WELFARE STATES WITHOUT WORK: THE IMPASSE OF LABOR SHEDDING AND FAMILIALISM IN CONTINENTAL EUROPEAN SOCIAL POLICY

Gsta Esping-Andersen

Estudio/Working Paper 1995/ September 1995

Gsta Esping-Andersen is Professor in the Department of Political Science, Universit di Trento, Italy. He was Visiting Professor (Spring Semester 1995) at the Center for Advanced Study in the Social Sciences of the Juan March Institute in Madrid. He is also a member of the Scientific Council of the Center.

1.

Introduction

Imperial Germanys social insurance reforms became the model for policy emulation in most European countries. As with Bismarck, social policy was motivated by concerns far removed from any egalitarianism. Chiefly, it was the creation of conservative elites who abhorred laissez-faire almost as much as the socialist menace, who sought to rescue the old order with moral discipline, social pacification, and nation building (Rimlinger 1971; Flora and Alber 1981). Bismarcks aim was not a welfare state but a welfare monarchy. In comparison with the universalistic egalitarian ideals that later came to define the British Beveridge plan or Scandinavian social democracys _Peoples Home_ ideal, the early architects of social policy on the Continent were authoritarian, etatist, and corporativistic. Much of this conservative legacy has been carried over in contemporary welfare state institutions. In most other countries the modern welfare state was built by social democrats or left-liberals; welfare state consolidation in postwar Europe was dominated by rightist or center-right coalitions -- Christian Democracy in particular. Even where christian democratic parties were marginal, such as in France or Spain, Catholic social doctrines still exert a visible influence on social policy. The partisan roots of the European welfare states are, for two important reasons, worth stressing. For one, the christian democratic _subsidiarity principle_ has institutionalized familialism in the sense of supporting the male breadwinnerfemale caring model with transfers. In particular the Southern European are characteristic for their very generous cash transfers and their almost non-existent provision of social services.
1

The conservative-catholic imprint affects also the politics of crisis and

- Error!Marcador no definido. -

retrenchment today. The kind of ideologically fuelled partisan battles that are fought in the Anglo-saxon nations, and even in Scandinavia, are conspicuously absent. The neo-liberal right is, on the Continent, a truly marginal player. These are probably the most consensual of all modern welfare states and this, as we shall see at the end of the paper, helps explain why the welfare state edifice remains so immune to change, notwithstanding the urgency of major and quite radical reforms. The implications of a social insurance dominated approach are, firstly, that social entitlements derive principally from employment rather than citizenship (as in the Nordic model), or from proven need (as in the more targeted systems in North America and the Antipodes). It is assumed that family dependents rely on the entitlements of the (usually male) breadwinner. A second implication is that social protection tends to be differentiated by occupational classes; that benefits will mirror accustomed status and earnings rather than redistributive ambitions. In brief, the realm of solidarity and risk-sharing is narrower and more particularistic. The degree of status differentiation varies between nations and programs. In Germany, unemployment insurance is unified, pensions are divided into broad occupational classes (workers, salaried, miners, and civil servants), and health care is a labyrinth of 1200 separate regional, occupational, or company based funds. In contrast, the French and Italian pension systems combine large occupationally mixed plans with a myriad of status-particularistic schemes. Italy's health care system is unified, while the French (and Belgian) is organized in terms of broad occupational funds. Civil servants tend to be especially privileged in the Continental European welfare states. As we shall see, the long-term financial solidity of social insurance is inversely related to the degree of status fragmentation. The influence of Catholic social teachings remains a fetter on state provision of services, particularly those related to social care and family social reproduction. As a result, the Continental European model is, in sharp contrast to the Nordic

- Error!Marcador no definido. -

countries, service-lean yet very _transfer-heavy_.

The consequences for employment, as we shall see, are substantial. On one side, the lack of social services contradicts women's growing desire for employment and helps account for Europe's overall employment stagnation. As women nonetheless increasingly desire to work, and since high wage costs make private care alternatives unaffordable, the system imposes a severe trade-off between female careers and fertility. It is in this context that we must understand why Italy and Spain, today, exhibit world-record low fertility rates. In broad terms, the accent on social insurance remained a source of strength and adaptability during the postwar decades' welfare state maturation. In comparison to the British or Nordic flat-rate schemes, social insurance was more easily upgraded in terms of benefit adequacy with rising incomes and an increasingly affluent middle class society. The substantially heavier financial burden which full income maintenance imposed was, until the 1970s, easily absorbed because of sustained growth with low inflation and unemployment (Gordon 1988). The perceived adequacy of social insurance across most social strata implies also very little demand for private sector coverage, especially in pensions. And since contributory social insurance instills a sense of individually earned contractual rights, these social security systems have enjoyed unusually broad public legitimacy. Despite mounting fiscal burdens, and the inescapable complaints about bureaucracy and the quality of services, the kind of anti-welfare state revolts that erupted in both Scandinavia, Britain, and the United States are
Thus, the 1990 ratio of social services (excluding health care) to social transfer payments is .16 in Germany, .12 in France, and .06 in Italy, whereas in Denmark it is .33 and in Sweden .29. (Note that the U.S. ratio is also a low .07, here mainly due to the encouragement of market alternatives). (Source: OECD, National Accounts. Detailed Tables. 1991). Put differently, a third of the Scandinavian welfare states is dedicated to servicing family needs; in the Continental European nations, typically only one-tenth. Non-health social services include day care and youth services, care of the aged and disabled, home help services, and the like, but also employment-related services such as rehabilitation schemes and employment exchanges.
2

- Error!Marcador no definido. -

conspicuously absent, be it in Germany, France or Italy. The primary focus of this paper is on the problems and tensions that these particular welfare state models face as a consequence of sweeping social and economic change in the past decades. Rather than boost public employment, as in Scandinavia, their main policy response to _de-industrialization_ and unemployment has been to induce labor force exit. This approach may have had favorable consequences in terms of restoring competitiveness in traditional industries, such as autos. The longer term effects, however, are quite problematic. It has resulted in very unfavorable population dependency ratios due to swelling numbers of retirees coupled with a stagnant and even shrinking workforce. The upshot is prohibitively heavy fixed labor costs which, in turn, discourage employment growth or, alternatively, spur the growth of informal sector jobs or self-employment. In brief, these systems find themselves locked into a self-reinforcing negative spiral, and are today particularly ill-suited to address pressures for greater labor market flexibility and women's demand for economic independence. In brief, the Continental Western European welfare states are coming into conflict with the emerging needs of a postindustrial economy. Before analyzing these problems in greater depth, I shall provide a brief sketch of the principal characteristics of the postwar European welfare state models. In the third, and final, section I shall examine possible future scenarios on the basis of identifiable current trends.

2.

The postwar model of social security.

The postwar welfare state consolidation on the Continent meant essentially an elaboration and extension of existing social insurance in both income maintenance (accident, sickness, disability and old age pensions, unemployment)

- Error!Marcador no definido. -

and health care. In some cases, like Germany and Italy, postwar reformers entertained ideas of a Beveridge-type universalistic system. However, the vested interests in maintaining the proven social insurance model were too strong. Pressures for more universalism translated instead into a patchwork policy of extending insurance to previously non-covered populations (such as the selfemployed); of consolidating the myriad insurance funds under one general umbrella; and of building basic minimum income support programs for those groups unable to participate in social insurance --principally persons with a weak labor force attachment. Hence, old social assistance programs were upgraded, or similar plans were introduced from scratch (such as the Italian social pension, or the French and Belgian social minimum). Together with the emergence of universal family or child allowance schemes, then, an element of universalism and citizenship-based rights was coupled with the dominant insurance model. If this is the general pattern, it also has its exceptions. Thus, Italy never introduced genuine unemployment insurance, and national health care (since 1978) came to be a mix of universal citizen entitlement and insurance (financed via premiums from earned income as well as general revenues). In fact, Italy's lack of unemployment insurance has, in the high unemployment era of the 1980s, additionally perverted the pension schemes-disability pensions in particular. The Dutch pension system, like Italy's health care, combines insurance principles (it is contributory) with universal coverage and (generous) flat-rate benefits. For employees, there is a mandated second-tier occupational pension insurance. Also the Dutch disability pensions have played a major (and highly controversial) role in reducing labor supply. While most countries have introduced national standards in their means-tested assistance programs, this is not so in the Mediterranean
Thus, disability pensions (which are not subject to normal contribution requirements) have been granted at the rate of three times as much in the high unemployment areas of the South as in the North. Besides the objective of diminishing unemployment, they have also been a chief means of political clientelism. On the specific clientelistic aspects of the Italian system, see Paci 1989. Italys only genuine unemployment protection scheme is the Cassa Integrazione, meant to guarantee incomes among the partially or fully laid-off workers in the primary sector economy. Since it is financed by the general pension funds (INPS), the currently very high redundancy rates add substantially to pension insurance deficits.
3

- Error!Marcador no definido. -

countries where social aid remains closer to the classical poor-relief model. Besides cautious universalization, the single most important postwar change involved the adoption of the adequacy principle in cash benefits. This entailed a shift from contribution- to earnings-based benefit calculations and, hence, the adoption of a non-actuarial pay-as-you-go practice. The move was motivated by familialist and status maintenance concerns. In essence, the principle of a _family wage_ was extended to social transfers, as was the principle that social benefits should mirror accustomed status differentials. As a consequence, Continental European, and especially Southern European, pension replacement rates are among the highest in the World. Pioneered in Germany (1957) and then subsequently followed by the other nations, benefit levels are now pegged to previous earnings (see Appendix Table 1). The average old age pension in the EEC is 81 percent of net earnings; around 66 percent for sickness and unemployment. National variations are substantial, particularly with regard to unemployment and sickness. The social minimum programs are much more residualistic. Hence, as seen in Appendix Table 1, the social minimum for the aged in no case approaches 50 percent of per capita GDP. The consequence, as comparative research shows, is high poverty risks in groups, such as widows or single mothers, which depend on non-insurance transfers. In contrast, poverty rates are very low among standard career households. Poverty rates (for the mid-1980s) among retired couples and non-aged families with a fulltime earner are as low as in Scandinavia. For households without a main (male) breadwinner, however, the safety net tends to be fragile. The child poverty rate for single parent households is 18.7 in Italy, 19.2 in France, and 26.7 in Germany (compared to 4.5 percent in Sweden).
4

These figures are based on my own calculations from the LIS (Luxemburg Income Study) data sets using the standard 50 percent of (equivalent) median income as the poverty benchmark. Similar results for 1980 (without Italy) are provided in Mitchell (1991). The Nordic (and to an extent the Dutch) universal flat rate pensions (with automatic supplements) ensure that none will fall below 50 percent of median income (see also Palme 1990). A recent EEC study (EEC 1993) has tried to simulate key cases. Thus, a single aged person with an interrupted career would receive about 60 percent of the standard pension in France, Germany and the Netherlands, only 23 percent in Italy,

- Error!Marcador no definido. -

Two issues are of particular importance for income replacement. The first has to do with employment and contribution requirements. Although the precise norms vary, the basic pension usually requires around 30 or 35 years' contribution. To arrive at maximum benefits, perhaps 40 or even 50 years (as in Germany) are required. These norms have been seriously weakened as governments sought to induce early retirement. In Italy, until recently, 25 years' employment sufficed for public sector workers, giving rise to the concept of _baby-pensioners_. Still, as far as pensions are concerned, a long and stable employment career is still assumed. Hence, it is quite inevitable that social insurance tends to be male-biased. German data bring this out well. Female manual workers average only 22 years' contribution (36 years among males), and female white collar employees, 27 years (38 among males). Only 8 percent of females (compared to 53 percent of males) arrive at 40 years' contributions (Scheiwe 1994: Tables 9.3 and 9.4). The consequence is very low average pension benefits for women.
5

The second issue regards upper ceilings for contributions and benefits. If benefit ceilings are drawn relatively low, higher income employees will naturally seek complementary private protection. A certain degree of redistribution would result if contribution ceilings were comparatively higher. A hallmark of the corporativistic principle in Continental European social insurance is that the two tend to go together. In most countries, both ceilings are quite high; in some cases, such as Italy, they have been abolished altogether (Palme 1990: EEC 1993). In Germany, though, replacement rates begin to fall for persons earning twice the average wage; in Spain, at around three times the average wage (EEC 1993:8890). The somewhat brusque decline in Germany (except for civil servants) explains why, there, complementary private pension plans are quite widespread among

compared to 93 percent in Denmark. For women who worked during their life, the average pension benefit is less than half the males. For women entitled to a widows pension, the benefit is about 60 percent of the males average pension (Scheiwe 1994: Table 9.5).
5

- Error!Marcador no definido. -

high income groups (Esping-Andersen 1990: Table 3.1).

When even the elites perceive social insurance as adequate, the welfare state is likely to enjoy solid loyalties. But this also means heavy public expenditures which, if the schemes were fully funded, might not provoke crisis. The problem is that, not only are they not, but that they crave more and more government subsidization. Fears of imminent fiscal collapse may nurture more private insurance, but so far its role has been marginal. Private pension spending (1980 data) accounts for 1-4 percent of all pension payments in Continental Europe, compared to 10-12 percent in Britain, Canada, and Japan, 17 percent in the US, and a full 20 percent in Australia (Esping-Andersen 1990: Table 4.3). Catering to the upper strata has, however, its price in terms of egalitarianism. The top-quintile pensioner group in Germany receives almost 30 percent of total public pension transfers, compared to 20 percent in the UK. In contrast, the bottom German quintile receives only 11 percent, compared to 17 percent in the U.K. (West-Pedersen 1994). In terms of private pensions, the distribution is much more skewed everywhere. The top quintile pensioners receive 65-70 percent of the total in Germany, as well as in the UK and the Netherlands. The Continental European welfare state is, in other words, distinctive for its generous social transfer benefits and underdeveloped social care services -- both a mirror image of its familialistic bias. The difference between Europe and Scandinavia in welfare state service commitment can be seen in Appendix Table 2. Systematic cross-national comparisons of service delivery are rare but, to
6

On the financing side, contribution ceilings are very high in France and Germany, and nonexistent in Belgium and Italy. As was the case in the recent U.S. social security reforms, the abolition or upward adjustment of contribution ceilings has been a means to raise additional revenues. There are some indications that growing concern with the financial viability of existing pension schemes is giving rise to private plans, be it of the occupational type or individual annuity schemes. In France, life insurance type plans are growing at the rate of 20-25 percent per year; in Belgium there is a rapid growth of company plans (Grasso 1991: Appendix). Also the Netherlands has recently liberalized conditions for private schemes. Note, however, the widespread kindergarten tradition for pre-school age children, a system that is virtually universal in Italy. Indeed, its origins are closely tied to the Church. Note also the sometimes important role played by religious voluntary organizations in the field of care for socially
7

- Error!Marcador no definido. -

exemplify, Gustavsson and Stafford (1994) show that Swedish day care covers about 50 percent of small children compared to 2 percent in the Netherlands. Zimmermann (1993:214) estimates the equivalent German rate at 1.4 percent, while Saraceno (1990:32) cites a figure around 5 percent for Italy; France provides public day care centers for only 3.7 percent of small children, but encourages a much larger network of private day care (OECD 1990:191).
8

Differences in service provision have profound effects on the employment structure. In Denmark and Sweden, public sector employment in health, education and welfare services reached 25 percent of the labor force in the late 1980s, compared to a range of 6-11 percent in Continental Europe (7 percent in Germany, 11 percent in both France and Italy) (Esping-Andersen 1990:158). The comparison with the Nordic countries is pertinent because also there high labor costs prohibit most families from consuming equivalent market services. The lack of either market or state provided care is a source of low female activity and fertility rates. The resulting overall low employment levels imply very unfavorable ratios of dependants to actives, meaning a very high marginal tax on the few that actually do work. In addition, as I shall discuss below, it is arguable that supressed female employment is becoming a chief obstacle to flexibilization.

Expenditures and finances

The great strength of social insurance is that it nurtures a sense of equity: you get what you have earned and you earn what you get. In terms of securing broad legitimacy, this is a strong point in the _Continental model_. Originally,

disadvantaged and needy persons. A study conducted by the European Community (1990:10) comes to fairly similar results. If private day care solutions are included, however, France and Belgiums coverage rate approximates 20 percent of small children.
8

- Error!Marcador no definido. -

social insurance was meant to be strictly actuarial: benefits supposedly directly related to contributions. Originally, financing was shared equally between employers and employees. Today, all are pay-as-you-go systems and their financial structure has changed. In part, as a means of dampening wage growth via the _deferred wage_ strategy, employers and government now absorb a larger share of the financial pie. Of total social security finances, employers now account for 50-60 percent and employees for 30 percent, the remainder being covered by general government revenues. Due to expensive early retirement and rising deficits, government subsidization of pension schemes has risen sharply in the past decade. The rise in benefit levels, combined with a massive increase in beneficiaries (due to both ageing, early retirement, and the inclusion of non-insured groups) over the past decades, has produced, internationally, extremely high financial requirements. The result is fiscal imbalances and rising labor costs. Fixed labor costs as a percent of the total wage bill (for a married worker with two children) hover around 50 percent in Belgium, France, Germany, Italy, and the Netherlands (EEC 1993:83). This, as we shall see, arguably contributes to Europe's chronic employment problem. As seen in Appendix Table 3, the financial status of the European social security systems has, with few exceptions, deteriorated. To cover contribution shortfalls, government subsidies have grown which, in turn, adds to current budget deficits. It is estimated that about half of Italy's current annual public debt (at about 13 percent of GDP) goes to cover revenue shortfalls in the social security system. There was a noticeable slowdown in social expenditure trends during the 1980s but, except for Germany and Belgium, total spending as a share of GDP was higher in 1991 than in 1980 (see Appendix Table 2). Annual growth rates (in constant prices) averaged around 1-1.5 percent in Belgium, Germany and Netherlands; 3 percent in France; 4.5 - 5.0 percent in Spain and Italy. Besides a

- Error!Marcador no definido. -

catch-up effect in the latter cases, the common thrust behind rising expenditures, and thus financial burdens, can be traced to the burden of social insurance in managing the social and economic problems of the past two decades. The rising financial burdens are mainly concentrated in three programs. One is rising expenditures due to high and persistent unemployment. _Passive_ unemployment compensation expenditures absorb (1992) between 1.5 and 2.0 percent of GDP in our countries (except Italy where they are masked as pension payments or Cassa Integrazione expenditures). The rising burden of transfers to the unemployed has, moreover, the unfortunate effect of crowding out _active_ labor market measures such as training, job creation, and mobility measures (OECD 1993). The second area is health care where rising costs are closely related to population ageing. Thus OECD (1988: Table 31) estimates that, by the year 2000, up to 50 percent of all health spending will go to the aged. Thirdly, as we have seen, population ageing and early retirement has produced a cost explosion in pensions. The financial difficulties of pension funds are aggravated by their occupational fragmentation: the decline of some labor force groups, like miners and manual workers, means fewer contributions and generally more early retirement. The ratio of pensioners to contributors has been worsening in all the European countries. In Italy, the number of pensioners as a percent of contributors rose from 60 percent in 1971 to 84 percent in 1987, and will soon reach parity. Record low fertility rates will seriously worsen the dependency burden in future decades. To put it differently, Europe's welfare states have become _pensioner_ states, not because ageing is more advanced than elsewhere, but rather because of their policy bias in favor of passive income maintenance and labor supply reductions. Thus, the ratio of social spending on the aged to the non-aged is generally much higher (the average is around 1.5) than in, say, Scandinavia. Italy is an extreme case where, in 1990, almost 5 times as much was spent on pensioners as on the non-aged. And, in contrast to the Nordic countries, the pro-

- Error!Marcador no definido. -

aged bias is strengthening over time.

Where pension schemes are occupationally exclusive, the fiscal imbalances will vary sharply according to structural change. Thus, today, member contributions in the German miners' pension scheme barely cover 30 percent of expenditures; for the worker's scheme, roughly 70 percent. In contrast, the salaried employee scheme is able to balance its expenditures with contributions (Statistisches Jahrbuch 1992: Tables 19.1 and 19.4.9). We have already noted the possible negative employment effect of high social contributions and labor costs. An additional factor has to do with the punitive tax-treatment (and social transfer effect) of wives' earnings that is built into some European tax systems. This additionally discourages female (full-time) employment in general, and mothers' labor supply in particular (Gustavsson and Stafford 1994; Zimmermann 1993). Thus, for a one-child family in which the male earns average wages, the wife's decision to work full-time (at earnings equal to the male) would have negative consequences for net disposable income. In France, the effect is modest (an elasticity of .93), but it can be quite severe in other countries (an elasticity of .71 in Belgium, .58 in Germany, and .52 in the Netherlands). This contrasts with perfect neutrality in Sweden. For families with three children, the punitive effect is substantially stronger. (Calculations based on EEC 1993: Table 24). Thus, if we combine the tax-benefit effect of wives' employment and the widespread absence of child and other family services, we have also identified a primary cause behind the generally low female activity rates. In turn, female inactivity has the double effect of holding down the size of the actively employed population which finances the welfare state, and perpetuating the family's dependence on the male breadwinner's earnings and entitlements. This translates
9

Thus, between 1980 and 1990 the ratio rose from 1.36 to 1.49 in France; from 3.18 to 4.75 in Italy; and from .95 to 1.14 in Belgium. In Denmark it fell from .76 to .55; in Sweden from .86 to .83. A slight decline is also registered for Germany (from 1.39 to 1.35). (Calculations based on OECD, New Orientations for Social Policy. Paris: OECD: 1994: Tables 1b and 1c).

- Error!Marcador no definido. -

into high transfer levels.

3.

A diagnosis of contemporary welfare state problems

The European welfare states, as they exist today, were molded in the social, economic, and demographic conditions that obtained in the _fordist_ era of full employment. Many of the assumptions that underpinned Continental European social policy in this era were also policy objectives. Primary among the underlying assumptions was stable and vigorous noninflationary growth with full employment. Hence, whether or not governments actively committed themselves to a keynesian counter-cyclical regime (which, indeed, was hardly the case in either France, Germany or Italy), they could assume growth via the sustained expansion of domestic and international demand for their manufacturing goods. Most European countries experienced after the war a heavy outflow of unskilled rural populations. This latent welfare problem was primarily resolved by the expansion of industrial mass production. Relying chiefly on the market to secure high wages and job growth (and, thereby, rising aggregate consumer demand), the chief task of welfare policy was to secure families during the passive stages of the male breadwinner's life cycle --old age in particular. At the microlevel, working class families could count on life-long income security. Thus was born the standard worker and the standardized life course. On the basis of these premises, the welfare state was a success. All evidence suggests that old age poverty diminished drastically with the upgrading and universalization of pensions (Mitchell 1991). The welfare state's relative passivity with regard to the active years of the life cycle, limiting itself to insuring against unanticipated sickness or disability, and against the (assumed) marginal risk of unemployment, remained until recent years a minor problem.

- Error!Marcador no definido. -

Contemporary demographic and labor market changes imply that this residual attention to the active phase of the life cycle is becoming problematic. Marital instability is rising everywhere, creating in its wake a bundle of new poverty risks. In countries like Belgium, Germany and France, single parent households now account for roughly 10-12 percent of all child families; and they have a high incidence of poverty: 27 percent are poor in Germany; 19 percent in France and Italy (own calculations based on Luxemburg Income Study data), and they are increasingly reliant on public assistance. In France, for example, the number of one-parent families on social assistance grew by 185 percent between 1978 and 1986 (Room 1990:52). A very similar story emerges among the unemployed. Between 1970 and 1986, the unemployed's share of all assistance recipients jumped from less than one percent to 33 percent in Germany; from 3 to 67 percent in the Netherlands (Room 1990: 62). The postwar full employment concept was usually limited to the male breadwinner. As agriculture and urban self employment declined, the working class, even the unskilled, could count on well-paid industrial employment. It was this context that permitted policy makers to assume (and actually encourage) women to be full-time, full-career housewives. In this respect, social policy and industrial relations acted in tandem. The family's virtually complete dependence on the male earner's income and entitlements meant that unions came to battle for job security (seniority principles, the regulation of hiring and firing practice), and the _family wage_. The principle of a family wage has, nowhere, become as institutionalized as in Continental Europe, in wage bargaining as in the social benefit structure. This helps account for the comparatively very high income replacement rates in social security. In other words, social policy both assumed and created the _standard worker family_. Indeed, in response to severe manpower shortages in the 1960s, the strategic choice of countries like Germany, Belgium and France to draw on foreign workers, rather than domestic women, only helped reinforce this logic. While women's labor force participation began to accelerate in Scandinavia and

- Error!Marcador no definido. -

North America from the 1960s onward, it remained basically stagnant in the Continental European economies until the 1980s, after which female activity rates have begun to rise. Linked closely to the model family, the postwar welfare state premised itself on a set of socio-demographic assumptions. Primary among these was the organization of the (male) life cycle between inactivity and employment. Thus, entry into the labor force was assumed to occur at age 15-16; retirement at 65, thus permitting 40-45 years of active employment (with social contributions) before retirement. As regards the female life cycle, it was assumed that a short period of employment in youth would be followed by more or less permanent withdrawal at time of marriage and birth. This would, in turn, safeguard the availability of women for full-time social care of children and, later, of the aged family members. Put as briefly as possible, the postwar welfare state assumed full-time, lengthy, and unbroken male careers followed by relatively few years of retirement prior to death, and full-time female careers dedicated to social reproduction activities. This principle was not only stronger on the Continent than elsewhere, it has also been perpetuated until today. Hence, their labor force is comparatively very masculine, and very protected. As elsewhere, these underlying conditions no longer obtain. And, again as elsewhere, the impending crisis of the system is chiefly attributed to population ageing. Certainly, the Continental European nations do exhibit an unfavorable age-profile. Yet, there are several reasons why ageing may not be the chief problem. The dependency burden is primarily a function of fertility and retirement behavior, and both are intrinsically welfare state issues. The crisis of ageing in Europe is, to a great extent, the bi-product of the welfare states' labor reduction strategy. A second set of conditions that no longer obtain has to do with employment structure. On the one hand, industrial employment is declining rapidly, and

- Error!Marcador no definido. -

services constitute today the chief source of job growth. On the other hand, it is widely acknowledged that the labor market requires greater flexibilization. The latter, as is well known, involves both negative and positive trends: on the negative side, a departure from the standardized, secure employment relationship in favor of more precarious, temporary contracts and greater wage differentiation; on the positive side, greater adaptability to new skill and technology requirements in the workforce (Rodgers and Rodgers 1989). In both instances, workers are decreasingly able to count on the _fordist_ promise of life-long job stability with steadily rising pay. The post-OPEC policy response in Continental Europe differed sharply from that of both Scandinavia and the neo-liberal de-regulation approach favored in Britain, the United States, and the Antipodes. Rather than following the _labor cheapening_ strategy of flexibilization and labor market de-regulation that characterizes the Anglo-saxon political economies, or the public employment-led strategy of the Nordic countries, the dominant approach in Europe has been to manage labor market problems via supply reductions. To a (relatively modest) degree this led to a re-export of foreign guestworkers, but the dominant policy came to concentrate on early retirement of male workers and discouragement of female participation, often combined with hours-reduction, the hope being that this would ease youth unemployment. This strategy, it is clear, is intricately linked to the social insurance system. Hence, besides marginal benefit adjustments and the like, a common trait in our group of countries is the visible lack of any major attempts to scale-back or reform social security (except very recently, as we shall see below). Indeed, the political economy of these nations came to hinge on preserving the welfare state edifice. There is a convincing case to be made that welfare state suppression of labor supply will have positive net welfare consequences if, that is, it is mainly unqualified workers, or workers with redundant skills, who are eased out of the labor market. A hallmark of the modern economy is that demand for unskilled workers is in rapid decline. One response would be to allow their wages to decline

- Error!Marcador no definido. -

accordingly --the American approach in a nutshell. This has clearly negative distributional consequences. The other, as in Scandinavia, is to provide sheltered publicly subsidized employment. The Continental strategy of easing them out via welfare programs is based on three increasingly questionable assumptions. One is that the resulting productivity dividend will more than offset the financial cost of passive income maintenance. The second is that those whose supply is effectively suppressed are, indeed, unqualified low-productivity workers.
10

And the third is

that the current surplus of unskilled manpower is transitory; that the problem will go away once the process of industrial restructuring has been completed.

As we have seen, social expenditures continue to rise as the ratio of inactives to actives worsens. Besides this source of revenue shortfall, the drastic curtailment of the average worker's active contributing years also adds to the financial disequilibrium. Thus, compared to the orthodox assumption of 40-45 active years, the average worker in our countries will enter into employment around age 18-20 and retire at age 55-59 which, with heightened probabilities of unemployment and job-loss along the life cycle, amounts to maybe an average of 35, or at most 40, years' active employment; i.e. a decline of 10-20 percent in terms of financial contribution. Our average worker will also collect pension benefits for many more years than the orthodox model assumed, partly due to earlier withdrawal, partly to longevity. Since 1950, male life expectancy (at birth) has risen 5.9 years; for females, 8.5 years. Hence, a male who today retires at age 60
The validity of the second assumption is quite shaky. We know that there are very little differences in early retirement behaviour between unskilled workers and qualified (even professional) workers. We also know that womens educational attainment today equals --and in many countries --even surpasses that of males. This assumption is also questionable since it will hold only if new labor force cohorts enter with the kinds of skills that are, and will be, in demand. However, it is widely recognized that existing education and training systems fall far short in this regard. The average academically trained employee will enter his/her first job at age 25-26 (in Italy) or 28-30 (in Germany). All data suggest that their retirement choices are not very different from manual workers. I.e. they are likely to contribute little more than 30-35 years. When we also consider their higher pension benefits and longer life duration, the contribution-pension cost ratio for this group will be substantially worse. For more detailed data on retirement behavior, see Kohli et.al. 1993.
12 11 10

12

- Error!Marcador no definido. -

can count on 17-18 years' pension, females on 20-22 years. Combining and averaging the two, the standard individual worker pays 15 percent less towards retirement and collects 30 percent more than was the case in the golden 1950s. The real demographic problem in Continental Europe is not ageing but low fertility and low activity rates. Hence, in contrast to the Nordic countries, German, Italian or Dutch women face a rather powerful trade-off between careers and family formation. Italy's fertility rate is today the World's lowest (1.3), followed by Spain, and Germany; Ireland and Sweden, albeit for different reasons, enjoy the highest. In fact, Gustavsson and Stafford (1994) argue that Sweden's increased fertility is positively related to access to day care facilities and flexible employment conditions (part-time and flex-time, and liberal conditions for paid absence). These are exactly the elements lacking in the Continental European welfare states' policy menu.
13

The consequence of basic demographic trends is a rising aged-dependency ratio. Age-dependencies, however, are manageable depending on productivity performance (GDP growth) and labor supply behavior. Thus, the demographic scenario is subject to changes in retirement decisions, to the average age of employment entry, to male and female activity rates, to immigration, and to economic performance. The point is that the European welfare states have managed these variables in such a way as to seriously aggravate the underlying problem.

The employment problem

The lack of day care and similar facilities does not unequivocally arrest womens employment. Participation rates among the younger cohorts are steadily increasing everywhere. This may be the main factor behind declining fertility. Part-time employment may help combine both if, that is, employers go along. In Italy they evidently do not, citing the punitively high fixed labor costs. However, as Kloosterman (1993) suggests, Hollands introduction of more flexible wage schemes and lowered social costs in the 1980s has given a formidable boost to female (part-time) employment, while female full-time jobs have hardly changed at all.

- Error!Marcador no definido. -

Why are the Continental European welfare states so uniquely incapable of generating employment growth? For the EEC as a whole, the employment share among the working-age population actually fell from about 64 percent in 1968 to 57 percent in 1988 (EEC 1989). In contrast, the United States, starting at the same level, arrived in 1988 with 72 percent; and Sweden with more than 75 percent. In part, the European decline has to do with male early retirement. Male participation rates have declined, 1960-1985, by 15-20 percentage points in Germany, the Netherlands, Belgium, and Italy, compared to 6 percent in the U.S. and Canada. And, in part, it has to do with relatively stagnant rates of female participation. Compared to around 80 percent in Sweden, overall female participation is only 50 percent in Belgium and Germany, 55 percent in France, and 40-45 in the Netherlands, Italy and Spain (OECD 1993). Since the mid-1980s, female participation has risen, especially in the young cohorts in Germany and the Netherlands where part-time employment has become common. Nonetheless, the total level of net job generation has fallen far short of supply, thus creating very high and chronic unemployment levels, especially among youth. Except for Germany, average unemployment levels have hovered around 10 percent through the last decade. The incidence of long-term unemployment is now between 50 and 70 percent of all unemployment in Belgium, Italy, Spain and Germany; about 40 percent in the Netherlands and France. And, as perhaps the most sensitive indicator of labor market exclusion, youth unemployment rates tend to be extremely high in countries like Italy (33 percent), Spain (35 percent), France (22 percent) and Belgium (20 percent). In other words, employment conditions in Continental Europe display all the characteristics of _insider-outsider_ labor markets. Thus, low levels of unemployment among adult male workers combines with huge populations of excluded or marginalized workers. In general terms, this can be ascribed to industrial job-decline combined with very sluggish service employment growth.

- Error!Marcador no definido. -

What is unique within our countries is the slow rate of service growth and, in some sectors, actual stagnation and decline (particularly in personal services). Thus, from the mid-1970s to the mid-1980s, the EEC average rate of service growth was less than half the American. As neo-liberal critics repeatedly assert, _Eurosclerosis_ is the step-child of the welfare state and industrial relations institutions working in tandem. The lack of service job growth (except in the high-end services, such as business, finance and the like) can be attributed to the classical cost disease problem (Baumol 1967). Since most personal and social services enjoy only modest, if any, productivity improvement, they will fail to create jobs if wages and labor costs follow trends in high productivity sectors. Egalitarian and high wage (family wage) structures, as exist throughout Europe, will therefore render private sector services, like laundry, carwash, bellhops, or day care, inordinately expensive. One solution is to furnish services via public sector employment. Another is to encourage wage flexibility. For obvious welfare state related reasons, neither alternative has been possible in Continental Europe. To safeguard the family wage, unions cannot allow low wage markets; social contributions raise fixed labor costs, thus making the hiring of one additional worker marginally very costly. And fixed social contributions have been rising steeply. Catholic familialism aside, the very heavy transfer burdens (and public debt payments) with which European welfare states are already burdened would prohibit any significant public employment strategy.
15 14

The result, then, is weak public and private sector service employment growth. With the continuation of traditional industrial relations practice and job rights legislation, the consequence, in turn, is an increasingly closed and rigid _insider_ labor market with ever narrower ports of entry. Hence, low female
Recent data on earnings differentials bring this out. While the lower quintile income earners in America, Britain (and also Australia) have lost substantial ground over the last decade, this has not been the case in either Germany or Italy (see OECD 1993; Gottschalk 1991). Public debt as a percent of GDP was, in 1989, 128 percent in Belgium, 99 percent in Italy, and 79 percent in the Netherlands. Germany (with 43) and France (with 36 percent) would clearly be better positioned were it not for the upward trend. (OECD 1991)
15 14

- Error!Marcador no definido. -

participation and massive youth unemployment in concert with a secure, predominantly male, insider labor force which relies precisely on the guarantee of life-long employment with high wages and heavy social contributions in order to assure family welfare across the life cycle. The point is that the particular kind of welfare state-family-work nexus that characterizes the Continental European model has an inbuilt tendency to eat the very hand that feeds it. Unable to promote employment expansion, it reverts to labor supply reduction policies which, for males, mean unemployment and pension costs; for women, the necessary continuation of male breadwinner dependencies. In both cases, this translates into extremely high labor costs and labor market rigidities because the _insiders_ are compelled to defend their employment security. As such, the labor market remains rigidly closed and incapable of major job provision. The strengthening of insider-outsider cleavages poses obstacles to greater flexibility. From the employer point of view, what counts is wage flexibility (setting wages according to productivity and profits), functional flexibility (such as greater adaptability to new technologies), and employment flexibility (capacity to hire and fire according to need). From the individual and family point of view, flexibility means the capacity to manage dual-career marriages with family obligations, the rising probability of family breakups, the increased possibility of mid-career changes, such as unemployment, re-schooling, and occupational change, and generally more differentiated and less standardized life cycles. The Continental European welfare state model conspires to inhibit either of these flexibility needs. The _insider_ labor force is trapped in its dependence on life-long job security and will thus, via its collective organizations, resist attempts to weaken the _standard employment relationship_. This entrapment is obviously tightened by the shortened span of active contribution years: if a decent pension requires at least 35-40 years' employment, workers will logically struggle to minimize whatever employment securities might emerge.

- Error!Marcador no definido. -

Of course, the welfare state may relax the conditions pertaining to eligibility, or otherwise subsidize career risks. This is, indeed, a major element in social security reform over the past decades. On the one hand, welfare states (especially the German) grant implicit contributions during education, unemployment, and military service, thus allowing individuals the possibility of arriving at standard contribution records despite early retirement and employment interruptions. On the other hand, they seek to nurture familialism by granting pension points (or even mothers' salaries) to women who chose to leave the labor force. Here, flexibility is bought at the expense of additionally imbalanced social security budgets and/or of consolidating the traditional genderdivision of labor. The alternative road to flexibility which, at present, appears much more pervasive, implies an exit from the conventional welfare state-employment nexus. Data here are, by nature, much scarcer, but there is little doubt that employers as well as employees are given particularly strong incentives to recreate the market, so to speak, behind the back of the welfare state. Thus, to escape high fixed labor costs, employers and employees jointly underdeclare wages (and thus contributions) up to the ultimate years of employment which are taken into consideration for benefit purposes. This is notoriously common in the Italian system. Similarly, early retirees and employers have a common interest in an informal employment relationship because the former no longer need to accummulate entitlements. Hence, an estimated 50 percent of early retired males in Italy continue to work; a similar figure has been cited for laid-off workers in the Cassa Integrazione. An alternative is to farm out production in the form of homeworking. According to the most conservative estimates, 700,000 women in Italy, and 500,000 in Spain, produce informally out of their home, being paid piece rates without any form of social contributions; other estimates for Italy put the number at 2 million (EEC 1989; Dallago 1990). Two other trends point towards a similar scenario. One is the rising proportion of self-employed --often disguised franchise workers -- especially in

- Error!Marcador no definido. -

Belgium, France, Italy and Spain. In fact, compared to other types of _a-typical_ employment like temporary workers, this is a main source of employment generation, especially in labor intensive services like restaurants, hotels and entertainment (OECD 1993; EEC 1989). The other is black economy activity which, however, has probably not grown as much as many assume and which, regardless, tends to involve mainly persons already employed (moonlighting), or recently retired. Nonetheless, there seems to be much higher levels of black economy activity in precisely those nations with high welfare state induced labor costs and employment rigidities, while this does not seem to be the case with high taxation nations, such as Sweden. Thus, in Northern Europe, black economy work may account for 5 percent or less of GDP, while in Italy and Spain the level is around 10-20 percent (EEC 1989).
16

All these forms of irregular employment are examples of labor market actors dodging the rigidities and high costs of the regular economy. They have in common one factor which is detrimental to the welfare state: they draw on its benefits without contributing fully to its finances. Thus the vicious circle is worsened further. As to flexibilization, there is mounting evidence that partial flexibilization policies --such as liberalization of temporary contracts -- may actually deepen the insider-outsider divide. Spain is an illustrative case since, here, temporary workers account for virtually all net job growth in the 1980s; their share is now almost a third of total employment. Yet, their chances of attaining a more permanent employment relationship is exceedingly small (10-15 percent). Indeed, it has been shown that the wage and job-security privileges of the permanently employed _insiders_ have strengthened precisely because firms can regulate their labor force at the margin through easily dismissable temporaries (Bentolila and Dolado 1994).

Zimmermann (1993:233-234) suggests, however, that there has been a noticeable rise in black economy work in Germany between 1960 and 1980.

16

- Error!Marcador no definido. -

4.

Future welfare state scenarios

All the advanced western welfare states have experienced a degree of socioeconomic transformation for which they were ill-prepared, mainly because their construction reflects an economic, family, and demographic structure that no longer obtains. There are substantial differences in nations' policy response to these changes. The Continental European countries have, so far, followed a distinct trajectory in which welfare state induced labor supply reductions figured prominently. One result has been to nurture the emergence of rigid _insideroutsider_ labor markets with consequent marginalization and peripheralization, particularly among youth and women workers. The strategy's reliance on existing social security programs has led to a major cost explosion to the point where several nations' social insurance funds are threatened with bankruptcy unless continuously subsidized by public moneys. In response to the self-reinforcing negative spiral of these systems, families and labor market actors often seek flexibility solutions outside the formal system: irregular employment which escapes tax contributions and employment legislation; women, facing a naked trade-off between employment and children, delay and reduce fertility. It should be evident that the more that individuals opt for informal flexibility solutions, the more they help tighten the noose around the system's neck. The system is such that individual rationality adds up to collective irrationality. Public confidence in the welfare systems' continued capacity to deliver on expected benefits appears in many countries to be falling (EEC 1994 Grasso, 1989). There has thus in recent years been a shift towards private pension savings
As popular discourse across many European nations suggests, strong insider-outsider divisions risk producing social dualisms. Thus, the French speak of the "societ de deux vitesses"; the Germans of the "Zweidrittelgesellschaft". In the case of France, the factual base of marginalization is brought out in a recent United Nations report (United Nations European Region 1993:49). In the age group 18-24, 26 percent were unemployed and another 25 percent muddled through by doing "petits boulots".
17

17

- Error!Marcador no definido. -

plans, either in the form of company occupational pensions or individual annuity plans, especially in Belgium, France, Germany, and the Netherlands, but also in Italy. Fears of impending benefit reductions or tightened eligibility rules also have the perverse effect of accelerating the run on early retirement schemes. Welfare state reform has, so far, been limited to marginal adjustments: dampening cost growth via delayed or diminished benefit indexation or via spending caps in health care. Some countries, especially the Netherlands, have taken more direct measures to reduce replacement rates and to encourage private sector insurance. Retirement age is now being raised, most recently in Italy, but this can for obvious reasons occur only very gradually and will, at best, see positive results only in the longer run. In order to combat fiscal fraud, but also to restore greater actuarialism to the system, there is a noticable trend to return to contribution (rather than earnings-) based entitlements in pensions. The reform currently being implemented in Italy may, in fact, spearhead similar reforms elsewhere. On the labor market front the policy profile is similar. Relaxation of parttime employment legislation has, in the Netherlands, helped promote a surge in women's employment; in Italy, however, hardly at all. The explanation is quite simply that employers have very little incentive to offer part-time contracts given the high fixed overhead costs. Similarly, there have been some marginal efforts to lower wage costs and flexibilize job-access for youth by government wage subsidization. However, it is quite clear that marginal adjustments either on the social insurance front or in the labor market will not suffice to bring the systems out of disequilibrium. And, as in the case of temporary contract workers in Spain, more sweeping reforms may actually have perverse effects if they are implemented only selectively. Of course, temporary contracts do help provide first-entry labor market opportunities for women and youth in particular. But all evidence, for Spain as well as for France, indicates that they nourish dualisms while hardly generating additional net employment growth. Partial flexibilization may actually deepen the

- Error!Marcador no definido. -

abyss between the core and periphery workforce. In general, it is evident that a system premised so heavily on the familywage logic and career-long contributions imposes severe rigidities on peoples' life cycle behavior at the micro-level, and suboptimal labor force participation levels at the macro-level. Hence, the most pressing issue for these welfare states is how to encourage, at once, a major expansion in the supply and demand for employment. Lower labor costs is one solution, but it does not address women's trade-off between work and family. Without an adequate supply of family care services, women's entry into the labor market may still rise (as it does), but at the long-term cost of extremely low fertility. The lack of adequate social care provision also risks creating a peculiarly perverse kind of inequity. A dual-career couple within the privileged occupational strata will both earn the _family wage_ and accumulate rights to two _family social wages_. For them, the lack of public day-care is a minor problem since they will be able to afford nannies and maids. This is not likely to be the case for most families; hence, households' earnings power may easily polarize. What will also likely polarize is families' capacity to manoeuvre and adjust flexibly in today's labor market. The individual earner's risk of change, mobility, or jobloss is much reduced in the context of a two-career household. In contrast, the traditional male breadwinner household will logically see such events as a threat. It is therefore logical that trade unions will continue to uphold traditional job rights and favor strong labor market protection policies. It is virtually certain that tomorrow's family and labor markets will demand a drastic reduction in the household's social care responsibilities. It is also becoming increasingly evident that the distribution of social risks along peoples' life cycle is shifting towards the active adult stages; in part because of changing family behavior, including divorce and the rise of single parent households; in part because the labor market will decreasingly offer the kind of classical life-time employment guarantees it once did. On both counts, the Continental European _pensioner_ welfare state model is singularly ill-equipped.

- Error!Marcador no definido. -

Since the financial situation of these welfare states is already so overloaded, it is difficult to envisage any major new reform departures along the lines sketched above. It is, however, equally difficult to imagine radical de-regulation strategies of the neo-liberal kind. Continental Europe's welfare systems were built by the conservatives and these very same forces remain essentially dedicated to their survival. Since, as I have suggested, the population is also locked into dependency on these systems, it is difficult to imagine broad-based support for any drastic welfare state roll-backs. It is symptomatic, indeed, that most conservative-led governments today seek solutions that imply a reinforcement of the existing edifice. Thus, in 1994 the German government passed a social insurance plan for care of the elderly, and the new conservative French government intends to raise additional taxes so as to create subsidized jobs for the unemployed, extend mothers' parental leave, and improve old age pensions. At the same time, of course, cuts are a must --especially in the most financially imbalanced systems such as Italy's. A common characteristic of cutbacks, however, is that they are explicitly meant to safeguard --not destroy -the existing edifice. This is plainly the principle behind the ongoing pension reforms in Italy: sustaining the program by slowly raising the retirement age and by strengthening the importance of contributions for benefit calculations. And those that propose more privatization do so half-heartedly. For one thing, it is obvious to all that if privatization means company welfare plans, the idea is simply unrealistic given employers' already burdensome fixed labor costs. A gradual, albeit probably slow, rise in individual private welfare plans is underway. But it is uncertain whether this, in the long run, will result in a major _system change_. In part, it is fueled by uncertainty regarding the viability of public schemes and if the latter are brought on a surer financial footing, the thrust may be limited to a relative small, high-income clientele. And, in part, a massive surge in individual or group insurance is unlikely unless governments grant favorable tax concessions. It is difficult to see how governments can afford such concessions today. If the combination of familialism and suboptimal activity rates define a

- Error!Marcador no definido. -

crucial element of the Continental European welfare state crisis, it is not easy to see a positive-sum solution in the short-to-medium run. Since a surge of costly social services is an unrealistic prospect, while women's integration in the economy is likely to rise, Europe's very low fertility rates are likely to continue. Financial crisis will clearly compel governments to scale back early retirement practices, and in this respect the labor shedding strategy will abate. But where is job growth going to come from? How is the deepening gulf between insiders and outsiders going to be closed?

- Error!Marcador no definido. -

APPENDIX TABLE 1.

1990 transfer benefit levels as a percent age of net earnings.*

Pensions Belgium Germany France Italy Netherland s Spain 80 77 83 89 67 98

Sickness 100 100 53 31** 70 60

Disability 113 60 46 56 69 32

Unempl. 79 63 80 26** 74 80

Social Min. 32 29 30 16 41 28

EEC 12

81

69

60

61

30

* **

The social minima are expressed as a percent age of per capita GDP. Note that in Italy most employees have the right to full wage continuation during sickness. The unemployment benefit does not refer to the Cassa Integrazione system for industrial redundancies, which will typically replace 70-80 percent of pay. Note also that the Dutch data reflect the 1987 reform's reduction of replacement rates.

Source: EEC, Social Protection in Europe. Luxemburg: EEC, 1993.

- Error!Marcador no definido. -

APPENDIX TABLE 2.

Social security expenditures and the role of services, 1990.

Social Security Expenditure* 1980 Belgium France Germany (W) Italy Netherlands Denmark Sweden United States * ** 28.0 25.4 28.7 19.4 30.8 28.7 1991 26.7 28.7 26.6 24.4 32.4 29.8

Services/Transfers Ratio**

N.A. .12 .16 .06 N.A. .33 .29

As a percent age of GDP Social services excluding health care. Note that France is 1989. OECD, National Accounts. Detailed Tables. Paris: OECD, 1993. And EEC, Social Protection in Europe. Luxemburg: EEC, 1993,.

Source:

- Error!Marcador no definido. -

APPENDIX TABLE 3.

Social security finances in select European countries, 1980-1991. The difference between contributions and benefit expenditures. 1980 1991 -3.2 +0.9 +5.2 -4.3 -2.4 -1.1

Belgium France Germany Italy Netherlands Spain

-5.9 +1.7 +3.1 -3.9 N.A. N.A.

Source:

OECD, National Accounts. Detailed Tables. Paris: OECD, 1993.

- Error!Marcador no definido. -

REFERENCES

Bentolila, S., and J. Dolado. 1994. "Spanish Labour Markets". Economic Policy April: 55-99. Dallago, B. 1990. The Irregular Economy. Aldershot: Dartmouth. Esping-Andersen, G. 1990. The Three Worlds of Welfare Capitalism. Oxford: Polity Press. European Community. 1989. Employment in Europe. Luxemburg: EC. European Community. 1990. Childcare in the European Communities 1985-1990. Bruxelles: The European Commission. European Community. 1993. Social Protection in Europe. Luxemburg: EC. European Community. 1994. Green Paper on European Social Policy. Bruxelles: DG5. Flora, P., and J. Alber. 1981. "Modernization, democratization, and the development of welfare states in Western Europe". In The Development of Welfare States in Europe and America, ed. P. Flora and A. Heidenheimer. New Brunswick, NJ: Transaction Press. Gordon, M. 1988. Social Security Policies in Industrialized Countries.Cambridge: Cambridge University Press. Gottschalk, P. 1993. "Changes in Inequality of Family Income in Seven Industrialized Countries". American Economic Review 2: 136-142. Grasso, F. 1989. "L'evoluzione degli squilibri del sistema pensionistico". In Welfare State e Redistribuzione, ed. R. Brunetta and L. Tronti. Milano: Franco Angeli. Gustavsson, S., and F. Stafford. 1994. "Three regimes of childcare: The United States, the Netherlands, and Sweden". In Social Protection versus Economic Flexibility, ed. R. Blank. Chicago: University of Chicago Press. International Labor Office (ILO). 1975 and 1993. Yearbook of Labour Statistics. Geneva: ILO. Kloosterman, R. C. 1993. "Three Worlds of Welfare Capitalism?" Unpublished paper, University of Utrecht. Kohli, M. , Rein, M., and A. Guillemard. 1993. Time for Retirement. Cambridge:

- Error!Marcador no definido. -

Cambridge University Press. Mitchell, D. 1991. Income Transfer Systems. Avebury: Aldershot. Myles, J. 1990. "States, labor markets and life cycles". In Beyond the Market Place: rethinking Economy and Society, ed. R. Friedland and S. Robertson. New York: Aldine de Gruyter. OECD. 1988. Employment Outlook. Paris: OECD. OECD. 1989. Measuring Health Care. Paris: OECD. OECD. 1990. Lone Parent Families. Paris: OECD. OECD. 1991. National Accounts. Detailed Tables. Paris: OECD. OECD. 1993. Employment Outlook. Paris: OECD. OECD. 1994. New Orientations for Social Policy. Paris: OECD. Paci, M. 1989. Pubblico e Privato nei Moderni Sistemi di Welfare. Napoli: Liguori. Palme, J. 1990. Pension Rights in Welfare Capitalism. Stockholm: SOFI. Rimlinger, G. 1971. Welfare and Industrialization in Europe, America and Russia. New York: John Wiley and Sons. Rodgers, G., and J. Rodgers. 1989. Precarious Work in Western Europe. Geneva: ILO. Room, G. 1990. New Poverty in the European Community. London: Macmillan and the Commission of the European Communities. Scheiwe, K. 1994. "German Pension Insurance". In Gendering Welfare States, ed. D. Sainsbury. London: Sage. Statistisches Bundesamt. 1992. Statistisches Jahrbuch fr Deutschland. Titmuss, R. 1974. Social Policy. New York. Pantheon. West Pedersen, A. 1994. "What makes the difference? Cross-national variations in pension systems and their distributional outcomes". Paper presented at Workshop on Convergence and Divergence, Soroe, Denmark, June 9-11. United Nations European Region. 1993. Welfare in a Civil Society: Report for the Conference of European Ministers for Social Affairs. Vienna: European Centre.

- Error!Marcador no definido. -

Zimmermann, K. 1993. "Labor responses to taxes and benefits in Germany". In Welfare and Work Incentives, ed. A.B. Atkinson and G.V. Mogensen. Oxford: Clarendon Press.