Corps de l’article

Why is the gender perspective so crucial for addressing the challenges of a long life society?

The economic downturn of the past two years, and the accompanying increase in unemployment and poverty in a number of OECD countries have brought to the fore in public debate the policy concern for the sustainability and adequacy of social protection and, more particularly, of old-age pensions. This concern is underpinned by national, European Union and UN forecasts of demographic ageing and extended life expectancy combined with declining fertility rates. These result in the shrinking of the active population that will have to support the growing welfare burden of the retiring “baby-boomers”. This increase in the population dependency ratios underlines the urgency of devising a multi-pronged strategy to respond to the challenges of a long life society (UN 2001a, 2001b; CEC 2003a; EPC 2000; Futuribles 1999). To give an idea of the magnitude of the burden, the European Commission estimated that public spending on pensions would rise between 3 and 4 percentage points of GDP in most European Union (EU) countries between 2000 and 2050, although there is a wide variation among countries, ranging from around 5% to over 20%. Even in the United States, where the demographic situation is more balanced, the General Accounting Office (GAO) of the US Congress sounded the alarm bell when it recently estimated that in 2009, one year after the first baby boomers will be eligible for Social Security benefits, the cash surplus of the fund will start a permanent decline, with income from contributions becoming insufficient to pay currently scheduled benefits by 2018, and the combined Social Security Trust Funds projected to be exhausted by 2042 (GAO 2003: 2).

In response to this grim perspective, the EU Commission has focused its three-pronged strategy for the sustainability of future pensions on raising the employment rates of all population groups, particularly women. Why so? In a nutshell, for the four following reasons:

  • First, at prime age, women have relatively low activity rates in a number of European countries;

  • Second, women have a higher risk exposure to poverty than men, particularly those who constitute one-person households (24% vs. 19% for men), and, worse, the risk of poverty for women heading a single-parent family is close to 40%;

  • Third, they represent the majority of older people (nearly 60% of people over 65 and close to two thirds of those over 75);

  • Fourth, the average level of women’s pensions remains significantly lower in many countries as compared to men’s pensions due to differing employment histories.

Raising employment rates is thus seen across Europe as an important instrument for sustaining the future of pensions, besides achieving greater social cohesion and equality of opportunity, which are basic objectives that underpin the European employment strategy and social model. However, as the Commission warns, the pace of current labour market and welfare reforms undertaken in member States falls short of the necessary adjustments that will be necessary for achieving, by 2010, the three benchmark targets fixed by the Lisbon and Stockholm ministerial Summits for employment rates, respectively 70% of total working age population (15–64), more than 60% for female employment and 50% for the employment of older workers aged 55–64 (CEC 2003b).

One of the major instruments to increase employment rates is the so-called active labour market policy (ALMPs). Activation policies were originally developed in the 1970s to cope with high unemployment and to contain the growing related welfare costs. They are now expected to provide one of the solutions to the problems raised by “long life society” by improving the capacity of the working age population to enter and to remain in the labour market and by “making work pay”. It is therefore important to draw lessons from past experience in addressing the challenges of unemployment and labour market exclusion and to highlight some of their shortcomings in achieving their more limited earlier target of reducing unemployment. Among these shortcomings, one may particularly underline a fragmented approach and inadequate attention to the effect of the major changes that have occurred in the labour force and in the labour market, most of which have an important gender dimension. Indeed, most activation policies focus on selected aspects of the labour market, notably on training, placement and counselling, rather than on attempting to be part of a broader co-ordinated policy mix that encompasses macro-economic measures, taxation, family policy and measures to improve work-life balance.

The present article looks at the changes that have taken place in the EU, occasionally also in the US, in the labour market and in workforce characteristics and how the welfare systems have adapted to them, highlighting successes but also problem areas. It concludes with some insights into possible measures to enhance female participation rates. The article draws essentially on some of the more general findings and conclusions of the recently published ISSA project on the policy implications of the links between labour market and social protection reforms (Sarfati and Bonoli 2002[**]) and on more recent data published in 2003 and early 2004 by various sources, including the EU Commission and OECD secretariat.

Arguably, there is a close link between developments in the labour market, the workforce and the welfare system, since employment patterns affect people’s access to social security benefits, while the rules determining access to social protection benefits influence decisions concerning employment for employers as much as for workers.

Shifts in the Labour Market and the Work Force

The welfare states established after the Second World War were designed to protect citizens, typically the male breadwinners on a life-long and full-time job, against labour market failure during short periods of unemployment. This was achieved by providing economic security to the breadwinner and his dependant family by income replacement through social transfers, by regulating employment protection, and by a commitment to achieving full employment. These characteristics of the breadwinner have been dramatically changed over the past three decades by the following major trends (Sarfati 2002):

First is the emergence of a persistent and high level of unemployment in the OECD region. In the larger European economies, unemployment trebled since 1970, rising from 4% to 11% of the workforce in the mid 90s. Although it declined in the following years, it rose again to 8.1% by June 2003 in the wake of the economic slowdown—representing some 14 million people, half of whom are long-term unemployed (CEC 2003c: 15, 209). Female unemployment is 2 percentage points higher then males’, with the incidence of long-term unemployment (LTU) 1 percentage point higher. In the US, by contrast, unemployment was more than halved from 9.5% in the early 80s to 4% in late 2000, but it has increased in the following two years to a still manageable 5.8%, with 8.5 million unemployed. Since the end of the recession in 2003, economic growth seems to be poor in jobs (EPI 2003 a, 2003b; Christian Science Monitor, 8.12.2003; Financial Times 9.01.2004; NYT 10.01.04), and the US Bureau of Labor statistics is relatively reserved about the kind of job growth prospects in the next decade (BLS 2004).

Second trend, the EU larger economies have suffered declining and low employment rates in the population, particularly among the young age cohort (15–24) and older workers aged 50 +, whose activity rates stood on average at 40% in 2002 and are not expected to grow in 2004 (CEC 2003c: 201), but differentiation among countries is huge. For example, Denmark and the Netherlands have youth employment rates well above 60%, and Sweden has close to a 70% employment rate for older workers. Youth and older worker employment rates are below or slightly above 30% in Belgium, Italy, Greece, France and Spain. This reflects late entry into, and early exit from, the labour market, associated with the afore-mentioned high incidence of unemployment. At the same time, female activity rates present a contrasted picture: while the smaller European and Anglo-Saxon countries enjoy high female employment rates, on average, female employment rates lag some 16 percentage points behind their male counterparts and the recent slight narrowing of the gap partly reflects declining male employment rates (52% vs. 68% in 2003; CEC 2003: 201).

Notwithstanding the previous trend, the third change consists of a massive increase of women’s participation in the labour force on both sides of the Atlantic, which is one of the most prominent features of post-war economic development in the OECD region. Over the past decade, employment rates rose substantially more (or declined less) for women than for men in this area. As a result, inactivity rates across the OECD declined by 2.2 percentage points for women vs. only 1.4 percentage points for men. Women also tended to benefit more than men where unemployment declined (OECD 2003a: 29). In the EU, female activity rates rose by 5.3 percentage points, reaching 60.9% by 2002, while men’s activity rates slightly declined over the same period by 0.6 percentage point to 78.4%. However, while the gap in female employment rates between EU and US has been closing rapidly (from about 18% in 1993 to 11% in 1999), it is still significant and may be attributed to the slower growth of the services sector (see below). Narrowing the gap is mainly due to improved female activity rates which are enhanced by structural changes in the economy—the shift to a post-industrial society—, and institutional changes in the labour market, notably the development of part-time employment.

This naturally leads us to the fourth trend, which consists of the rapid growth of so-called “non-standard jobs”—part-time, temporary, independent or quasi-independent jobs—involving a third of the active population, particularly women, but also young people and older workers. On average, around three-fourth of all part-time jobs are occupied by women, and their share in part-time work is higher than 60% across the OECD region (OECD 2002: 68–69). Part-time (less than 30 hours per week) constituted a substantial contribution to job growth, particularly for women, youth and to a lesser extent older workers. In some countries, it offset declining full time employment, while in others it contributed half the total job growth (OECD 2003a: 49).

Looking at the shares of different forms of non-standard jobs in total employment in the EU over the past decade (1992–2002), there was a decline of total self-employment from 16.2% to 14.6%, and from 12.7% to 10.7% of female self-employment. By contrast, part-time employment increased by four percentage points on average for all workers from 14.2% to 18.1%, but by close to five percentage points for women (from 28.8% to 33.5%). The share of fixed term contracts in total employment increased by about 2 percentage points from 11.1% to 13.0% (after having peaked at 13.6% in 2000); the share of female fixed term employment remained 2 percentage points higher than males’, rising from 12.5% to 14.2% over the decade (CEC 2003c: 209). These figures show that, across the EU, women are not only more likely than men to be employed in part-time jobs but also in temporary ones, though the latter tended to increase slightly more for men in recent years. The incidence of temporary work is much higher in Spain (a third of both women and men), and relatively higher in Portugal and Finland (15–20%). Moreover, in Finland, Belgium and the Netherlands, well over half of all temporary workers were women. Temporary working is more prevalent among the younger age cohorts and is often combined with education. Thus, in 2000, just fewer than 54% of women and almost 58% of men in fixed-term jobs were under 30. Among employees aged 30 and over, however, the proportion of women with temporary jobs exceeded that of men throughout the EU, and the gap was particularly large in Belgium, the Netherlands and the Nordic countries. While temporary jobs can be seen as an option out of unemployment, many of these jobs are vulnerable to spells of unemployment and are an involuntarily chosen route by a third of men and women due to lack of opportunities for permanent jobs. However, in most EU countries, more women than men were in temporary jobs out of choice (Eurostat 2002).

In the US, non-standard employment remained common even in the period of strong economic growth of the 90s, employing more than a quarter of the total workforce in 2001 down from a third in 1995, but still 31% of women, down from close to 34% in 1997. By share of work arrangements, women held more than half of all non-standard arrangements, of which two-thirds of regular part-time jobs, close to 60% of contacts by temporary work agencies, about 45% of on-call assignments or day labour and 36.7% of the self-employment. As regards choice, three-fourths of women in regular part-time preferred their current work arrangements, and slightly less than half of women “temps” or on call wanted a regular full-time job (EPI 2003c).

Atypical (or non-standard) jobs are often associated with low pay, limited career opportunities, limited access to training and less social and employment protection than full time jobs. They obviously involve fewer contributions to the welfare system and reduced pension entitlements, particularly for the self-employed, persons with discontinued careers, temporary jobs or short part-time jobs with less than 10 hours per week. The incidence of short part-time jobs is far from negligible, constituting no less than about 10% of total employment in the Netherlands, around 6% in Australia, Canada, Denmark, Norway, Switzerland and the UK, and about 4% in Finland, Germany and the US (Fitoussi and Passet 2000: 64).

A recent survey of non-standard jobs in the US shows that wages for part-time, temporary work and on-call/day workers are lower than wages of regular full-time workers, even after adjusting for personal characteristics such as age, education and race. Health insurance coverage falls below that of regular full-time workers, as only 14.8% of women and 12.4% of men in non-standard jobs are covered by employer provided health insurance, compared to 66.8% of women and 70.8% of men in regular full-time jobs. Typically, pension coverage is also lagging—with only a fifth of women and 11% of men in non-standard jobs covered by employer-provided pensions, compared to 66.8% of women and 70.8% of men in regular full-time jobs. In spite of improvement in benefits during the economic growth, these gaps are still important (EPI 2003c: 2).

While for youth and middle-aged men atypical jobs often constitute a stepping stone—to standard jobs for the former, to retirement for the latter—, for women, these very often constitute the only opportunity for jobs, with few career perspectives or a guarantee of a decent living standard and social coverage, particularly for pensions. Arguably, some part-time jobs, notably for women, are not necessarily precarious as they can also be of long tenure. Thus, for example, in the EU in 1998, while only 10% of men with a full-time job had a temporary contract, this proportion rose to over 30% of male part-timers. By contrast only 15% of female part-timers had a temporary contract, indicating that part-time is a more “regular” form of work for women (Auer and Cazes 2003: 45). On the other hand, the Third European Survey of working conditions (2000), carried out by the European Foundation for the Improvement of Living and Working Conditions, indicates a high prevalence of long tenures of ten years or more for both men and women not only in full-time (45% of men vs. 40% of women), but also in part-time jobs (36% vs. 33%, respectively). However, the incidence of precarious status of part-timers of either sex is seen in the greater likelihood that they hold a temporary contract than full-timers (16% of women part-timers and 21% of men part-timers hold a fixed term contract or are temporary agency workers compared to 11% of women and 9% of men full-timers) (European Foundation 2002: 17, 26). As regards the tenure of temporary workers, data for 2000 shows that a third of both men and women have contracts of less than 6 months duration, and about another third had contracts of between 6 and 12 months. Longer contracts of over 3 years were limited to fewer than 16% of men and just over 10% of women temporary workers (Eurostat 2002).

The rapid growth of non-standard jobs has been clearly heavily influenced not only by the rise and persistence of unemployment, but also by the significant change in thecomposition of families, which constitutes the fifth trend. This is characterized by the drastic increase in single parent and single-person households, many of whom are women. The latter raise their children alone and are often obliged to accept low-paid, low-skilled and precarious jobs if not to remain outside the labour market altogether and depend on welfare. Lone motherhood limits their capacity to enter and remain in gainful employment and to contribute and benefit from social protection. Among the single-person households, many are older people living alone and having few resources. The situation will worsen as the population ages, as more than a third of persons aged 65–79 are living alone, and their percentage reaches 45% for those aged 80 + (Oxley 2000). The trend towards one-person households is indeed on the rise across the EU and is expected to continue in the forthcoming decades, rising from 14 million in 1961 to 42 million by 1995 and reaching between 51 and 71 million by 2025 (Eurostat 2003a). Among these the very old age group (80 +) is forecast to increase by eight million between 2010 and 2030—a 44% growth. The majority of these very old people will be women, many of whom will be living in poverty and requiring assistance and care, both formal and informal (CEC 2003d: 14). This indicates the urgency of tackling poverty and exclusion at an early stage in life!

The sixth trend concerns demographic ageing, which coincides with extended life expectancy beyond retirement, declining fertility and early exit from the labour market. The combined effect of these trends is to dramatically increase population dependency ratios in the forthcoming decades, requiring major adjustment of both the labour market and the welfare systems. To give an idea of the magnitude of this trend—the number of active persons per person aged over 65 has been halved since 1950, representing 4 persons in 2000 and just about 3 by 2020, while persons aged 60 + will increase from 20% now to 33% by 2050 (UN 2001). This situation is aggravated by the decline in the number of people apt to work and by unemployment, while the expansion of non-standard work reduces the contribution capacity of the population and the extent of coverage. The need for adaptation is made all the more pressing by the growing demands on public expenditures from the rising costs of health care and pensions, which taken together already represent some 80% of social expenditure, at a time when governments’ room for manoeuvre is limited by budget constraints. These constraints are there to stay, as they result from a combination of factors such as the relatively high level of current public deficits and debt in a number of countries, the generally limited scope for raising taxes in the current economic slowdown, exacerbated international competition in which the risk of relocation and “offshoring” to avoid heavy taxation is high, and, last but not least, the prevalence of neo-liberal orthodoxy, which promotes a lean or “small” State and favours tax cuts to stimulate investment or demand.

Turning to migration would only marginally alleviate the problem of the shrinking workforce and the sheer numbers required would not be acceptable politically, economically and socially either in the host countries—as shown by the rising xenophobia in Europe and the worldwide concern for security—or in the “emerging” economies, where such massive exits would create a brain-drain putting at risk their own development.

The main answer therefore seems to reside in promoting the activity and employment rates of the working age population (Sigg 2002). So we need now to turn to the attempts to tackle this problem.

How Have Welfare Systems Adapted to These Changes?

In response in particular to growing and persistent unemployment and to early exit from the labour market, welfare systems have become much more employment-oriented, focusing on activation policies, without losing their original emphasis on income replacement. Activation now extends beyond able bodied individuals of working age to include groups previously considered as “legitimately non working”, particularly mothers and single mothers, older people and people with disabilities (Bonoli and Sarfati 2002; OECD 2003b).

Arguably, activation measures were successful in reducing the levels and duration of unemployment, but also experienced some shortcomings, notably by selecting the “fittest” and excluding from the labour market those who are more difficult to train or to place, or those who face “exogenous” obstacles to entering the labour market, notably women with caring responsibilities, particularly when they are single parents. Moreover, even in the most successful instances, these policies have contributed to maintaining gender segregation, even in the more egalitarian Nordic countries. To illustrate these policies and their outcomes, two of the many examples contained in our ISSA project will be mentioned here, notably Denmark and the UK, which show different approaches to activation.

The greatest success among activation initiatives is certainly Denmark’s, where these policies date back to the 70s. The extent of their performance can be gauged by the dramatic contraction of unemployment from a 12.4% peak in 1993 to 4.3% by 2001, the lowest since 1976, and an impressive increase in employment. Total activity and employment rates in 2001 were among the highest in the OECD region and substantially higher than the EU figures (respectively, 79.9% and 76.2%, vs. 69.7% and 64.1% for the EU average). These high figures also apply to female employment rates (72% vs. 55% EU average!). Another important achievement is the very low long term unemployment (LTU) rate of 0.9%, representing less than a quarter of total unemployment compared to 3.3% LTU, which is more than 40% of total unemployment in the EU. The youth unemployment rate was 8.5% in 2001, compared to 14.6% in the EU. While total unemployment generally increased in the EU in 2002, Danish youth unemployment slightly declined, while it somewhat increased in the EU. As regards the employment rate of older people aged 55 +, though relatively low, it has increased by almost 10 percentage points since 1996 and now stands at 58%, well above the European average (38.5%) and even above the EU Stockholm target of 50%. Danish women also fare much better in this age group, with their activity rates rising by one percentage point to 52.9% in 2002 vs. 32.5% in the EU (CEC 2003c: 209, 213).

Denmark is thus seen as providing a unique combination of stable economic growth and a solidaristic welfare system that protects its citizens from the consequences of rough structural change. Its “flexicurity” model, as it has come to be known, combines high levels of flexibility through job mobility and job rotation and a generous and long duration of unemployment benefit system accompanied by extensive activation programs. The latter are based on rights and duties and provide massive support in terms of qualifying training, individual monitoring, placement and subsidized jobs to the unemployed. This approach has the advantage of improving both the labour force’s employability and the ongoing transformation of the economy. Its strength resides in the achievement of positive individual and collective outcomes as a result of broad-based negotiations and strong involvement of social partners—employers and trade unions—and local authorities. It shows that labour market policies cannot generate ordinary jobs by themselves, and that demand driven growth is a prerequisite, which Denmark achieved by also expanding fiscal policy in the early 90s and by reforming the credit system (Madsen 2002). It is also remarkable that these positive outcomes were reached without wage inflation and with little change in income distribution. Moreover, they not only have not incurred balance of payments deficits (except for 1998), but also were rather producing rising public budget surpluses! (Madsen 2004: 6) The incidence of poverty among the unemployed is extremely low and incommensurably lower than in the rest of the EU (Andersen 2002: 81), while the incidence of child poverty of 4% is the lowest in the EU, where it averages 17%, that reaching 20% or more in Germany, Italy, Spain, Portugal, Ireland and the UK. The latter has the highest level, in spite of successful government programs to combat it and the rate still stands at 25% (see below) (CERC 2004: 118).

Despite this highly performing labour market, the Danish activation policies have some shortcomings, particularly, as noted by Madsen, the ongoing process of selection which implies a continuous testing of employees’ productivity, the “creaming” of the more qualified, while gradually excluding from the labour market those who fail to meet the productivity criteria. The proportion of people excluded from the labour market and living on transfer income has almost trebled from 6% in 1960, to 15% in the late 70s, growing to some 25 % at the end of the 90s, with ethnic minorities being over-represented among cash benefits recipients and unskilled workers among voluntary early retirees. So there are, as Madsen puts it, some snakes in the Danish paradise (Madsen 2002). But this trend of an increasing proportion of people living on transfer payments is visible in other Nordic countries, and beyond, with the rather high incidence of people living on disability benefits, about the double or more as those depending on unemployment compensation. On average, 14% of the working age population in OECD countries classify themselves as disabled, with a third being severely disabled (OECD 2003b).

Madsen also acknowledges that the past performance was facilitated by economic expansion. He cautions against the risks of a downturn in the business cycle, which could raise the cost of activating the unemployed at an early stage, requiring sharp increases in public expenditure at a time of falling revenues and growing political pressures for cuts in activation budgets. This problem is compounded by the need to comply de facto with the strict budget criteria of the EU Stability and Growth Pact (to which Denmark has not adhered, but which must be heeded for competitive reasons) (Madsen 2004).

We will turn now to the UK, where activation policies were launched in the late 90s in order to make work possible for people who can work, providing security to those who cannot, and “making work pay”. These policies consist of the following three main clusters of measures (Millar 2002):

The first aims at helping parents to combine work and family life. It includes, besides improvements in maternity leave, which now statutorily covers all employed women, extended eligibility to maternity allowances, new rights for part-time workers (who represent one fifth of the workforce)—mainly women—as regards hourly pay rates equal to full-timers and the same access to company pension schemes, sick pay, training, holidays and other leave. In addition, and more importantly, was the adoption of the National Childcare Strategy, which provided for a public budget allocation for the provision of quality and affordable childcare for all children aged 0–14, including after school hours and during holidays. This is a clear departure from past policies where childcare was outside the remit of government responsibility. Local authorities are entrusted with its implementation, in partnership with private and voluntary groups.

The second set of initiatives focuses on the poorest areas and provides resources to deprived neighbourhoods so as to reduce long term unemployment and “worklessness”, improve health, reduce crime and raise educational standards.

A recent submission by the Trades Union Congress (TUC) on the government’s review of policies to eliminate child poverty notes that some success has been achieved in reducing the proportion of working-age people living in workless households from 14.1% in 1995 to 11.5% in 2003. However, when looking at the family type, the picture is gloomier. Indeed, while the proportion of workless families grew substantially in the past two decades for all family types, the worst hit were lone parents with children, among whom worklessness rose from 35% in 1979 to 56% in 1999. An important concern here is the high incidence of poverty in workless households, which is five times as high as that of families with someone in employment (65% vs. 12% after housing costs are included). A closely related preoccupation is the extent of child poverty, which has dramatically increased since 1979, peaking at 33% of all children in 1998/9, and receding to 31% in 2000/01. It still represents 3.8 million children and the government’s declared objective is to halve it by 2010 and to eradicate it by 2020 (TUC 2003a: 1, 3, 5).

Among the third group of activation measures, known as the New Deals, probably the most important, according to Millar (2002), aim at helping various population groups into work—young people, disabled, lone parents, older workers.

The innovative approach of the UK activation strategy is the focus on “worklessness” (or the absence of paid employment), rather than unemployment, because of its close link with poverty and receipt of benefit. Indeed, about half of the 5 million people who receive state benefits are sick or disabled, one-in-five are lone parents and the remainder are unemployed. Moreover, a third of all children live in poverty, of whom just over half live in workless households.

This policy has succeeded in involving many more women who were previously excluded from activation. The extent of the success of these targeted measures can be gauged from recent data that indicate a 10% decline between 1996 and 2000 in the numbers of workless households and a 20% decline in the number of children in workless households (Cooper-Green et al. 2001). The latest official data on the New Deal for Lone Parents shows that over half of the program entrants since October 1998 moved into jobs by end September 2003 (e.g. of 540,630 starts, there were 254,230 job entries), more than half of these were women (51.7%) (TUC 2004a).

But, while recognizing the importance of opening up opportunities for people previously excluded from support, Millar (2002) highlights potential problem areas:

  • First, the contentious role and extent in such activation measures of compulsion to work. Should women with small children be expected to work full-time and care for their children? The focus on paid work may adversely affect parental care and may discourage other forms of work or socially useful activity, notably voluntary and community work.

  • Second, the policy focus on workless households rather than workless individuals means that the main effort is to get one adult into work, while it would be appropriate to encourage two-earner families.

  • Third, the UK government is not in favour of direct job creation, despite the fact that publicly supported or semi-supported employment can help the most disadvantaged groups back to work and would thus deserve attention.

  • Fourth, there is need to define and implement rights and obligations of employers in respect to training.

  • Fifth, the current debate is still open on where activation focus should be. Is it on getting people into jobs, leaving the market to do the rest? Or should the focus be on developing human capital to improve the skills and hence retention and career development prospects of people (that is, their “employability”).

  • Lastly, the new approach to activation requires institutional and cultural change—with close cooperation and co-ordination between the different parts of the administration. Some agencies deal with service provision, others with exercising control and sanctions. Coordination is also necessary between local authorities and employers, as well as with voluntary organizations. Lastly, there is a broad range of new skills and knowledge that personal advisers have to acquire for the implementation of the measures.

Arguably, while activation based on the principle of “making work pay” has scored success in labour market entry or re-entry, in the past decade, it has also generated a new phenomenon of the “working poor”, many of whom are women. In response to this problem, new forms of ex-post social intervention had to be devised in order to keep such workers on the job rather than returning to welfare, such as the negative income tax and child credits in the UK, the US and France, for example (Bellorgey 2002; Millar 2002; Sinfield 2002).

More generally, while activation measures across the EU succeeded in consolidating labour market attachment and in reducing poverty, their main drawback is that they by-pass the non working population. This concerns the significant numbers of unprotected unemployed, who represent between a quarter and a third of the unemployed in the EU, and those who face non financial obstacles to returning to work, such as the unavailability of affordable and quality child-care facilities or transportation. In such circumstances, stronger financial incentives based on benefit reduction may have little impact on decisions to take up paid work. In any case, a strategy aiming at employment promotion by cutting the unemployment benefit can only have a limited outcome while reducing people’s capacity to acquire skills and search for a job (Bonoli and Sarfati 2002). The Danish example of generous and long-term benefits over several years, combined with activation, job rotation opportunities and sabbatical leave, childcare arrangements publicly subsidized and parental leave seem to be a more effective answer.

So, clearly, governments now dispose of a broad set of activation policies, ranging from workfare programs in the US and New Deals in the UK to the Nordic active labour market policies and other variations across the OECD region. But these policies leave too many people on the side of the road, marginalized either in dead-end jobs or in low income statuses as unemployed, handicapped or early retired and this may be more severely felt in a long life society where care of dependants becomes more demanding for the welfare system as much as for the active population, particularly for women who tend to be the main “informal” care providers.

The OECD secretariat, which promoted active labour market policies (ALMP) as the core of its Jobs Strategy, launched in 1994, has now come to recognize its limitations, despite its noted successes, in particular, its failure to address long-term unemployment. As noted by Martin (2000: 88) “the track record of many active programs is patchy in terms of achieving their stated objectives.”

Commenting on the complexity of existing ALMP evaluations in Europe, Gazier (2000: 11) points out that the evidence from the few comparative international studies shows positive and limited effect for such policies. Depending on the institutional set-up and national context, most of these policies are weakly efficient, and sometimes . . . can be inefficient. However, they are worth the while in spite of their possible perverse effects (e.g. “creaming” effect, windfall effect, displacement and substitution effect, crowding out effect, revolving door effect). Moreover, there is a strong complementarity between “active” and “passive” labour market policies. The EU “Employment Strategy”, launched in 1997, and its related and recently updated guidelines (CEC 2003b) rely strongly on early activation, and set quantitative objectives to be achieved in terms of timing and of percentage of trainees among the unemployed, as well as for employment rates of youth, women and older persons (the Lisbon and Stockholm targets). It also includes among its four pillars, the promotion of equal opportunities in access to training and in employment.

Another feature of recent welfare reforms is the greater attention granted to women’s need for an independent income, through recognition of their entitlements for coverage during periods of child rearing, and sometimes also for care for other dependant family members, particularly with regard to calculating their pension rights (Ginn 2003).

The development of parental leave schemes and child credits has also contributed to providing income guarantees to mothers and to a lesser degree, to fathers. But the persistence of non-individualized social rights, combined with gender segregation, persistent pay inequalities and inconvenient working time arrangements make it difficult to reconcile work and family responsibilities and may encourage this group to leave the labour market or never enter it.

Derived social rights related to family status also result in unequal treatment with regard to the level of social benefits. These reflect persistent inequalities on the labour market and asymmetrical roles within the family. They are amplified by social security systems. Thus, for example, people receive different benefit levels depending on their family status. To address this problem, Meulders and Jepsen (2002) recommend that family policies and related social protection reforms should be scrutinized from the vantage point of “gender impact assessment” to make sure that there is a process of gender mainstreaming in all planned measures. For example, promoting part-time employment is not gender neutral, since it would be mostly women who would take it up, with the result that it has a long-term effect on wages, promotion and social security rights. The same argument also applies to career breaks. Arguably, these are rather complex and ambiguous matters that are widely discussed in the literature. However, such scrutiny is part of the European Employment Strategy, namely on the potential impact of policy measures on labour market entry and withdrawal, as well as on their potential discriminatory outcome in terms of “mainstreaming” of women’s work (Jepsen and Meulders 2002). Women’s labour market participation and work-family balance are issues that concern society as a whole, whether it is for achieving social cohesion and inclusion, social equity, or . . . addressing the demographic problem of the welfare state. Thus, for example, contrary to expectations, an EU opinion survey shows that men complain more than women on stress resulting from long or inappropriate work schedules and the difficulty they have to reconcile work with other pursuits, including care for children. The authors therefore underline the need to consider work schedules and other work arrangements for the whole workforce and not only for women (European Foundation 2002).

As already pointed out earlier, while women labour force participation and employment rates have significantly risen in the advanced economies, including those of women with small children, non-employment of women due to family responsibilities remains rather high in a number of EU countries for both prime-aged and older women (25–54 and 55–64), standing respectively at 18% and 21% (Eurostat 2003b). As women’s educational attainment and job experience at a younger age have increased during the past two decades, it is not surprising to note that about two thirds of non-active prime-aged women, but only half of older women, have previous work experience. Most of the prime-aged non-employed women belong to two-parent households with children, while some are heading single-parent households. These characteristics point to the kind of policies and incentives that may be necessary to attract women with family responsibilities back to the labour market, notably through better access to continuous training and skill-upgrading, assistance in child care and other care, family-friendly work scheduling, career advancement prospects, possibilities to alternate non-standard with regular full-time jobs, as well as a taxation system better adapted to two-earner families.

What Can Be Done to Enhance Female Activity Rates?

A long life society implies an emphasis on the means to achieve an inclusive society and a balanced social protection system. Much has been said and written about the demographic ageing “time bomb” and the need to reverse the trend of early retirement by extending working life by several years, and the related social security contributions to counter the increasing dependency ratios of the population, particularly in Europe. Much has also been said about the rigidity of employment protection legislation and of minimum wages, which have been considered by some as the main obstacle to the promotion of job creation in Europe. But, the key issue in this equation, as already stated, is an efficiently functioning labour market, a concept that is too often overlooked or underestimated. This type of market is characterized by high participation rates of the different components of the working age population, two-way mobility between activity and inactivity, between different employment statuses (part-time or temporary and full-time work), between paid work and training or socially useful activities, and between work and (gradual) retirement. Such a labour market may be more important than the age structure of the population in guaranteeing the long-term sustainability of today’s welfare states, and in ensuring decent living standards and social cohesion for the population.

However, as noted earlier, despite the general progress in women’s participation and employment rates, inactivity rates (other than unemployment) among women constitute two-thirds of all non-employment of the working age population (OECD 2003a: 81–82). The policy concern for inactivity stems from its close correlation with persistent poverty and exclusion. It affects, in particular, lone parents, especially mothers, and their children. Indeed, children growing up in poor and workless households are more likely to drop out of school and have fewer chances of finding a regular job with decent pay, and this fact leads to a higher incidence of crime and drug addiction. The chances for non-employed to enter or return to the labour market also depend on the “closeness” of their labour market status—those who were previously employed (even for a short time), or who are in training or education, and even those who are unemployed, are more likely to find a job than those who have had no earlier work experience, those qualified as “homemakers” and the retired, as shown by Table 1, which indicates large differences in the degree of mobility into employment according to the extent of previous relative “distance” from the labour market. An overwhelming majority of inactive persons, both men and women, aged 15–54 that are either middle-skilled or high skilled would like to work, as shown in Table 2. Understandably, therefore, policy focus in many countries is not only on improving incomes for such families through transfer payments, but also on identifying the barriers that prevent the non-employed from entering or re-entering the labour market and improving their earning capacity.

Arguably, cultural, social, economic and institutional factors influence women’s activity rates, as do the quantity and quality of jobs offered. This is reflected in the very different labour market outcomes across countries, which are also influenced by government policies that may stimulate or deter supply and demand for female work. Major determinants include available and affordable childcare and elderly care, access to education and training, decent levels of remuneration, flexibility and choice in work schedules, availability of part-time work, adequate welfare coverage particularly pensions, as well as equality of opportunity at work. Government policies play a role in these areas as much as labour-management practices and collective bargaining do.

To enhance labour supply, assistance in childcare and care provided to dependant family members is crucial, as is clearly evidenced by the Nordic countries (Andersen 2002; Overbye 1998; Esping-Andersen 2000). As already mentioned, activity rates of prime-aged women are very high in these countries, around or above 80%—that is even higher than in the US and other Anglo-Saxon countries often mentioned as good performers in this respect—, whereas the Southern European countries have very low female activity rates at or much below 50%. Several closely related factors explain this shortfall.

The inadequacy and the high cost of childcare services are the main obstacle to female employment. Esping-Andersen (2000) considers this factor as much more central to the lacklustre employment situation in Europe than the oft referred to labour market rigidities. He considers that it discourages employment of women to a much greater extent than either the employment protection legislation or high minimum wages which, according to a number of liberal economists, are the main features of these labour market rigidities (Esping-Andersen 2000: 104).

Table 1

Big Differences in How Easily Non-employed Groups Move into Employment : Movements between Main Activity Statuses for Working-age Persons in Europe, 1997-1998 : One-year Transition Probability (percentages)a

Big Differences in How Easily Non-employed Groups Move into Employment : Movements between Main Activity Statuses for Working-age Persons in Europe, 1997-1998 : One-year Transition Probability (percentages)a

Population-weighted averages for 12 European countries (Austria and the 11 EU countries reported in Chart 2.11).

Source: OECD Employment Outlook 2003, Table 2.2.

-> Voir la liste des tableaux

Table 2

Inactive Persons of Working Age Who would Like to Work (now or at some time in the future), 1997

Inactive Persons of Working Age Who would Like to Work (now or at some time in the future), 1997

.. Data not available.

– Values not reported because of the small number of observations.

a) Question V66: “Would you like to have a paid job, now or in the future?” b) Includes job displacement, dismissal and end of job contract. c) Population-weighted average for countries shown. d) Population-weighted average calculated for countries with data for both surveys: Denmark, France, Germany, Hungary, Italy, Netherlands, Poland, Portugal, Spain, Sweden, Switzerland and UK. e) Among non-active persons who are not seeking employment, share of those who would nevertheless like to have work (now), data for 2001.

Source: OECD Employment Outlook 2003, Table 2.A.1.1; The International Social Survey Programme (ISSP), 1997; European Union Labour Force Survey (EULFS), 2001.

-> Voir la liste des tableaux

The availability of paid parental leave also plays an important role in enabling women to maintain their occupational status and progress in their careers. Thus, for example, women in the Nordic countries have access not only to high quality subsidized childcare facilities, but also to good quality education and training. As a result, they actually tend to leave part-time jobs and take up full-time employment (CEC 2003c: statistical annex).

Further evidence to the relevance of affordable child care for reducing exclusion can be drawn from the already mentioned high incidence of poverty and welfare dependency and non-employment among single-parent households (the rate of their dependency on social transfers is 8% across the OECD region, but over 12% in Austria, Belgium and the UK; OECD 1998). The level of provision and child care costs, the age at which children start school and the availability of pre-school care influence the outcomes of programs to make work pay (OECD 2003a: 117; OECD 2002b, 2003c).

The UK experience shows the difficulties in tackling high inactivity rates and poverty. A recent report by the Trades Union Congress (TUC) on inactivity thus points out that although the country has one of the best records in the EU on unemployment, it also has the worst record in terms of labour market exclusion, measured by the share of inactive working age people who say they want to work (22% compared to 9.9% EU average in 2002), and the reasons for this inactivity are mainly bad health, and family and caring responsibilities, both of which have a much higher incidence in the UK than in other EU countries. While family responsibilities explain 26% of all inactivity in the UK, the figure rises to 51% for lone parents, compared to 13% in Sweden and 23% in Denmark. The report assumes a close linkage between caring responsibilities, poverty and low health status (TUC 2004b). Another TUC paper points out that in Spring 2003, there were 2,141,000 economically inactive people of working age in the UK who wanted jobs—that is significantly more than the 1.5 million unemployed. Among these, the long term sick and disabled were the single largest group (35.1%), followed by persons looking after a family or home (30.3%). Government policy to promote employment rates for these groups seemed to meet some success, notably for lone parents, whose employment rates rose from 43.5% in 1997 to 53.1% in 2003 (TUC 2004c: 2,3). At the same time, in spite of huge government effort and investment, there are still not enough childcare places and their high cost (nearly a quarter of average household) prevents mothers in low-income families from returning to work (TUC 2004d; The Guardian, 26 Jan. 2004).

The incidence of lone parenthood, poverty and the increased risk of labour market exclusion is also noted in the US, where the recession of the past two years and the more recent increase in joblessness particularly affects the family income of unemployed, low-skilled single mothers, whose numbers grew by 19.2% between 2001 and 2002, while job openings in retail trade and services, where they were most likely to be employed, declined by 22.5% and 16.1% respectively (EPI 2003a).

Levels of education and skills obviously also play an important role in labour market attachment. Lack of qualifications limits employment rates. Employment rates are generally strongly and positively correlated to higher educational attainment, and the gap in employment rates of groups with different educational levels is substantial. This applies in particular more to women than to men (but also to older workers). Thus, in 1996, close to two-thirds of women who were not in the labour force had no educational qualifications beyond basic schooling, compared to 35% of those who held jobs. By 2001, employment rates of low-skilled women remained at the significantly low level of 37%. The latest finding of the EU show that although the share of low skilled in the working age population has sharply decreased since the mid-90s, it still amounts to 40% overall, and concerns a fifth of the 25–30 age cohort for both men and women (Sarfati 2002; CEC 2000, 2002). Educational attainment and employment rates are closely linked and, in most EU countries, explain high employment rates for women with tertiary education as compared to those with lower level education, and a much lower gender gap in employment as illustrated by Table 3. A major effort thus needs to be made in investing and improving human capital, which can contribute to better pay and career development and enhance productivity. In spite of these obvious arguments, as noted by Gazier, the least efficient activation measures are those providing training to the unemployed and job subsidies. The most successful activation measures are placement, personal monitoring and highly focused professional training (Madsen 2002; Pennings 2002; Gazier 2003). So there needs to be more rethinking of better adapted education and training systems, both at the initial and later stages.

Also, with the shift towards the post-industrial society, education plays an important role. Indeed, the higher the level of educational attainment, the higher the likelihood to be employed in the services sector, particularly for women at all skill levels (the reverse is true for employment share of both men and women in agriculture and manufacturing!). Arguably the service sector offers both highly skilled and low-skilled jobs, but, as the OECD secretariat points out, once “trapped” in low-skilled and low-paying jobs, there is very little chance for upward mobility, particularly for women, who are more likely to move from such jobs into inactivity (OECD 2003a: 95). Across the OECD region (with the exception of Japan and South Korea), employment rates are much higher, and the gender gap much lower among women with a tertiary education than among low educated women (OECD 2002a: 71).

Tableau 3

Women’s Employment Rates and the Gender Employment Gap by Educational Attainment, 2000 Persons aged 25 to 54 years

Women’s Employment Rates and the Gender Employment Gap by Educational Attainment, 2000 Persons aged 25 to 54 years

a) Percentage point difference between the employment rates for men and for women.

b) For above countries only.

Source: OECD Employment Outlook 2002, Table 2.2.

-> Voir la liste des tableaux

On the other hand, many qualified women have to accept jobs below the level of their educational attainment; they often suffer wage inequalities and have to cope with the “glass ceiling” when they reach the top jobs, particularly in the private sector. Although the wage gap between men and women has tended to shrink, it is still important throughout the industrialized countries (about 25 to 30 percentage points) in both low-skilled and high-skilled jobs, despite the adoption of non-discrimination legislation. It is often due to different work statuses associated with part-time and discontinued careers. These factors constitute disincentives to job entry.

The emphasis on qualifications and tertiary education, while it can open up routes to better paid and more interesting jobs, this may not be the only response to better jobs. Arguably, future employment growth will no doubt be concentrated in the service providing sector—notably health, caring and personal services, especially in the context of demographic ageing—, but also teaching, professional and business services, all of which are major employers of women. However, the latest forecasts from the US Labour Department estimate that, in the US, most of these jobs will be low or little skilled (BLS 2004). Moreover, by contrast with the experience of past decades, recent trends in US unemployment (2000–2003) show a negative correlation between higher educational attainment and exposure to long term unemployment, which was twice as high for people with a university degree compared to people with high school degree or less . . . (though there was no breakdown by gender) (EPI 2004).

On the demand side, the availability of part-time jobs is, of course, important for women wishing to enter or return to the labour market. This depends much on the extent of the development of the services sector—both public and private—, which is highly positively correlated with female employment as much as with the provision of services on which they depend and which are not limited to child care (the supply side) (Sarfati 2002; Cressey 2002; Ughetto 2002). Esping-Andersen argues that restrictive macroeconomic policies reduce real private consumption expenditures and hence demand for services; they thus depress job growth potential in the services sector where women are predominantly employed. He therefore sees the solution to EU unemployment in a reduction of the obstacles to labour market entry of housewives (Esping-Andersen 2000: 102, 103, 107–108).

Evidence for this argument can be seen in the latest EU data which shows that the dynamism of job creation in the services sector was most pronounced in services where demand tends to grow with income—notably in education and health services, home health care services, social and personal services and recreation. These obviously are linked to the increase in female employment and to . . . job creation especially in the public services. However, currently, there is a rather limited scope for developing public services due to the budget constraints already mentioned, although the perspective of the massive withdrawal of the “baby-boomers” will shortly offer many job openings.

But innovative approaches to working life flexibility go beyond the issue of women’s ability to take on part-time work. They depend on corporate concerns with work organization, working patterns and location, transportation, opening hours of public services and shops, the provision of caring and other community services. These must translate into a societal reality and political commitment to equal opportunities along with offering more job opportunities. This has been demonstrated in a pioneering way in Northern Italy and emulated in the UK, France and Germany (Cressey 2002; Boulin 2003).

The quality of the job offered is also important for providing decent living standards at work and in retirement. Part-time jobs are not necessarily precarious in terms of tenure. Recent findings for Europe actually indicate that women working part-time enjoy more stable jobs—with 10 years’ tenure or more—, than women in full-time jobs (Doogan 2001). In the US, in the period 1999–2001, nearly all non-standard workers saw a considerable real wage growth as a result of the combined effect of the decline in the share of workers in non-standard employment, low unemployment rates and strong labour demand. The only exceptions were for women in on-call or day labour jobs and contract company jobs, and male self-employed independent contractors. The gap in hourly pay between women on regular part-time compared to full-time worker was less than 15%, and less than 11% for women working for a temporary work agency (EPI 2003c: 3, 5). However, as already noted, many part-timers work for short hours for lack of better alternatives, and are therefore excluded from employment and social protection coverage.

While good quality jobs include opportunities for skill upgrading and career development and hence provide more chances to remain in employment, people in low quality jobs face greater risks of unemployment and exclusion from the labour market. I have already discussed the higher exposure of women to such risks.

The problem of discontinued careers and employment in “non-standard jobs” thus emerges once again during labour market withdrawal and more seriously, during retirement, when income replacement rates are low as a result not only of inadequate contributions, but also of these workers’ decreased capacity to participate in occupational pension schemes which tend to complement, sometimes significantly, income from public pay-as-you go schemes (first pillar). This exclusion of women equates with poverty in countries where public schemes are low-level flat rate, as is the case in the UK or Switzerland, for example (Ginn 2002).

In the UK, low paid and undervalued part-time women workers suffer from financial hardship throughout their working lives and are not eligible for, or cannot afford to contribute to, their own occupational pensions. Women part-timers earn less than 60% of the hourly rate of full-time male workers. According to the UK Equal Opportunities Commission, only 37% of women working part-time have access to any kind of pension scheme, and 44% of women employees work part-time! Even where access to an occupational pension scheme is possible, women are not likely to join if there is no employer contribution, and many women cannot build up entitlement to an occupational pension because of discontinued work history with unpaid caring responsibilities. Currently, 51% of women do not receive a basic state pension in their own right (see in the preceding section the comment on derived social rights) and they are penalized by the sex-based annuity rates of the pensions industry. Some ethnic groups are particularly adversely affected. Thus, for example, only 3% of Bangladeshi women now have an occupational pension. The TUC therefore underlines the need to undertake a reform in eligibility rules, including restoring the link between pension entitlement and average earnings, increasing substantially the basic state pension, introducing compulsory employer contributions to pension schemes and prohibiting sex-based annuity rates. Other related measures to reduce poverty among women include raising national minimum wage, ending the gender pay gap for both full-timers and part-timers within ten years, provision of more affordable child-care and a higher childcare tax credit, paid parental and caregiver’s leave and more earnings-related maternity pay (TUC 2003b: 2, 4).

In Switzerland, where the low-paid—particularly women in retailing—earn too little to qualify for occupational pensions, legislation was introduced at the end of 2003 to lower the earnings threshold for access to occupational pensions. In the Canton of Geneva, a special arrangement was introduced as of 1 January 2004, which enables households employing domestic workers to pay contributions (“chèque-service”) to ensure coverage by the welfare system (covering unemployment, pensions, invalidity, maternity and accident insurance; Le Courrier, 3 Dec. 2003).

So, in spite of the gradual adaptation of national pension systems across the EU to facilitate reconciliation between family and work responsibilities for both parents, the EU Commission acknowledges that the significant coverage discrepancy in pension entitlements, particularly under second pillar schemes of women atypical workers, will persist for a long time to come (CEC 2003a).

By Way of Conclusion

The current concern about the future of pensions and the resulting policy focus on reducing unemployment and promoting employment rates of the population have shown that the gender dimension is important in this context. It must be part and parcel of the broad range of current social and economic reforms—be it those targeting poverty, economic or cultural exclusion, achieving social equity and justice or providing incentives for labour market entry among people who are willing and able to work. The same applies to family policies, which need to facilitate balancing working lives and family responsibilities, stem the decline in fertility rates, and lead to reforms in taxation and social protection systems, which need to provide positive incentives to enter and stay at work.

Within the European Union, the current strategy to ensure the adequacy and sustainability of pensions in the context of demographic ageing emphasizes a three-pronged strategy of substantially raising the employment rates, reforming the pension and tax systems and reducing public debt. It recognizes the prime importance of increasing employment rates of women, as well as the postponement of the currently low effective retirement rates (CEC 2003a).

In this context, the development of the services sector is of prime importance as its job creation potential in general, and for women in particular, has been demonstrated. Virtually all net employment growth in OECD countries in recent years has taken place in the services sector. To recall again the Danish experience, employment in services in 2001 constituted 74.1% of total employment (62.9% for men, 86.5% for women), up from 69.7% in 1990. Similar results are found in other Nordic countries, but major gaps exist in other countries particularly in Southern Europe. The determinants of differing national performance include:

  • Slower productivity growth in some services which may contribute to job creation (for example in hotel and catering, personal and domestic services, retailing);

  • Higher productivity growth in business services, by contrast, can promote high quality employment;

  • The rise in incomes can produce shifts in demand for services (notably due to the continued growth of women’s labour market participation rates);

  • The expansion of the welfare state and related social and health services opens up many opportunities for female employment;

  • The level of income tax and consumption taxes can positively or negatively affect the demand for personal services; and

  • Cultural factors may affect service consumption patterns (Gadrey et al. 1999).

There is a strong positive correlation between female employment rates and the development of services. According to one estimate, closing the gap between the EU and the US in the level of employment in services would mean some extra 30 million jobs in Europe (Larsson 1999). That is more than twice the number of the currently unemployed!

Using activation and employment promotion measures to convert the “safety nets into a springboard”, to paraphrase the World Bank’s metaphor, may be appropriate for those who can benefit from it, but there are too many disadvantaged groups in the population who may find it difficult or even impossible to gain a foothold in the labour market. These include women who are disproportionately present among “atypical workers”, among the low-earners, among single-adult households with or without family responsibilities.

Labour market and welfare systems need to be reformed to take into account the changing characteristics of the labour force and life-cycle that no longer corresponds to the three-tiered sequence of education/training, followed by employment and culminating with retirement on which post-war welfare systems were based. Periods devoted to training and caring activities have become longer as those in employment shrank.

Employment promotion should therefore target a “socially sustainable employment”, which gives people greater choices in and outside work, in the family and in society, as well as the capability and flexibility to move between them during their life course. This can include consideration of working time no longer on a daily, weekly or monthly basis but throughout working life. This would help reconcile work and family responsibilities but also work and training or other non-paid activities, as well as early retirement through the institution of “time savings accounts” or “social drawing rights” (Boulin and Hoffmann 1999; Supiot 1999). Greater flexibility in working time envisaged over the life cycle does not necessarily result in additional cost for employers if it can be adapted to fluctuations in labour demand within the company. Thus, rather than a trade-off, there is a potential for a win-win outcome.

The “Transitional Labour Markets” approach, in line with such proposals, was recommended by Schmid and Gazier to facilitate transitions between different employment and social statuses. It aims in particular to provide greater autonomy and choice during different stages in life and relies on increased solidarity, which could be achieved by a policy-mix that addresses unemployment, ageing, gender discrimination and exclusion (Gazier and Schmid 2002; Gazier 2003). The Nordic countries have shown that such an approach is feasible and that there is no necessary incompatibility between high levels of both social protection and economic performance.

More generally, the changes in the labour market and the policy of increasing the population’s employment rates require a broader approach to the constitution of social protection rights beyond the strictly salaried “subordinated” employment relationship. There are already a number of examples in Europe where such a link is blurred, for example paid vocational training outside the workplace or time off for trade union or staff representation, remunerated by the employer (Freyssinet 2002). On the other hand, public pension schemes grant “pension credits” for periods not worked but which are acknowledged as serving a useful social purpose—such as education, unemployment, child rearing, care for the elderly, sickness and disability, and… military service. While the extent of coverage and areas giving rise to such entitlements for pension purposes vary widely among countries, the “caring credits” are particularly relevant for the gender dimension, being more family-friendly and contributing to reduce or avoid poverty among retired women. The problem arises in particular in countries where occupational pensions constitute an important proportion of the retiree’s income. And since there is increasing pressure worldwide to transfer more of the pension coverage to individuals and to the private sector, this aspect needs to be considered.

As Ginn points out “Comparison of EU countries shows that social and employment policies, exemplified in the Nordic welfare states, can do much to help reconcile women’s dual roles in reproducing society and in paid employment. If a trade-off between women’s social reproduction and economic production can be avoided, or at least minimized, the affordability of state pension schemes is improved by their social insurance contributions and taxes.” (Ginn 2003: 96). This also constitutes a win-win solution for society and the individual.

But reform initiatives encounter tough and very vocal opposition from many quarters, particularly from the core labour force with vested interests in the status quo, as well as from business circles who feel that the proposed reforms do not go far enough to liberalize the labour market and the welfare state. There therefore appears to be some urgency to raise awareness, particularly among existing institutions for social dialogue on the different societal issues and the socio-economic implications represented by different policy choices. There is need for a greater solidarity between genders, between different age cohorts and generations, and among labour market “insiders”, the more peripheral “non-standard workers” and the “outsiders”. The existence of representative social institutions which have a negotiating capacity at enterprise, local, sectoral and national levels is a prerequisite for achieving consensus on acceptable outcomes for all population groups, as has been amply demonstrated by the Nordic countries, in Austria and the Netherlands, but also more selectively on some specific reform issues in traditionally more adversarial industrial relations contexts such as those prevailing in Italy, Ireland, Spain or the UK (Sarfati 1999, 2003; Baccaro 2002; Cressey 2002; Lourdelle 2002). But it also requires government commitment to such dialogue and consensus building.

As the ISSA book shows, there is certainly more than one route to successful reforms of both the labour markets and the welfare state. And, while activation and employment are important means to achieving and preserving social cohesion, they are not an end in themselves.