced them to pay close attention to the present. In
sofar as the government was itself an investor in capital and infrastruc-
ture, one issue was how to make such investments without advantaging
future governments and hurting the government™s own current elec-
toral chances. Another issue was how to tax income without unions
demanding compensation. As is argued later, the provision of wage
protection through the welfare state was part of the solution, but such
protection itself presented a problem. Even if voters understood that
social insurance would be desirable from a long-term perspective, the
inability of current and future voters to make binding agreements
with each other created a time-inconsistency problem. This made
underprovision of social protection a dominant strategy for short-
term-oriented, vote-maximizing parties.
2. Collective-action problems. It is dif¬cult to withhold the bene¬ts of
growth from those who refuse to support it. In the postwar set-
ting, this meant that individual unions inevitably were tempted to
raise their own wages even while bene¬ting from the favorable mar-
ket conditions created by the restraint of other unions. The pro¬ts
freed up by their restraint did not remain in the same sector; rather,
they passed through the national capital market, boosting investment,
productivity, and labor incomes economywide. Firms for their part
were tempted to underinvest in research and development (R&D)
and technical training in the belief that these investments bene¬ted
competing ¬rms that did not help to defray the costs. Therefore,
to sustain economic growth, collective-action problems had to be
3. Distributive con¬‚ict. Like a messy divorce in which the family jewels
are sold off to pay the lawyers™ fees, a society riven by distributional
con¬‚ict will be prone to dissipate the resources needed to sustain
prosperity and growth. In particular, different groups of workers will
only be willing to restrain their wages if they are con¬dent that they
reap a fair share of the bene¬ts of that restraint: both a fair share
vis-` -vis employers, who control the use of pro¬ts, and a fair share
vis-` -vis other groups of employees, who may take advantage of their
restraint sometime in the future. Additionally, an even distribution of
the fruits of their labor today may be the only credible promise for an
even distribution of those bene¬ts tomorrow. Wage moderation, in
other words, may presuppose wage solidarity and redistributive social
Welfare Production Regimes

Centralized and concertized bargaining of the form that emerged in
northern Europe in the decades following World War II addressed these
three problems simultaneously and became a central building block in a
unique continental European welfare production regime. The coordina-
tion of bargaining across sectors encouraged individual unions to exercise
wage restraint by convincing them that other unions would do likewise.
Distributive con¬‚icts could be solved through bargaining because the out-
comes of such bargaining would have an immediate effect on relative wages
and because future deviations from the compromise could be addressed
through future bargaining. Finally, centralized bargaining allowed govern-
ments to assume a supporting role by providing unemployment, health, and
retirement programs “ central institutions of the welfare state “ that reduced
workers™ uncertainty about their future welfare and, therefore, their temp-
tation to engage in short-termism and industrial con¬‚ict (Korpi and Shalev
1979; Lange and Garrett 1985).
In turn, for workers who were less exposed to labor-market uncertainty,
investing in their ¬rm™s success through wage restraint, acquisition of ¬rm-
and industry-speci¬c skills, and shop ¬‚oor cooperation became a more at-
tractive strategy than militancy. If a worker lost her job, she could depend
on a sizable income and a package of bene¬ts (such as healthcare) from the
state, and she would likely be hired back into a job with similar pay owing
to standardized wage rates (what has previously been referred to as wage
protection for the unemployed). At the macro-level, tax policies penalizing
dividends, and conspicuous consumption reassured workers that wage re-
straint would translate into higher investment (Stephens 1979; Katzenstein
1985; Eichengreen 1997).
On the employer side, ¬rms had to worry that the decision to invest
would encourage their workers to raise their wage demands in order to
appropriate the pro¬ts generated by that investment. But if wages were
determined in economy-wide rather than enterprise-level negotiations, an
individual ¬rm™s investment decision would no longer affect the wages it
had to pay. In these circumstances, centralized wage negotiations led to a
higher level of investment and, insofar as productivity was raised, to higher
wages in equilibrium (Hoel 1990).
Interlocking directorships and cohesive employers associations operat-
ing under close government oversight avoided the underprovision of tech-
nical training and R&D. Firms that would have otherwise been reluctant
to provide training to their workers for fear that they would be poached by
competitors were restrained by the threat of sanctions by both government
A Brief History of Modern Welfare Production Regimes

and industry associations. In addition, wage protection through the collec-
tive bargaining system, which ensures that similarly skilled workers are paid
approximately equal wages, and income protection through the state en-
abled employers to place some of the cost of training on employees, thereby
essentially creating an employee collateral in the company (in the case of
¬rm-speci¬c skills) and the industry (in the case of industry-speci¬c skills).
Institutionalizing union representation on corporate boards and in govern-
ment agencies in charge of economic, social, and educational policy made
it easier to monitor the parties™ compliance with the terms of these agree-
ments. It facilitated, to use a game-theoretic term, common knowledge
about the cooperative equilibrium.
It is dif¬cult to quantify the effects of cooperation on investment in
human and physical capital, but the evidence that the period from the late
1940s to the late 1960s was characterized by an unprecedented growth in
the capital stock and by a rapid expansion of both formal and vocational
training is unambiguous. The particular forms that these investments took
are discussed in Section 2.2.2.
Together, then, centralized bargaining, social protection, vocational
training systems, and collective representation of interests combined to
overcome the three key obstacles to growth. In turn, the capacity of coun-
tries to embrace this solution hinged on the presence of a set of historically
speci¬c structural conditions. First, cooperation was facilitated by the ex-
ceptional scope for rapid growth after the war. The European economy was
functioning below capacity. The in¬‚ux of labor from Eastern Europe and
internal migration from low-productivity agriculture to high-productivity
industry limited upward pressure on wages and supported the modern sec-
tor™s growth. Above all, a backlog of unexploited technologies was left over
from the years of war and depression, ready to be imported from the United
States. For all these reasons, the return on investment was high. Wage
restraint supporting that investment was generously rewarded, and insti-
tutional experimentation, even if it involved risks, carried high potential
Second, centralization was facilitated by a relatively homogeneous labor
force, which made it easier for workers to reach understandings about wage
relativities and for employers to live with wage compression. The European
adaptation of Fordist production methods, which relied on high-speed-
throughput technologies and incremental innovation, an extensive division
of labor, and semiskilled workers, was hindered very little by wage com-
pression that pushed up the cost of unskilled labor and depressed the wages
Welfare Production Regimes

of the most highly skilled, because European industry did not made heavy
use of workers in either tail of the feasible skill distribution. This skill struc-
ture was partly a result of a preindustrial craft tradition in many European
countries, partly a result of widespread training facilities created during the
war to boost the supply of skilled workers for the war economy, and partly
a result of investments in a quality system of primary education by govern-
ments eager to modernize, but constrained by limited resources for basic
education that would feed into the private training system. Almost every
worker, therefore, received a good primary school education, supplemented
by some period of vocational or on-the-job training. Additionally, a growing
number of workers received long vocational training through apprentice-
ships, vocational schools, or some combination of the two. Unskilled work-
ers gradually disappeared from the labor force or formed a small “¬‚exible”
segment that could more readily be dismissed during economic downturns.
Third, where unions were strong but politically uni¬ed and organized
along industry lines, compromise between different segments of workers
was easier to accomplish. Shifts in skill boundaries through reorganiz-
ation and upskilling prompted less opposition than was the case in a craft-
based union system like that in the United Kingdom, and this facilitated
compromises with management over investment in new technology and
training. It also appears to have been an important component to the con-
tinental European success story where unions and employers in the ex-
posed manufacturing sector were the largest, best organized, and politi-
cally most in¬‚uential actors. Because unions and employers in this sector
had a strong incentive to reach wage settlements that would maintain in-
ternational competitiveness, economy-wide wage moderation was easier to
accomplish (Swenson 1989, 1991a).
A similar argument applies to the organization of employers. In many
European countries, extensive cross-shareholding and strong representa-
tion on boards by large universal banks, interlocking directorships, and
well-organized employer associations already existed prior to the end of
the war and made it easier for employers to coordinate their behavior. This
“coordinating capacity” (Soskice 1990) facilitated the creation of central-
ized bargaining institutions by making it possible for many employers to
act jointly in the case of industrial con¬‚ict and by increasing the capacity
for the implementation of agreements with unions. Peter Swenson (1991a)
describes this capacity well in the case of Swedish employers using their
lock-out capacity during the 1930s to force the union confederation (LO)

A Brief History of Modern Welfare Production Regimes

to assume control over strike funds and cut off support to militant construc-
tion unions.
Finally, government policies supported cooperative bargaining by alle-
viating economic insecurity, addressing the distributive concerns of unions,
and penalizing noncooperative behavior. Tax policies rewarded investment
and punished consumption, as noted before. Subsidies and low-interest
loans were channeled to sectors where unions displayed wage restraint and
to ¬rms willing to support apprenticeship training and ¬nance R&D. Coun-
tercyclical monetary policies and ¬scal stabilizers limited uncertainty about
the future. Expansion of the welfare state encouraged workers to make
additional investments in risky skills and addressed the problem of distri-
butional con¬‚ict by supporting the maintenance of a “social wage” that
satis¬ed egalitarian norms.
But how could the state be relied upon to perform these functions? This
central question for political economy will be explored in detail in Chap-
ter 4; however, some preliminary observations from a historical perspective
can be made here. First, the mainstream parties that emerged from Europe™s
experience with left- and right-wing extremism before and during the war
were more inclined to emphasize economic growth and social insurance
than radical redistribution. The Cold War reinforced their pragmatism and
moderation, and, with the notable exceptions of Britain, the antipodes, the
United States, and (to a lesser extent) Japan, proportional-representation
electoral systems gave party elites a strong incentive to seek compromise
in order to form governing coalitions. Generally speaking, center parties
allied with moderate left parties to pursue policies of labor peace and so-
cial insurance ¬nanced by moderately progressive tax and transfer policies.
The center and left both bene¬ted from redistribution, but only to the point
where it did not seriously undermine investment incentives or jeopardize
middle-class interests.3
Equally important, the phase of mass mobilization had gradually given
way to a more or less “frozen” party system characterized by stable voting
blocks. Some societies were highly segmented or “pillarized” by religious
and other divisions, and no pillar could reasonably aspire to become hege-
monic in a PR electoral system. This reduced the incentives of parties to try
to buy off each others™ constituencies through populist tax cuts and de¬cit
spending.4 In turn, elites™ emphasis on growth, distributional justice, and

3 These conditions will be made much more precise in Chapter 4.
4 The notion of a frozen party system was put forth by Lipset and Rokkan (1967).

Welfare Production Regimes

consensus building encouraged voters to judge government performance
on precisely those dimensions. It also helped to build cooperative relations
with unions on which the government depended for labor peace and the
provision of other public goods such as constructive involvement in voca-
tional training systems.
From the perspective of unions, proportional representation and center-
left coalitions guaranteed that future governments would not move radi-
cally against their interests, thereby reducing future uncertainty. Insofar as
parties owed their electoral success as much to the efforts of highly central-
ized organizations of capital and labor (the latter in particular) as to their
own industriousness, governments also had an incentive to consult and in-
volve labor organizations in the preparation of new legislation and to seek
their consent in its implementation. This secured the consent of powerful
unions but also limited policy ¬‚exibility. In effect, the existence of these
disciplined mass organizations enabled the mainstream political parties to
credibly commit to the consensus policies of postwar Social Democracy.5
In addition to these domestic institutional conditions, the international
trade and monetary regime gave governments an important measure of ¬s-
cal and monetary policy autonomy by cushioning currencies against spec-
ulative attacks and by permitting governments to restrict and direct the
international ¬‚ow of capital. Likewise, the General Agreement on Trade
and Tariffs (GATT) only brought down trade barriers slowly and allowed
many exemptions to help European countries build their own industry
(Ruggie 1983). Nontariff barriers were particularly dense in services where
European governments argued that special considerations justi¬ed heavy
state regulation and exclusion of foreign competition.
Public utilities were widely considered natural monopolies that required
state ownership or tight regulatory control, and in areas such as telecommu-
nication and postal services there was a national security interest in keeping
foreign ¬rms at bay. Regulation or nationalization of banking and insur-
ance were considered necessary to protect markets against mass bankruptcy
and to allow governments to steer the national economy in the event of a
crisis. Protected service markets could also be used more directly as an
employment buffer against business cycle swings, stabilizing the economy
and facilitating the government™s commitment to full employment. Finally,
protection of services against competition was seen, rightly or wrongly,
as a means to ensure universalism in service provision and as inherently

5 Again, this insight will be developed further in Chapter 4.

A Brief History of Modern Welfare Production Regimes

inseparable from the goal of modernizing society by extending telephone,
postal, transportation, and other services to rural and less-developed re-
gions. Infrastructure, broadly speaking, was a precondition for industrial
development, and the future depended on the spread of industry. By com-
bining economic growth and full employment with some measure of social
justice, regulating and sheltering services from international competition
became part and parcel of the European growth strategy.
Table 2.1 summarizes the structural preconditions for the emergence of
institutions and policies that overcame the problems of growth described
previously. For the most part, the table refers to the situation that existed
immediately after World War II. Each column provides a standardized mea-
sure of the variables discussed in this section. The ¬rst is based on countries™
GDP per capita in 1950 and the ratio of this ¬gure to per capita income
in 1940. In both cases, lower numbers imply greater catch-up potential.
The index numbers have been inverted so that higher numbers indicate
greater catch-up potential. Unsurprisingly, the three big losers of the war “
Germany, Italy, and Japan “ score the highest on this measure.
Second, the strength and unity of the labor movement is a multiplica-
tive index of unionization rates and the organizational unity of the union
movement.6 Here all northern European countries score high with the
exception of Finland. The same pattern is evident for the capacity of busi-
ness coordination, which is based on a simple division of countries into
three groups according to the strength of business associations and cross-
shareholding. The next indicator, concentration of educational attainment,
is calculated from the distribution of the adult population across three cat-
egories of formal education in 1950. Not surprisingly, the United States
and Canada come out at the bottom, while education in most European
countries is concentrated in primary and secondary degrees. Finally, I have
used Lijphart™s measure of consensual democracy to capture the extent to
which the political system encourages political compromise. The compos-
ite index of structural preconditions in the last column is the mean of all
¬ve indicators.
In countries where the structural preconditions were propitious, cen-
tralization of wage bargaining generally progressed the furthest, and social
protection attained the highest levels. It is not easy to demonstrate this
conjecture with any degree of statistical precision because there is a lack of
systematic data for centralization and social protection that extends back to

6 Both measures were standardized to vary between 1 and 2 before being multiplied.

Welfare Production Regimes

Table 2.1. Structural Preconditions for Economic Policies and Institutions, ca. 1950

Strength Coordination Concentration Index of
Catch-Up and Unity Capacity of of Educational Consensual Structural
Potentiala of Labor b Businessc Distributiond Democracy e Preconditions
Canada 0.16 0.06 0.00 0.00 0.03 0.05
United 0.04 0.00 0.00 0.00 0.22 0.05
United 0.43 0.20 0.00 0.49 0.00 0.22
France 0.52 0.02 0.50 0.52 0.07 0.33
Italy 0.70 0.16 0.00 0.47 0.77 0.42
Switzerland 0.15 0.13 0.50 0.38 1.00 0.43
Japan 1.00 0.10 0.50 0.09 0.64 0.47
Belgium 0.54 0.18 0.50 0.37 0.77 0.47
Denmark 0.41 0.45 1.00 0.05 0.83 0.55
Netherlands 0.50 0.14 0.50 0.87 0.82 0.57
Germany, 0.70 0.24 1.00 0.33 0.63 0.58
Austria 0.69 0.66 1.00 0.20 0.52 0.61
Finland 0.59 0.08 0.50 1.00 0.92 0.62
Sweden 0.29 1.00 1.00 0.22 0.68 0.64
Norway 0.50 0.46 1.00 0.56 0.62 0.63
a Average, after standardization, of two indicators: (i) GDP per capita in 1950 and (ii) GDP per capita
in 1950 relative to GDP per capita in 1940. Source: Cusack (1991).
b Average, after 0“1 standardization, of three indicators: (i) unionization rates in 1950; (ii) political and
organizational unity of labor movement (0 = several competing and divided confederations, 1 = a
single dominant confederation; 0.5 = weak confederation or moderate division); and (iii) whether
unions are organized by industry (= 1), craft (= 0) or a mixture of principle (= 0.5). Sources: Visser
(1989) and Ferner and Hyman (1992).
c Standardized index with three values based on the extent of cross-shareholding and strength of employer
d Standardized index based on the standard deviation of the share of adults with a primary, secondary, and
postsecondary degree in 1950. Source: OECD™s Education Database as compiled by Robert J. Barro,
and Jong-Wha Lee (www2.cid.harvard.edu/ciddata/barrolee).
e Arend Lijphart™s index of consensus democracy, based on the effective number of parliamentary parties,
the percentage of minimal winning one-party governments, executive dominance, electoral propor-
tionality, and corporatism. The index covers the period 1945“96, but there is little change over this
period in any of the countries in the table. Source: Lijphart (1999, p. 312).

the war. In addition, as noted earlier, some of the institutions and policies
that emerged after the war were an extension of institutions and policies
that existed before the war. For these reasons, it is impossible to establish
an unambiguous causality between preconditions and subsequent institu-
tional developments. However, it is noteworthy that the index of structural
A Brief History of Modern Welfare Production Regimes

preconditions is highly correlated with a centralization of a wage-bargaining
index for the 1970s (r = .9),7 as well as with an index of social protection
that is explained later (r = .8). Since truly centralized bargaining systems “
distinguished from the compulsory wage-setting measures that were insti-
tuted in some countries during the war “ did not arise until the 1950s and
1960s, this is at least consistent with a story that emphasizes the importance
of the structural-institutional preconditions highlighted here. Likewise,
most of the cross-national differences in social protection that were evi-
dent in the 1970s and 1980s were the result of policies initiated in previous
decades (discussed later).
After institutions and policies were in place, they reinforced one an-
other through a set of interlocking complementarities (Hall and Soskice
2001). Governments supported centralized bargaining because it facili-
tated rapid growth that aided the electoral fortunes of the governing parties.
Strong unions and employers™ associations also helped maintain the political
monopoly of mainstream parties by providing ¬nancial and organizational
support. Politicians, in turn, nurtured the institutions of centralized bar-
gaining by granting representational monopolies to the peak associations
of capital and labor, rewarding unions for their restraint and attending to
their distributional interests.
By and large, this ideal type description is best approximated in the
Nordic countries, where centralized industrial relations and secure social
democratic governments committed to full employment produced both
wage compression and a rapid expansion of pensions and other social rights.
Union cooperation was facilitated by drawing representatives of the main
labor market organizations into the preparation and implementation of
literally every new piece of social or economic legislation. As we move south
and west from Scandinavia, one or more elements of the postwar model “
centralized bargaining, government commitment to full employment, and
egalitarian wage and social policies “ are weakened, but in no case are all
three completely missing.
Germany, Austria, Belgium, the Netherlands, and Switzerland all de-
veloped highly coordinated and ordered systems of industrial relations in
the 1950s, although in none of these cases was bargaining centralized at the
national level. While these countries differed from their Scandinavian coun-
terparts in their level of public service provision and employment, they too
passed generous entitlement legislation such as Adenauer™s 1957 pension

7 The centralization of wage bargaining index is from Iversen (1999, Chapter 3).

Welfare Production Regimes

reform in Germany, which greatly increased social spending. The gen-
erosity of the Dutch social security system was second to none in Europe.
Likewise, governments instituted extensive legal and regulatory protection
for employment, and social insurance schemes introduced in the 1930s
were bolstered by legislation and the creation of employee workplace
In the European context, France, Italy, and the United Kingdom deviate
more sharply from the ideal type. All suffered coordination problems be-
cause of fragmented labor movements and weak employers™ organizations.
In France and Italy, unions were divided by ideological and confessional
cleavages, and many had allied with politically marginalized communist
parties. The latter saw the unions not just as vehicles for workers™ material
progress but also as an important organizational resource for the effort to
mobilize and expand a mass base. Partly for this reason, governments in
both countries were reluctant to pursue policies that might strengthen the
communist unions. The French state, however, could rely to some extent
on a strong independent bureaucracy to overcome coordination problems
and pursue aggressive investment and industrial restructuring policies, and
although Italy was one of the worst performers in terms of in¬‚ation and
unemployment, cooperative solutions did emerge at the local and regional
levels in areas where unions and the left were strong and entrenched.
By contrast, Britain never devised effective institutional solutions to any
of the three obstacles to growth. After the war, a reform-minded Labour
government nationalized several industries, but otherwise industrial poli-
cies were kept at arm™s length, in part as a result of a market-based ¬nancial
system that did not lend itself to French-style dirigisme (Zysman 1983).
The ¬nancial system was also an important impediment to the pursuit of
full employment. Because British banks were heavily oriented toward inter-
national banking, they opposed devaluation. Consequently, when the gov-
ernment tried to address internal imbalances through demand stimulation,
it would often ¬nd itself reversing policies so as to not cause a politically
unacceptable depreciation of the pound (Hall 1986, Chapter 4). The result-
ing “stop-go” pattern was clearly not conducive to farsighted investment
and wage strategies. Finally, even though representatives of employers and
unions were consulted on economic policy matters, legislation designed to
difuse distributive con¬‚ict did little to induce wage restraint for the simple

8 The laggard here is Denmark, partly because its industrial structure is so dominated by small

A Brief History of Modern Welfare Production Regimes

reason that no individual union had much incentive to cooperate with the
government™s incomes policy given that membership was divided among a
large number of mainly craft-based unions.
In several respects, the postwar British political economy can be seen
as a problematic hybrid of northern European and U.S. capitalism. As in
northern Europe, unionization rates were much higher in Britain than in
the United States, but like the labor movement and business community
in the United States, that in Britain was far more fragmented than that in
northern Europe. Indeed, it can be argued that because of the concentration
of American unions in the automobile industry and because unions in that
industry were fairly well organized and engaging in pattern bargaining,
U.S. wage setting was more coordinated than in Britain. Outside the much
smaller unionized sector in the United States, wage pressures were kept in
check by highly competitive labor markets.
The U.S. political economy is notable for the absence of any govern-
ment employment guarantee or commitment to egalitarian wage and social
policies. Although some aspects of the New Deal survived under Truman
and Johnson, notably social security, there was neither the economic ur-
gency nor the political support for interventionist policies, extensive social
protection, or comprehensive incomes policies of the kind that emerged
in Europe. American production capacity and infrastructure had not been
destroyed in the war, and American ¬rms enjoyed a signi¬cant productiv-
ity advantage vis-` -vis their European and Asian competitors after the war.
Against this backdrop, the main issue was not how to accelerate domes-
tic output but how to cultivate global demand for U.S. products. Aid to
help rebuild Europe and stop any westward spread of communism, and a
stable trading regime that kept foreign markets open to U.S. exports and
investments were key ingredients in the U.S. global strategy. The Marshall
Plan, Bretton Woods, and GATT became the main institutional vehicles
for forwarding this strategy.9
Many of the key postwar U.S. welfare programs, such as food stamps,
Medicaid, and public housing, were the result of mobilization by poor black
constituencies and civil unrest, notably the Inner-city riots of the 1960s,
and they had a distinctly redistributive impetus and effect. In a similar vein,

9 This is not to say that postwar international institutions were unmediated expressions of U.S.
hegemony. Rather, as argued by Ruggie, they were a compromise between the American
desire for a liberal international economy and the social and political reality of Western
European countries, which required some capacity of governments to shape policies to the
needs of domestic social stability (Ruggie 1983).

Welfare Production Regimes

af¬rmative action sprang from a racially based mobilization of the lower
classes. With the major exception of Johnson™s introduction of Medicare
(and before that Social Security), most welfare programs were of little di-
rect bene¬t to the white middle class; therefore, they were vulnerable to
rollback when violence and protest subsided. Majoritarian electoral institu-
tions eliminated the need for a compromise with the poor, and both major
parties have always been to the right by European standards. The impor-
tance of electoral institutions for partisanship and redistribution is discussed
further in Chapter 4.
In terms of protection against risks, the American middle class turned
to education. Catering to centrist voters, the federal government and the
states poured huge sums of money into the public university system, and
individuals spent an ever-increasing share of their income to send their chil-
dren to college. There was never an effective vocational alternative “ partly
because there was no deeply rooted craft tradition, and partly because busi-
ness was inadequately organized to provide the necessary funding, training,
and monitoring of such an alternative “ consequently, a college degree was
the only ticket to the American dream that could be bought with money
and effort rather than luck.10

2.2. Comparison of Welfare Production Regimes11

2.2.1. Social Protection
The national differences in the social protection systems as they evolved
during the ¬rst three decades after the war can be fairly well summarized by
a comparison of employment and unemployment protection “ two dimen-
sions emphasized in the theoretical discussion in Chapter 1. Data for the ¬rst
dimension are provided in Table 2.2. The ¬rst column is OECD™s measure
of legal and quasi-legal employment protection legislation (EPL) govern-
ing individual hiring and ¬ring. The composite EPL index is based on
provisions in the legal code as well as on collective bargaining agreements,
hereunder what constitutes just cause for dismissals, required length of ad-
vance notice, mandated severance pay, compensation for unfair dismissals,

10 A few large ¬rms such as Boeing and IBM could promise life-time employment to workers
who would go through many years of company training because of the position of these
¬rms as technological and market leaders. Of course, as this position of leadership was
eroded, so was the life-time employment system.
11 This section builds on Estevez-Abe et al. (2001).

A Brief History of Modern Welfare Production Regimes

Table 2.2. Employment Protection in Eighteen OECD Countries

(1) (2) (3) (4)
Employment Collective Index of
Protection Dismissals Company-Based Employment
Legislation (EPL)a Protectionb Protectionc Protectiond
Sweden 2.8 4.5 3 0.94
Germany 2.8 3.1 3 0.86
Austria 2.6 3.3 3 0.84
Italy 2.8 4.1 2 0.81
Netherlands 3.1 2.8 2 0.80
Japan 2.7 1.5 3 0.76
Norway 2.4 2.8 2 0.66
Finland 2.4 2.4 2 0.64
France 2.3 2.1 2 0.61
Belgium 1.5 4.1 2 0.56
Denmark 1.6 3.1 2 0.53
Switzerland 1.2 3.9 2 0.49
Ireland 1.6 2.1 1 0.36
Canada 0.9 3.4 1 0.30
New Zealand 1.7 0.4 1 0.29
Australia 1.0 2.6 1 0.27
United Kingdom 0.8 2.9 1 0.25
United States 0.2 2.9 1 0.14
a Index of the “restrictiveness” of individual hiring and ¬ring rules contained in legisla-
tion and collective agreements (the high numbers indicate the more restrictive regimes).
Source: OECD (1999b). (Weight: 5/9).
b Index of the “restrictiveness” of collective dismissal rules contained in legislation and col-
lective agreements (the high numbers indicate the more restrictive regimes). Source: OECD
(1999b). (Weight: 2/9).
c Measure of company-level employment protection as explained in text. The French case
has been assigned a score of 2 even though company-level protection is weak. The reason
is that the Inspectorat du Travail can and does intervene to prevent redundancies, and this
is not captured by OECD™s legal measure of employment protection. See Berton, Podevin,
and Verdier (1991) for a description of the French system. Sources: Income Data Services
(1996); OECD (1998), pp. 142“52; David Soskice (1999). (Weight: 2/9).
d Weighted average of columns (1) and (2) after each indicator has been standardized to vary
between 0 and 1.

Welfare Production Regimes

the rights of employee representatives to be informed about dismissals and
other employment matters, and the rights of workers to challenge dismissals
in the courts.
The composite EPL index is constructed for both regular and tempo-
rary employment, but the skill argument is only relevant for the former
because neither employers nor employees have much of an incentive to
invest in ¬rm-speci¬c skills when employment is time-limited. The regular
employment EPL index is calculated for two periods, the late 1980s and
the late 1990s, but because it is nearly perfectly correlated between the two
periods (r = .99), the ¬gures are simple averages (shown ¬rst column of
Table 2.2). The index is based on the regulation of individual contracts and
does not incorporate measures for protection against collective dismissals.
In OECD™s latest update of the index (OECD 1999b), a separate index was
created to re¬‚ect the regulation of collective dismissals, which is shown in
the second column of Table 2.2.
Neither of the OECD measures fully take into account the employment
protection that is built into the ¬rm governance structure or into the work-
ings of the industrial relations system. As the OECD acknowledges, “non-
legislated employment protection tends to be more dif¬cult to measure and
may therefore be under-weighted” (OECD 1999b, p. 51). Japan illustrates
the problem because companies in Japan offer greater protection against
dismissals for their skilled workers than the EPL index would suggest (See
OECD 1994a, pp. 79“80). Indeed, dismissals and layoffs are extremely rare
in Japan compared to other countries (OECD 1997, Table 5.12).12 Instead,
large Japanese ¬rms engage in special work force loan practices with their
suppliers, called “Shukko,” which enable them to retain workers during
recessions. In other countries, and to some extent also in Japan, ¬rms must
consult with works councils or other employee representative bodies before
making decisions about layoffs, and industry unions are often in a strong
position to oppose collective layoffs. This relationship is only partly re-
¬‚ected in the EPL index because it considers only the need for ¬rms to
notify works councils or unions about impending dismissals, not the power
of unions or works councils to prevent or modify the implementation of
decisions to dismiss.
These “private” employment protection arrangements are captured in
column 3 of Table 2.2 by a simple index that measures the strength of

12 These data are not fully comparable across countries, but the very low ¬gures for Japan
leaves little doubt that the numbers are much smaller in Japan than elsewhere.

A Brief History of Modern Welfare Production Regimes

institutions and practices at the ¬rm level that increase the job security of
especially skilled workers in a company. The measure is based on three
criteria: (1) the presence of employee-elected bodies with a signi¬cant
role in company manpower decisions; (2) the existence of strong external
unions with some monitoring and sanctioning capacity (especially through
arbitration); and (3) the systematic use of employee sharing practices be-
tween parent companies and subsidiaries or across companies. Where at
least two of these conditions are met to a considerable degree, a score of 3
was assigned; where all three are largely absent, a score of 1 was assigned.
Intermediary cases were assigned a score of 2. With the exception of Japan,
the index of company-based protection is consistent with the rank-ordering
implied by the composite index.
The ¬nal column combines the OECD and company-based measures in
a composite index that captures both the legal and more informal aspects of
employment protection. The index is a weighted average with the following
weights: 5/9, 2/9, and 2/9. The ¬rst two weights are adopted unchanged
from OECD™s own weighing scheme (OECD 1999c, p. 118), and re¬‚ect
the fact that collective dismissal rules tend to build on individual dismissal
rules, which are already part of the EPL index. Because the in¬‚uence of
employee representative bodies over ¬rm-level manpower decisions is also
partly captured by the EPL index, I assigned the same (low) weight to the
company protection indicator. Some would quibble with the assignment of
these weights, but the relative numbers are not very sensitive to changes in
these weights, and I think most agree that the resulting index of employment
protection provides a reasonably good summary of the differences across
Looking at the ranking of countries, it is not surprising to ¬nd the Anglo-
Saxon countries at the low end and Japan and many of the continental
European countries at the high end of protection. Belgium, Denmark, and
Switzerland are in the lower half of the table, most likely because these
countries have relatively large small-¬rm sectors, but in terms of actual
numbers, the break in the employment protection index is between this
group of countries and the Anglo-Saxon countries.
The measurement of unemployment protection is more straightforward,
although there are some nontrivial issues concerning the administration of
unemployment bene¬t systems. The most obvious and commonly used in-
dicator is the unemployment replacement rates, the portion of a worker™s
previous wage that is replaced by unemployment bene¬ts (see column 1 of
Table 2.3). I here consider a “typical” worker, de¬ned as a 40-year-old
Welfare Production Regimes

Table 2.3. Unemployment Protection in Eighteen OECD Countries

(1) (2) (3) (4)
Net De¬nition Index of
Unemployment Generosity of Suitable Unemployment
Replacement Ratesa of Bene¬tsb Jobc Protectiond
Denmark 60 76 3 0.91
Netherlands 58 74 3 0.89
Switzerland (40) 94 2 0.86
Belgium 57 99 2 0.82
Austria 43 78 3 0.81
Germany 43 66 3 0.77
Norway 40 40 3 0.64
Sweden 30 52 3 0.63
France 48 44 2 0.54
Finland 45 20 2 0.43
Ireland (38) 59 1 0.37
Japan 10 48 2 0.33
Canada 32 49 2 0.30
New Zealand 31 44 1 0.27
Australia 32 30 1 0.22
Italy 5 18 2 0.18
United 23 15 1 0.11
United States 14 26 1 0.10
a Net unemployment replacement rates for a 40-year-old representative worker. Source: Re-
stricted OECD data reported in Esping-Andersen (1999a, Table 2.2, p. 22). Net ¬gures
for Ireland and Switzerland are missing and have instead been estimated by taking gross
replacement rates for these countries as proportions of average gross replacement rates
and then multiplying these proportions by average net replacement rates. Source: OECD,
Database on Unemployment Bene¬t Entitlements and Replacement Rates (undated).
b The share of GDP paid in unemployment bene¬ts as a percent of the share of unemployed in
the total population. Average for the period 1973“89. Sources: Huber, Ragin, and Stephens
(1997); OECD, Economic Outlook (various years), OECD, Labour Force Statistics (various
c Index that measures the restrictiveness of the de¬nition of a suitable job in the administration
of bene¬ts to unemployed. 1, Any job quali¬es as a suitable job; 2, skilled unemployed are
given some discretion in rejecting jobs they deem unsuitable to their skills, but choice is
restricted in time and/or to certain job categories, 3, skilled unemployed exercise wide
discretion in accepting or rejecting jobs on the grounds of the suitability of the job to their
skills. Sources: OECD (1991, pp. 199“231); European Commission, Employment in Europe
(various years); and national sources.
d Average of columns (1)“(3) after each indicator has been standardized to vary be-
tween 0 and 1.

A Brief History of Modern Welfare Production Regimes

industrial production worker, “averaged” across several different family
types (single, married to working spouse, and married to nonworking
spouse). Additionally, I focus on net replacement rates after adjusting for
cross-national differences in tax systems and non-income subsidies for the
unemployed (such as rent support). Given that taxation of unemployment
bene¬ts varies considerably across countries, gross replacement rates (for
which much more detailed data exist) can be misleading.
As in the case of employment protection, the Anglo-Saxon countries
again score at the bottom. But note that the three continental European
countries falling in the lower half of the employment protection index “
Belgium, Denmark, and Switzerland “ now ¬gure at or near the top of the
table. On the other hand, two countries “ Italy and Japan “ have very low
replacement rates compared to their position on the employment protec-
tion indicators. The pattern is broadly similar, though not identical, when
looking instead at the actual amount of money the government spends on
unemployment bene¬ts (as a share of GDP), compared to the number of
unemployed people (as a share of the population). As before, the three
countries in northern Europe with relatively low employment protection
are among the ¬ve countries with the most generous unemployment bene¬t
Table 2.3 also includes a more qualitative measure of the administration
of unemployment bene¬ts: the restrictiveness of the de¬nition of a suitable
job. All national unemployment systems stipulate that in order to receive
bene¬ts a person cannot refuse a suitable job, but what constitutes a suit-
able job varies signi¬cantly from one system to another. In principle, such
variation is important for our purposes. For example, if a skilled worker is
required to take any available job, regardless of whether it is commensurable
with the worker™s skills, high unemployment bene¬ts are of limited value
from the perspective of reducing the riskiness of speci¬c skills investments.
In practice, it is dif¬cult to get any precise comparable ¬gures for this vari-
able. Consequently, I am using only a very simple three-tiered classi¬cation
based on a variety of national and international sources. Though basically
reinforcing the pattern revealed by the other two indicators, it does affect
the rank-order position of some countries slightly.
As in the case of employment protection, I combined the various indi-
cators into an index of unemployment protection (see column 4). With the
possible exception of Italy, this index gives a good sense of cross-national
differences in the extent of unemployment protection. The number for Italy
probably underestimates the extent of protection because of quasi-public
Welfare Production Regimes

insurance schemes that do not show up in the of¬cial statistics. Thus, about
a third of (mainly large companies) covered by Casa Integrazione have
replacement rates between 70 and 80 percent, and there are normally good
unemployment bene¬t schemes for artisans (i.e., craftsmen) administered
at the regional level by associations representing small ¬rms, in cooperation
with regional governments.13
In addition to employment and unemployment protection, Chapter 1
distinguished a form of insurance called income protection. Income protec-
tion is secured both through the collective wage bargaining system in the
form of negotiated standard wage rates and through the public tax and
transfer system.14 Income protection helps reduce the variability of after-
tax and transfer income and, hence, helps to manage the risks associated
with investing in skills that cannot easily be transferred from one job or
occupation to another. Table 2.4 seeks to capture income protection with
three different measures. The ¬rst is an index of total taxes and transfers
(both weighed equally) as a proxy for publicly mediated income protection.
The second is d9/d1 earnings ratios used as a proxy for (the inverse of )
protection of wages through the private wage-setting system. The third is
after-tax and transfer Gini coef¬cients, which are a function of both wage
and public income protection. In principle, Gini coef¬cients are the key
data for our purposes, but because they are only available at the level of
households, they do not directly measure individual income protection.
Even if individual income is highly protected, after-tax and transfer house-
hold income can be quite dispersed owing to differences in the labor force
participation rates of household members. As in Tables 2.2 and 2.3, the
¬nal column is an index of protection based on all three indicators (after
As can be seen from the correlation matrix at the bottom of Table 2.4,
the different indicators are all fairly highly correlated with one another, and
the income protection index produces an ordering of countries that is quite
similar to that for the unemployment protection index. Japan, Ireland, and
the Anglo-Saxon countries are at the bottom of the scale, while the Low
Countries and Scandinavia are at the top. Italy has moved up compared
to the unemployment index and is probably more accurately located, but
Switzerland is now an outlier. Without Switzerland, the correlation be-
tween the income and unemployment indexes rises to .75. In Chapter 4,

13 Michele Salvati provided this information.
14 More precisely, this was called income protection for the unemployed in Chapter 1.

A Brief History of Modern Welfare Production Regimes

Table 2.4. Income Protection in Eighteen OECD Countries, 1973“1995

(1) (2) (3) (4)
d9/d1 After-Tax and
Taxes and Earnings Transfer Gini Index of Income
Transfersa Ratiosb Coef¬cientsc Protectiond
Sweden 0.77 2.07 19.40 0.94
Belgium 0.78 2.34 21.67 0.85
Denmark 0.61 2.18 21.50 0.81
Norway 0.50 2.00 21.50 0.80
Finland 0.41 2.08 20.00 0.79
Netherlands 0.89 2.56 26.00 0.74
Germany 0.42 2.80 26.16 0.52
Austria 0.57 3.54 “ 0.46
France 0.67 3.25 29.40 0.46
Italy 0.37 2.33 31.33 0.44
Switzerland 0.16 2.72 30.50 0.32
Australia 0.03 2.81 28.50 0.31
Japan 0.01 3.07 “ 0.26
United States 0.10 3.21 30.00 0.24
Ireland 0.33 4.06 “ 0.21
Canada 0.20 4.19 27.80 0.18
United Kingdom 0.22 4.06 32.00 0.10
Correlation matrix:
(1) Transfers as 1
percentage of GDP
(2) d9/d1 earnings 1
(3) After-tax/transfer .68 1
(4) Income protection .79 .85 .85
’.60 ’.53
Unemployment .68 .65
a Taxes and transfers is the mean of total taxes as a percentage of GDP and total transfers as
a percentage of GDP, after both measures have been standardized. The transfer data are
from the OECD, National Accounts, Part II: Detailed Tables (various years), and the tax data
are from the OECD (2002).
b d1/d9 earnings ratios are the gross earnings (including all employer contributions for pen-
sions, Social Security, etc.) of a worker at the top decile of the earnings distribution relative
to the worker at the bottom decile (OECD, Electronic Data Base on Wage Dispersion, undated);
c Ginis are calculated on the basis of LIS data on posttax and transfer income for households
with working-age adults.
d Average of columns (1)“(3) after each indicator has been standardized to vary between 0
and 1.

Welfare Production Regimes

Table 2.5. Percentage of Population over 25 with a Postsecondary Education

1950 1960 1970 1980 1990 2000
North America 13.6 18.3 20.9 33.6 44.0 51.6
United States 13.6 16.5 21.3 29.8 45.2 50.1
Canada n.a. 20.0 20.4 37.4 42.7 53.0
Europea 1.9 4.5 6.3 10.0 13.9 19.6
France 2.1 3.0 8.5 11.4 18.4
Germany n.a. 1.8 3.1 6.9 10.4 17.5
Italy 1.5 2.1 2.6 4.1 9.0 14.7
Sweden n.a. 7.5 8.3 15.4 18.3 23.1
Difference 11.7 13.7 14.5 23.6 30.0 32.0
a Average for the following thirteen countries: Austria, Belgium, Denmark, Finland, France,
Germany, Ireland, Italy, Netherlands, Norway, Sweden, Switzerland, United Kingdom;
b 1955.

Source: OECD™s Education Database as compiled by Robert J. Barro and Jong-Wha Lee

it is argued that Switzerland is an outlier when it comes to redistributive
policies because the Swiss government has a collective executive that gives
right-wing parties veto power over redistributive policies. As noted in Chap-
ter 1, the pattern in other PR systems has been coalition governments that
exclude the right.

2.2.2. Skills and Training Systems
The differences between the American and European educational systems
were clearly visible already in the immediate postwar period. Thus, in 1950,
14 percent of the American population over 25 years of age had a postsec-
ondary degree compared to less than 3 percent in most European countries.
By 1970, the ¬gure had risen to more than 21 percent, compared to 6 per-
cent in Europe (see Table 2.5). The pattern is very similar for Canada,
and the gap between North America and Europe has been growing over
time (contrary to a simple catch-up hypothesis). In Europe, the univer-
sity was the providence of the privileged few during the 1950s and 1960s,
and whereas most postsecondary education in the United States leads to
general degrees, in Europe more of this education is taking place through
professional schools using more targeted occupational curricula.
The ¬gures for upper-secondary education also hide more subtle dif-
ferences in the content of education. In the Anglo-Saxon countries, uni-
versity education tends to be very general, and even engineering and
A Brief History of Modern Welfare Production Regimes

Table 2.6. Skill Systems in Eighteen OECD Countries

(1) (2) (3)
Vocational Median Length of Vocational Training
Training Sharea Tenureb Systemc
Austria 22 6.9 Dual apprenticeship
Germany 34 10.7 Dual apprenticeship
Sweden 36 7.8 Vocational colleges
Norway 37 (6.5) Vocational colleges
Belgium 53 8.4 Mixed
Japan 16 8.3 Company-based
Finland 32 7.8 Vocational colleges
Italy 35 8.9 Company-based
France 28 7.7 Mixed
Ireland 6 5.3 Weak
Netherlands 43 5.5 Mixed
Switzerland 23 6.0 Dual apprenticeship
Denmark 31 4.4 Mixed
Canada 5 5.9 Weak
Australia 9 3.4 Weak
New Zealand 7 n.a. Weak
United Kingdom 11 5.0 Weak apprenticeship
United States 3 4.2 Weak
a The share of an age cohort in either secondary or postsecondary (ISCED5) vocational
training. Source: UNESCO (1999).
b The median length of enterprise tenure in years, 1995 (Norwegian ¬gure refers to 1991).
Sources: OECD Employment Outlook, 1997b, Table 5.5. For Norway: OECD (1993, Table 4.1).
c The character of the vocational training system according to whether most of the training
occurs at the company level (as in Japan), through a dual apprenticeship system (as in
Germany), through vocational colleges (as in Sweden), or through some mixture of the latter
two (as in the Netherlands). Where vocational training is weak, I have not distinguished
between the type of system. Sources: Streeck (1992); Finegold and Soskice (1988); Soskice
(1999); Crouch, Finegold, and Sako (1999).

business schools provide very broad training that is not linked to par-
ticular industries or trades. By contrast, in Japan and most continental
European countries, many university degrees are more specialized and there
tends to be close linkages from engineering and trade schools to private
But the part of the educational system that most clearly distinguish con-
tinental Europe (and Japan) from the Anglo-Saxon countries (and Ireland)
is vocational training (see Table 2.6). The share of an age cohort that goes
through a vocational training in the latter (column 1) varies between 3 and
Welfare Production Regimes

11 percent (counting short-term postsecondary degrees), and there is little
involvement of companies in the training system. In the former countries,
the percentage of an age cohort going through vocational training is gen-
erally between a quarter and one half of an age cohort. At 16 percent,
the ¬gure for Japan is somewhere in the middle, but much training in this
country goes on in large ¬nancially secure companies and is not recorded in
the data.
As noted in Chapter 1, vocational training is closely related to how well
adults perform on Statistics Canada and OECD™s international literacy tests.
In particular, those who leave the formal educational system early are much
more likely to do well on these tests in countries with extensive vocational
training programs than in countries emphasizing formal education. This
fact would appear to be a result of the additional training afforded by the
presence of good vocational training opportunities and the incentives to
work hard in school that such opportunities create for young people at the
lower end of the academic ability distribution.
The main difference among the countries with strong vocational training
systems is in the emphasis on company as opposed to industry-level training.
Whereas in Japan, and to a lesser extent in France and Italy, the emphasis
is on company training, the remaining countries have some combination
of on-the-job training and school-based training, with heavy involvement
of employer organizations and unions. Formally, the systems can be di-
vided into the apprenticeship systems of the German type, the vocational
school systems of the Swedish type, or mixtures between the two, but they
all combine theoretical, industry-speci¬c, and direct workplace training
(column 2). The relative importance of the three is dif¬cult to gauge, but
Belgium, The Netherlands and the Scandinavian countries tend to place
more emphasis on school-based training (i.e., provision of non-¬rm-speci¬c
skills) than do Austria or Germany.
The difference in skill systems can also be gauged by median enterprise
tenure rates “ the median number of years workers have been with their
current employer (based on national labor force surveys). These numbers
contain relevant information about the ¬rm-speci¬city of skills because
¬rms and individuals investing heavily in such skills become increasingly
dependent upon one another for their future welfare. The greater the in-
vestment, the higher the opportunity costs of severing the relationship and
the lower the incentive for either party to do so. Indeed, short tenure rates
may be not only an indicator of the absence of ¬rm-speci¬c skills but also a
positive measure of presence of general skills. The reason is that general skills


. 2
( 11)