Figure 1 Female participation rates and birth rates in the Netherlands,

Transcription

1 Pensions and female labour careers in The Netherlands Ton Janusch Ministry of Social Affairs and Employment Jan Dirk Vlasblom The Netherlands Institute for Social Research/SCP This assessment is based on a first analysis of the policies in discussion and is not to be regarded as the position of the Government of the Netherlands. 1. Female labour careers As pensions are mainly deferred payments, a large part of the pension gap between men and women will disappear as a result of a diminishing career gap. In the Netherlands, the last decades this career gap has become smaller. Since the 70s female participation has risen dramatically. It started off at the level of around 25% in the sixties, and it grew almost continuously over the next decades, with an increased growth rate since the mid-80s. At the same time, the birth rate of women, after a strong decline between the introduction of the contraceptive pill in 1965 and 1975, remained fairly constant during this period of increased participation. In 2005, the birth rate is slightly above the level of 1985, the year the increase in participation started, and since then increased to the present fertility rate of over 1.7. Declining birth rates as a result of increased female participation is, at least in the Netherlands, not a necessity. Figure 1 Female participation rates and birth rates in the Netherlands, Dotted line: female participation rate (left scale), continuous line: birth rate (right scale) These changes in labour force participation come more or less by generation as the next figure shows. Younger generations of women do show higher participation rates from the start of their career onwards. Also, the so-called child gap seems to be almost gone for the younger generations. This shows that part of the now-existing pension gap is caused by the women who were in their working age in the sixties. Younger cohorts did have a longer career, and will therefore have higher pension claims, as they had more years of participation. 1

2 Figure 2: Participation by age and cohort, all women Source: Schippers and Vlasblom (2004) It is difficult to link this gradual increase in participation rates to specific (labour market) policies. It presumably started by the increase in the level of education of women. Starting from the World War II, girls got better educated. The gap to men has now been closed in this respect. This more equal background in educational level, in combination with a strong feel for the need of equal treatment in the 70s, made that the attitudes to working women changed. This change started with the high educated women, but by and by, it has now spread across the whole Dutch society. Over the last decades, people let go their objections to working women, and working mothers in particular (Jongen, 2010, Portegijs et al. 2006). As the picture below shows, over the period the percentage of individuals who did not consider working mothers a problem increased. Interesting to see that the level of acceptance was (and presumably still is) lower for men than it is for women. Figure 3: Share of individuals (10-74) that agree with working mother not a problem Source for this figure: Jongen,

3 Together with this change in attitudes, the tax and social security system was changed to reflect this new situation. Over the period many household aspects of the income taxation were abolished. The rationale for this was to give women an incentive to have their own labour career. However, as the graph shows, these changes cannot be seen as the cause of the increase, as the increase started earlier. It can as well be argued that women who were economically independent, wanted to be treated as such by the institutions. At least, the change in the system did signal a change in the way society regarded working women: was the nonworking woman the default in the sixties, the default now is the working woman. At the same time, the tax and social security system still allows for the existence of a non-working partner, although remaining elements are at this moment gradually been phased-out. Together with this change in the organisation of the tax and social security system, the possibilities for women to work increased. Of prime importance is that the possibility of working part-time became available with most employers, including more responsible and managerial functions. In the Netherlands part-time work is not to be equalled with marginal jobs. Also, since the late 90s, there is an increase in the amount of childcare facilities. A system of childcare subsidies is in place, which evolved over a couple of years. This allows parents to get affordable childcare, in order to be able to combine work and care. To underline the abovementioned importance of autonomous developments, a comprehensive policy was only created with the Child Care Act of Together with the increase in female labour force participation, the existing male-female wage gap started to decline. On average, there is still a considerable gap in gross wages (Fransen et al. 2011). It is not uncommon to split this gap in an explained and an unexplained part. Such an analysis shows that part of the gap can be explained by aspects of sorting. However, a major part of the gap however, is due to differences in careers: neglecting labour market career during early years results in larger differences at higher ages (Fransen et al. 2011, Russo en Hassink 2008). The almost absent wage gap for the group (0% in the public sector, 8% in the profit sector) shows that the still dominant decision of women getting to work part-time and the man staying in his full-time job, is not the outcome of optimal household economics, but the result of preferences and routines. As far as there is a male-female gap in careers, it is mainly in the weekly number of hours worked. As the next figure shows, for all cohorts, the number of weekly working hours, after an initial start of around 36 hours per week, drops to an average of around 24 hours per week, and stay in a part-time job, also when the children no longer form care responsibilities. This pattern is rather consistent over cohorts. The only slight change is in the number of hours worked by the youngest cohort at the start of their career. The figure suggests that younger women tend to start, on average, in a small full-time job. As it comes to pensions, this gap has two effects: firstly, especially in the profit sector, wage profiles tend to correlate with the fact whether or not individuals work part-time (Russo en Hassink 2008). Secondly, as pensions are related to the total income earned in life, working less hours does also lower the build-up of pensions. 3

4 Figure 4: Weekly working hours per age cohort With respect to the gender gap in pensions, the last graph points out that care-credits during the first years of the child s life, do not solve the problem: especially in the hours worked, women tend to shift to part-time work at the birth a the first child (average working hours drop from 36 to 24 per week between the age of 25 and 35). However, at the time the children go to school, or even later on, when they leave home, there is no increase in the number of weekly working hours. The same applies, although to a lesser extent with respect to the participation rate: those women who stop working after the birth of a child tend to remain outside the labour market during long periods (some of them never return). Therefore, care credits over a small period of years, will not solve the pension gap, as many men keep working full time over this whole period. Again, this suggests that a policy that is aimed at gender equality in both labour and care careers does a better job in diminishing the care gap (and also in increasing female economic independency over the entire life course) As a result of the changed labour market pattern of Dutch women over the last three decades towards more participation and more gender equality, in the coming years the pension gap will be lower than it is now. The main cause for this is the trend towards more gender equality in labour careers. The introduction of a pension system that more or less pays a pension out of general means without a direct link to any working life career does give an opposite signal: in such a system, interruptions of the career are made cheaper to the individual, and thus more individuals will take the opportunity of such a career break. (Although it may be doubted to which respect individuals take pensions into account when making labour market and care decisions). In this way it gives the signal that gender inequality is rewarded by the system, and as such it gives a signal that is inconsistent with the goal of gender equality in the labour market. So, if care credits have to be incorporated into the pension system (which is as shown not the best way to deal with gender gaps in pensions), it should at least be done in such a way that it signals that within the household equal sharing of all tasks should be the default. Deviations should be possible, but should not come cheap (or even free) to the individual or to the household. One possible way of doing this is to let the working partner pay the pension premiums for the partner that takes a break: in this way the (future) costs to the interrupting partner are (at least) partly paid by the non-interrupting. This shift in costs between partners will induce both partners to rethink their division of tasks, it will give the right incentive within the household, and it will lessen the problem of free riding (taking more care benefits than the own pension contribution will even make for). 4

5 2. Care credits in the Dutch pension systems 2.1. State basic pension (AOW) It would be problematical to give a scheme of care compensation in the form of care credits or otherwise a logical place in the basic state pension (AOW). The AOW provides a flat-rate pension at the level of 50% of the net minimum wage for each partner in couples (EUR 748 per month gross), or 70% for singles (EUR 1,074 per month gross). The entitlement as such and the level of the AOW benefit are independent of the labour market and income position during the active phase of the life-course. A period of whole or partial retreat from the labour market regardless of the reason of that retreat does not influence the level of the benefit. The AOW ensures that a relatively high minimum income is provided for all persons aged 65 and over, and is in this way an effective prevention against poverty at old age, also for those who, to take on care responsibilities, have had a limited working career. From the perspective of the aim of the AOW to provide a sufficient income at old age there is no need to give special attention to the pension situation of carers. The AOW would lose its function as a base for supplementary pension arrangements if the benefit is dependent on the level of those supplementary pensions Occupational pensions A fundamental issue is that in the Netherlands the occupational pensions, as a part of the total labour agreement, are the responsibility of the social partners. Implementation of care supplements in the occupational pensions along the lines of statutory regulation would mean that the legislator would get a role in defining the pension agreements which exceeds its responsibilities. As the first pillar deals with poverty, the aim of the second pillar is to provide an adequate replacement income which enables retirees to continue the living standards during the active part of the life-course. This relation between pension and salary would be lost if for care periods the pension would be based on a higher than the actual salary. This would also imply a subsidy for carers, at the cost of other participants in the pension fund, which needs a sufficient legitimacy. Collective risk-sharing is one of the main characteristics of the second pillar. Labour market decisions in relation to care being to a high degree a personal choices rather than a necessity dictated by circumstances, the question arises whether such an internal subsidy is self-evident. This aspect supports the point this is a matter to be decided by the social partners rather than by the legislator. An implementation of care compensation in the supplementary pensions would for defined contribution (DC) systems imply the introduction of supplements on the contributions, and for defined benefit (DB) systems that of a surplus on the actual pay to which the pension entitlements are calculated. In the latter case, this would mean the introduction of a virtual pay for the determination of the individual entitlements and their costs for the pension fund, 5

6 alongside the actual pay as the basis for besides the regular use for salary payments, taxes and social security premiums the remittance of the pension contribution. The practical consequences for the administrations of employers and pension funds need further analysis as is the case for the costs, both initial and structural. In case of care for children, it seems logical to use the criteria which are used for the parental leave, i.e. the presence of a child up to a certain age (presently, in accordance with the Parental leave directive, 8 years in the Netherlands) with whom the employee has a formal relation as parent or guardian. While in the case of children a care situation can be assumed, if not for the employee then for the partner, that is not the case for other situations, e.g. for a parent in need of care. For those situations it would therefore be necessary to establish the need of such care and the actual giving of care by the employee. The first aspect presumes a formal procedure by which the care-need of the person in question can be determined. It seems highly problematical to find ways by the employer and/or pension fund can establish whether the employee is indeed giving this necessary care. Moreover, such systems should allow for the diverse ways this care is given, e.g. shared by a number of relatives. It is hard to see how such a system could be implemented without substantial administrative costs. Another issue would be if the care compensation is dependent on an actual loss of pay, in relation to the kind of entitlement, individual or family-based. An individual entitlement would be consistent with the Dutch system of parental leave (an individual, non-transferable entitlement) and the individual labour relation between employee and employer. Individual entitlements would, however, exclude the possibility of transfer of the compensation between the partners, which is relevant in cases where one of the partners decides for a retreat from the labour market. Shared entitlements on the other hand would necessitate a system in which employers and/or pension funds can establish if the employee is entitled for the compensation and, if partners would be allowed to split their common entitlement between each other, to which extent. Although the possibility of a transfer would in those cases prevent the household losing the entitlement, it does not contribute to lessening the pension gap between the caring and not-caring partners, and would even increase that gap. Other points which have to be further examined are the situation of single parents (who lack an employer when leaving the labour market), divorced parents (especially in the case of a family-based entitlement), and the employees for whom no occupational pension scheme exists Social context Schemes of care compensation presume life courses in which the regular labour market participation is temporarily stopped or diminished as the consequence of care responsibilities, to be resumed in its regular pattern after this period. In the Netherlands the working patterns before and during care periods are highly differentiated. A substantial segment of women start their working career part-time. This is also becoming more common with men. In the same way the working patterns of the partners during the care period are diverse. While the full-time/part-time combination is dominant, the 6

7 partner s job can vary between a marginal contract and an almost full-time occupation, while an increasing number of couples opt for both having a part-time job to share the caring. The working patterns before and during a care period are in this way highly variable, as is the resulting effects on the total and individual income situation. Any scheme of care compensation should take this variety into account, and have an outcome which makes sense in the individual situations. The same would be true for other care situations, e.g. for parents. Some carers are already working part-time, and can combine this care with the existing working pattern, while someone working full-time might be confronted with the necessity to reduce the working time, while other full-time workers can combine this with the care as a result of flexible working hours. Another relevant aspect is that, care responsibilities being a major reason for women to decide to go working part-time, in the Dutch situation the part-time working pattern does not end with the period of care responsibilities. Part-time working in most cases being a personal choice rather than a necessity dictated by circumstances, it would be experienced as a bonus for choices which otherwise also would have been made Consistency in policies The introduction of a scheme of care compensation would imply a fundamental change of the recent Dutch policies aimed at a higher female labour participation, both in persons and in hours. The most important of these are the gradual phasing out, also for families with younger children, of the tax benefit for single earners (around EUR 1,950 per year); and the introduction of a supplementary tax benefit for the least-earning partner in families with children, up to a maximum (of around EUR 1,850 per year) increasing with the income. These and other policies are based on the principle that the presence of children does not any more preclude the earning of an income, and are aimed at a greater labour market participation of women, from the perspectives of both financial independence of women and an increase in the labour supply. Schemes of care compensation on the other hand run the risk to stimulate a retreat from the labour market. Moreover, they are implicitly based on the assumption that care situations are not compatible with a working career. The introduction of a scheme of care compensation would in the Dutch situation lead to conflicting messages and contradictory incentives. 3. Conclusion Systems of care compensation in the pensions should form a careful balance between protection and incentives. Protection of a sufficient income at the old age is called for with a view on those situations in which labour market participation is for whatever reason no realistic option. At the same time, such systems should not form a disincentive for women to have a career and be economically independent. 7

8 The Dutch system of a state pension which, independent of the working career, guarantees protection against poverty, and supplementary pensions which reflect the working career, combines both elements. From the Dutch perspective a system of care credits or other policies by which carers are compensated for their loss of pension rights would imply a fundamental return in the present policies aimed at a more substantial labour market participation of women. It would be conflicting with the message to women to value a working career and an independent income, counteract the introduced financial incentives for women to stay working and, when possible, to increase their labour participation. While such schemes might lessen the pension gap, it will be at the cost of the career gap, which will result in the persistence of the main cause of the present pension gaps. Protection is called for if the present female life courses are inevitable. In the Dutch context, there is however no reason to assume a priori that the change in attitudes and preferences which has led to an almost total eradication of the pattern of women leaving the labour market as a result of care duties will not continue and the next period lead to more equal working careers. Such a development should not be hindered by contrary incentives. Especially the fact that this is an issue of deep-rooted patterns, consistent policies are needed if changes are to be achieved. For the same reason, there is a case for care compensation for those situations in which a continued labour market participation is discouraged, e.g. by a lack of social acceptance, or retraction from the labour market is stimulated, e.g. by paid leave arrangements. The design of such compensation schemes should not discourage a development to a future more substantial female labour market participation. Finally, an introduction of care credits in occupational pensions would in the Dutch situation raise questions which are not relevant in universally covering state pensions. The fundamental question touches the responsibility of the social partners for the occupational pensions. Which risks are to be covered in the pension schemes is, as an aspect of the total labour agreement, a matter to be decided by the social partners rather than the legislator. The fact that occupational pensions are dependent on the existence of a labour relation as such and, subsequently, with a specific employer and a specific pension fund, leads to not readily answerable questions as the way the care compensation should be organised for those deciding to leave the labour market, the financial consequences for individual employers and pension funds of the decisions partners make on the division or transfer of their entitlements, and the position of employees which are not covered by a pension fund and of the self-employed. 8

9 Appendix The old age pension system in the Netherlands Introduction The Dutch pension system has three main pillars: the state pension (AOW) forming the basis for supplementary occupational pension schemes, and individual supplements (e.g. personal savings, private insurance). This description will be limited to the first two pillars. First pillar: AOW The AOW has the following characteristics: It provides a flat-rate pension at the level of the net minimum wage (circa EUR 1,250 per month for a couple both partners of which are 65 or over). The coverage is universal, independent on the work status (employees, civil servants, selfemployed, or inactive). The entitlement is solely dependent on the years of residence in the Netherlands (complete insurance is reached by uninterrupted residence from the age of 15 years). The height of the income in this period is irrelevant for the level of the benefit. The benefit is paid on an individual basis. The benefit is 50% of the net minimum wage; for singles the benefit is 70%. From 2015 there is no additional benefit for partners younger than 65 years without a sufficient income. The AOW is funded by premiums and taxes. Second pillar: Occupational pensions Relevant characteristics of the occupational pensions are the following. The occupational pensions are, as a part of the total labour agreement, the responsibility of the social partners, who negotiate the pension agreements and govern the pension funds. The role of the legislator is mainly limited to aspects as governance, obligatory coverage for all employers and employees within the range of the fund, and financial security (with a supervisory role for the national bank DNB). Collective risk-sharing is one of the main characteristics of the second pillar. All the actuarial and investment risks are shared by the participants in the pension scheme, each paying a fixed percentage of her or his salary. The pension funds (around 550) are organised along the lines of economic sectors or individual companies. The pension funds cover around 95% of the employees. 9

10 The occupational pensions are capital funded, financed by earnings-related contributions of both employer and employees. A typical pension is, including the AOW, 70% of the average salary in the insured period. 10

11 References Fransen, E., J. Plantenga en J.D. Vlasblom (2011). Why do women still earn less than men? Decomposing the Dutch gender pay gap, In: Applied Economics, p Jongen, E.L.W. (2010). Childcare revisited. The Hague: CPB Netherlands Bureau for Economic Policy Analysis (CPB Document 200). Portegijs, W., M. Cloin, I. Ooms en E. Eggink (2006). Hoe het werkt met kinderen, Moeders over kinderopvang en werk. Den Haag: Sociaal en Cultureel Planbureau. Russo, G. en W. Hassink (2008). The Part-Time Wage Gap: a Career Perspective. In: De Economist, jg. 156, nr. 2, p Schippers, J.J. en J.D. Vlasblom (2004). Het einde van het kinderdal. In: Economisch Statistische Berichten, jg. 89, nr. 4448, p