MALE NORMS AND FEMALE ADJUSTMENTS: THE INFLUENCE OF CARE CREDITS ON GENDER PENSION GAPS IN FRANCE AND GERMANY Open Access

Recent reforms to the various different European pension systems have induced many substantial changes. While pension systems and their reforms imply several general challenges, women have to face additional structural disadvantages. This is the case because of the direct relationship of the pension norms to ‘ideal’ labour market biographies. Most European countries’ pension systems are based on compulsory insurance schemes that are related to work biographies.

In some countries, child care credits have been introduced and developed to adjust the original male breadwinner-oriented pension calculation. Of these countries we focus here on France and Germany in order to gain a better understanding of gender pension gaps in these kinds of systems and to have a closer look at the effects of adjustment measures.

The pension formulas in Germany and France incorporate two factors that depend on individual realization: (1) the duration and (2) the level of paid contributions. If different rights over pension resources are built up, it is these two factors that are responsible. On average, German and French women build up half the pension of men.

Based on the argument that this gender bias is caused by care responsibilities, in particular of mothers, both countries introduced measures to level out these disadvantages. No other reason is recognized as entitling one to non-wage related additional pensions. However, following the calculation logic above, such measures are insufficient for a substantial reduction of gender pension gaps. We argue that even though care-caused individual difficulties are more or less levelled off by care credits, the generally gendered structure of the opportunities to build up equivalent pension entitlements remain and even intensify over time. Balancing the expected losses of the last general pension reforms does not improve the situation for women, it only tends to prevent their situation from becoming even worse. In addition, these ‘repairing measures’ do not exclusively work out positively, as will be shown.

The aim of this paper is to contribute to a better understanding of child care pension credits, favoured by the EU (EC-Report 2003, 2006), that have been implemented in specific systems. It will show the extent to which the individual factors of pension calculations, that are also partly relevant within other European pension systems, are indeed individual and the extent to which gender pension gaps are reduced by the care credits in these two countries.

What changes in the different pension systems have in common is the aim of strengthening individualization, equalization, and, to some extent, privatization (National Action Plans 2002, 2005; EC-Report 2003, 2006). Europe's pension diversity is not only caused by the growing complexity of their institutions, measures and legislation, but also by different norms to legitimize and calculate pension entitlements. While some European countries include a non-means-tested basic pension, as for instance The Netherlands, other countries relate their systems solely to the labour market. France and Germany belong to the latter group.

Both countries’ pension systems have much in common, and over the past few years they have both been subjected to far-reaching reforms. On an individual level, their pension norms require life-long full-time labour market participation. Compared with this, both countries show gendered pension entitlements: on average, women's pensions are some 50% of those of men (for Germany – AVID 1996: 60f; for France – Veil 2002: 68). Different measures were introduced in both countries to redress some deficiency of women's contributions on an individual level to improve women's non-derived old age pensions of which the most important ones are pension care credits.

However, in both countries the pension norm itself, i.e., the assumptions regarding the life course institutionalized in the national pension formula for building up full pensions, was not changed. This norm was introduced in times of male-breadwinner labour market participation and subsidiarity-based family concepts. The calculation norm still requires life-long full-time employment, while the measures that were introduced, including care credits, are amendments in terms of adjustments to this norm. The situation of women is analysed here in terms of what can be expected for their entitlements taking the latest reforms into account.

Calculations are determined by labour market participation through wage-related contributions to compulsory insurance schemes; in the case of care credits, those contributions are fully or partly financed by the state. Full pension entitlements depend on two putatively individual calculation factors: (1) the duration of insurance, i.e., the length of time contributions have been made and (2) the level of income, or to be more precise, the level of paid contributions. This is the case for all pension schemes, explicitly for the public insurance schemes (Table 1).

The question is whether the recent reforms enable women in France and in Germany to build up full individual pension entitlements or whether there are still structural and general obstacles resulting in gendered old age pensions. In other words, do the adjustments to the traditional labour-market oriented male life-course norm equalize men's and women's opportunities to build up old age pensions? This article takes a closer look at both calculation factors, the duration and level of contributing, and the influence of the adjustments to answer this question.

Each country has its own life-course norm for reaching a specific level of old age security. The German and the French pension entitlements are built up through labour market participation. In the past few decades another factor for building up pension entitlements has found its way into both the French and the German systems: care credits. This means that there are two ways to reach the pension norm, the original one (which is also changing) and the amended one for mothers.¹ Two different forms of care credits have been introduced both in France and in Germany to improve mothers’ old age entitlements. One is a paid-up, i.e., non-contributory insurance period, entitling to pension benefits equated to paid contributions, for a period of two years (annuités), respectively, three years (Kindererziehungszeiten), and a level of up to entitlements gained by average income. The second one is, under certain circumstances, pension raising in different ways: in Germany periods of up to an additional seven years per child might be taken into account to raise pension entitlements of these years up to the maximum entitlements of average income-earners (Berücksichtigungszeiten), and in France it is possible to raise pensions for larger families (majoration). Since pension determining factors are interdependent and result in the later total pension, each division of pension entitlements is to some extent artificial. Nonetheless, the division in years of insurance and level of contributions is part of both calculation systems. They form the two individual parts of pension calculations as argued above.

Both systems’ norms for building up full pensions requires life-long labour market participation. In France, the norm means 40 years (160 trimesters) of pension insurance (as of 2003). The last reform in 2003 introduced the idea that in the long run the norm will require 42.5 years (170 trimesters; NAP [France] 2002, Appendix: 33). If this number of years is not complied with, the loss is non-proportional, i.e., for half the number of years insured, the pension level does not lie at 50% of a full pension but at 27%. It is the last 10 trimesters that raise the pension level (Observatoire des retraites 2002). Only within the civil service does half a career lead to half a pension.

In Germany, the pension norm, the notional Eckrentner, requires 45 years of pension insurance. These 45 years are attained by only 47% of men and 7.5% of women (Rürup 2003), showing that the assumptions are ideal, but neither are they realistic for men. In the future, this gender difference will change because of higher unemployment rates of men and child care credits for women. In the German pension calculation, not only the duration of insurance but also the age of retirement has very significant consequences due to the bonus-malus system. This system leads to pension cuts of 0.3% per month in cases of early retirement which is 18% for a period of five years. While the minimum retirement age is changing, and will soon be 67 for those born after 1969 (Rürup 2003), the consequences are actually worse for women due to the fact that their minimum retirement age was recently raised from 60 to 65, to bring it up to the same level as men.

Adjustments to the original norms in forms of introduced or developed care credits have improved mothers’ opportunities to build up non-derived pension entitlements. In France, introduced in 1971 and in force since 1975, two years of pension insurance are given per child (annuités pour enfant). Since the 2003 reform, fathers may also take these two years of pension credits for child rearing (Veil 2004). These two years were also given if the entitled parent did not interrupt work. Children, therefore, facilitated early retirement. This regulation was changed, so that as of 2004 care credits are linked to labour market interruptions of at least two months.

In Germany, child care credits, introduced in 1986, were increased from one to three years for children born after 1992. If mothers work during these years, any additional contributions are limited in level and are not added to the duration of insurance. As a result, working mothers’ twofold effort is not valued.

The level of these care credits also differs. While the three years of care credits in Germany comply with that of average income, i.e., 1.0 point (EP) per year for children born after 1992 (originally introduced in 1986 it was 0.75 point for one year), the French two years of care credits are linked to the individual income level. As of 2008, the best 25 years of pension insurance (increased from ten years) are the basis for calculation. Therefore, care credits in France work out best for mothers with a long-term high income.

Indeed, child care credits in principle help mothers to comply with the number of insurance years that are necessary to build up full pension entitlements. However, it is questionable if these years fit in with the actual situations of mothers in Germany and France.

In France, the infrastructure of institutionalised child care facilities is well developed according to the view of child rearing as an affaire d'Etat. Although most crèches are located in or near big cities, in Paris in particular (Wagner 2004: 23), there are many efforts to develop professional child care facilities. Various possibilities for the youngest children are institutionalized or are slowly becoming institutionalized (overview Wagner 2004). Nevertheless, guarantees for child care facilities are only given as of the age of three (école maternelle). Fifty percent of all children up to the age of three are taken care of mainly by the parents themselves, i.e., most of the time by the mother (Wagner 2004: 33). On Wednesday afternoons, when all facilities, including schools, are closed, children older than three also have to be taken care of.^2,3 However, considerable effort is taken in France to enable parents to have other people care for their child, also, for instance, through allowances for day mothers (AFEAMA, aide à la famille pour l'emploi d'une assistante maternelle, introduced in 1990 and extended in 1994). The intention is to make them also attractive, i.e., affordable, to medium or even lower income earners (working-mother model).

In Germany, 100% child care facilities are provided as of the age of three. Most of them (70%, Table 2) are part-time facilities. Facilities for children younger than three cover just about 7% (with huge differences between the former eastern and the former western Germany⁴).

Therefore, it is quite functional to compensate the three years lack of care facilities with three years of pension entitlements. And although these care credits have very positive consequences for mothers’ pension entitlements for those years, the required interruption of labour market participation leads to difficulties later on. One of the consequences is the difficulty of re-entering the labour market both because of a lack of practice and skill improvement (Schmid 2005) as well as reservations on the part of employers (Buchmann et al. 2003). This measure is even seen as a ‘disemployment strategy’ where women are concerned (Dingeldey and Gottschall 2001: 35, 37; Wagner 2004: 43).

There were, until recently, some reasonably well developed crèches in the east of Germany, and so the average employment interruption of east-German mothers is relatively short (for the generation born between 1951 and 1955 it is three years, for west-German mothers of this generation it is 9.6 years, AVID 1996; for related differences in women's pension entitlements see IGM 2001: 11).⁵ In addition, while full pension entitlements depend on life-long labour market participation, the participation rate of people aged 55 and over is very low (in France 40.7% of men, 35.2% of women, in Germany 53.5% of men, 37.5% of women, Eurostat 2006: 5). This will change, however, since conditions for early retirement are currently being abolished. Nevertheless, it is the participation rate of older women in particular that is disproportionately low, and there are developments, as in the east of Germany, to push older women out of the labour market (IGM 2001: 10; Rürup 2003). And although there is a system of minimum pension in France (minimum vieillesse comparable with the German Rente nach Mindesteinkommen) it is related to the 40 years of insurance and trimmed pro rata temporis (Veil 2002: 86–88).

The analysis of the effects of care credits on the duration factor of the pension norms leads to two conclusions. First of all, the introduction and expansion of care credits have to be evaluated positively. Women do have much better opportunities to get more, in ideal cases even full pension insurance years. The second conclusion, however, is that such an ideal is still difficult to attain because of insufficient care facilities. The ideal case requires that German mothers stay at home for three years and work at least part-time afterwards. We cover this point in more detail below. In France, in contrast, more effort is taken to retain women in the labour market. Although they receive care credits for two years and their child will have guaranteed care as of its third year onwards, there are also many initiatives for the provision of various forms of child care facilities even for the youngest children. Tax relief for the employment of (socially insured) child-minders is only one form. Nonetheless, statistics on the number of women's insurance years present another picture: on average, 28.7 years of insurance as opposed to the norm of 40 years (Table 3).

To summarize, in principle the required number of years of pension insurance can be complied with by mothers in both France and Germany, but in practice, they are not.

Apart from complying with the years of reference (of age, but of insurance in particular) to build up full pension entitlements, there are various factors that influence the level of entitlements. In both countries, this level depends on wage-related contributions, i.e., the level of pension entitlements is related to that of income. Care credits affect this level in several ways but other measures also do so. We briefly analyse here six different pension-level determining factors and how they are affected by care credits: part-time employment, gender segregated labour markets and gender pay gaps, shifts in the pension calculation, statistical discrimination, tax systems, and derived pension entitlements.

In both countries, women more often than men work part-time (in Germany 4.9 vs. 25.8%, in France 3.9 vs. 17.6%, Eurostat 2006: 5). One main reason for this is caring for children. And while women in both countries work less after having children, men work slightly more than before (EC-Report 2006b: 20).

German women usually alternate between part-time and full-time employment: 64% of west-German women, 80% of east-German women (AVID 1996; Stegmann 2003). Specific measures were introduced for mothers to compensate for their loss of pension entitlements due to part-time employment while rearing children. In other words, pension entitlements of women who have children under the age of ten and who are employed part-time or receive a less than average income are raised as if they earned an average income.

Child care for younger children (Kindergarten, age three to six, as well as primary schools, age six to ten) is normally part-time. Therefore, this measure to raise mothers’ pension entitlements corresponds again to the fact that child care facilities are insufficient. Part-time employment is an inevitable option for mothers with children of that age. This pension raising is limited in two ways: (1) it is possible up to the average income and (2) it is possible up to 50% of one's own actual income. The first limitation corresponds with the level of pension care credits for the first years, i.e., the paid-up non-contributory insurance period. The conditions provide incentives for mothers to accept lower wages. Therefore, the valuation of child rearing is again relative and only meant to diminish the losses of the less well-paid.

The second limitation has to be criticized in the context of part-time and its specific form of precarious employment. In west Germany, 10% of all women are precariously employed without social insurance. A total of 77% of the exclusively precariously employed in Germany are women. Eight percent of west-German women will face considerable difficulties not only in pension entitlements but also in income because they are in ‘assisting work’ (mithelfende Tätigkeit), 3% of them in the long term.

Due to part-time employment, the growth in the labour market participation of German women results in gendered participation: 89% of the part-time employed are women (Rudolph 2001: 26, 29). Such employment, which we might refer to as ‘female’ due to the participation ratio, is much less subject to tariff arrangements. For employees in private households, for instance, social contributions are not parity paid but have mainly to be paid by the employee (14.5%, whereas the employer only has to pay 5%). Tariff arrangements could result in occupational child care facilities, additional pension entitlements, or other positive labour market related rights (see below).

In France, women on average also alternate between several forms of employment (Fagnani 2000; Gauvin 2001; Veil 2002: 89). Although the number of part-time employed is lower than in Germany, it is not low in total and especially not in relation to part-time working men (Table 4; Majnoni d'Intignano 1999: 35; Eurostat 2006: 5).

Moreover, the last pension reform introduced incentives for mothers to work part-time. The official argument for supporting part-time employment is that it facilitates the resumption of work (Brossé-Verbiest and Wagner 2003). Pension entitlements are not raised since French pensions are calculated on the basis of the best 25 years of insurance (as of 2008). This means that a period of part-time employment does not necessarily influence the level of public pension entitlements. However, the former calculation norm, which included only the best ten years of insurance, provided better options for employees with non-linear employment.

In contrast to the situation in Germany, social contributions for child-minders in private households are covered by the state (Wagner 2004: 22). This disburdens or even enables the parent(s) to employ a childminder and at the same time it provides this childminder with social security. In addition, there are hardly any limitations pertaining to precarious employment in France, in contrast to Germany (Geringfügigkeitsgrenze). However, besides such rather positive incentives in France, there is another family policy that tends to have a negative effect on mothers’ labour market participation: the APE (allocation parentale d’éducation). Introduced in 1985 and extended in 1994, this form of financial support for parental care is an incentive for women to quit the labour market (Veil 2004: 19f; Wagner 2004: 43).

The gender-segregated labour markets and the gendered wage situation continue to be the main problem for women's pensions. There are numerous reports on this general problem, and therefore this article makes do with a few examples.

In France, the average income of regularly and full-time employed women is some 82% of that of men (NAP [France] 2002: 23; see also Majnoni d'Intignano 1999; Barrat and Meurs 2003). The percentage of women in top managerial positions is less than 15%, whereas their participation rate in service is at 90% (Insee 2003). There is a wage gap in particular between male and female cadres (Insee 2002) which is also reflected in gendered pension entitlements (Veil 2002: 67).

French calculations of women's average non-derived pension entitlements anticipate some 81% of men's pensions for 2015 (NAP [France] 2002: 23). German calculations for today's direct pension entitlements for women result in only 33% of men's direct net pensions (Table 5; Stegmann 2003).

Although this may change, gender segregation in labour market, wage and pensions will continue for a long time to come. This shows that the adjustments for mothers do improve their situation. However, the gender segregation is far from being reduced by these ‘reparations’.

French and German pension reforms introduced shifts in their pension constellations that influence the final pension levels. Pensions as a whole tend to be a constellation of multiple-interlinked modules. For instance, all contribution financed pension insurances are supported by government subsidies, as are most private pension schemes (Schmid 2002). However, in order to compare developments in pension constellations, the different conditions in France and Germany to build up entitlements can be broadly divided into three pension schemes: the contribution based public scheme, the contribution based additional occupational scheme, and the private scheme.

In both countries, the contribution financed public scheme is of special importance, and for women in particular due to the fact that fewer women than men have additional pension entitlements and where they do, the amounts they accrue are lower. In west Germany, the dependency-rate on public pension entitlements exclusively differs the most (43% of women as opposed to 17% of men), as does the level of entitlements (38% of women are entitled to less than DM⁶ 800, AVID 1996; for France, see Veil 2002). Widows pensions are the highest in both France and Germany, while the dependency on these derived pensions is highly gendered.⁷ However, bearing other forms of households in mind and facing the development of welfare states towards individualized social rights, we focus here on non-derived entitlements. From these direct entitlements women in both countries receive about half the pensions of men.

In Germany, the level of women's old age pension built up individually within the public scheme is about 50% of that of men and entitlements of east German women are some 62% of those of east German men (ASIB 2001). The main reasons for women's lower pension entitlements are reflected in the two ‘individual’ calculation factors: (1) the average number of years of insurance of women were, especially in the west, much less (west: women 25.5 years, men 39.7 years; east: women 40.6 years, men 43.7 years) and (2) the average level of annual entitlements was much lower (1.0 points on average for men, 0.7 points for west and 0.8 points for east German women, Table 6). Within the first particularly west-German phenomenon, not only non-employment periods but also low income employment periods (below the insurance threshold) result in this low number of participation years. The second reflects wage differences but also part-time employment.

In France, the differences are caused mainly by the average number of insured years which is 30 years for women and 41 years for men (calculated in trimesters: 121 versus 161, NAP [France] 2002: 22). We focus later on the non-proportional importance of the last trimesters.

While women depend on public pension entitlements much more than men, the benefits from this scheme continue to decrease. In France, since 1983, the average basic pension has depreciated, and the 1993 reform results in 20% lower pensions on average (Crevel and Wagner 2003: 46). In both countries, the benefits were continuously subject to political considerations so that different technical procedures determined the level of the public pensions’ benefits (for Germany, DRV45 2003: 76; for France, DRV45 2003: 106, Veil 2002: 88).

Both systems aim at disburdening the PAYG financed public scheme by focusing on additional pension entitlements. In France, pensions are expected to diminish by about 40% if no additional investments are made (Crevel and Wagner 2003: 46). In Germany, the intention is to extend the system by additional state subsidized investments. Such additional schemes form part of the average pension insurance of Frenchmen and Germans: on average, the French employed take part in 2.8 old age systems (DRV45 2003), east-Germans depend mostly on the public pensions only, while west-Germans build up pensions in two systems for old age (AVID 1996; ASIB 2001: 79).

In Germany, the level of importance of the different pension schemes is meant to change by introducing the so-called Riester pension. Today, about 82% of the employed are insured within the public scheme that covers about 78% of old age income (DRV45 2003). The relationships between the schemes are currently: 85% public schemes, 5% additional occupational scheme, and 10% private scheme. The last reform is meant to change this ratio to 60–70% public entitlements and up to 40% additional occupational and private entitlements (Table 7; Rürup 2003).

While it is planned that the level of public pension entitlements be trimmed, all pension entitlements together are meant to result in not less than 67% of the net pension level.⁸ However, to reach this objective it is necessary to build up additional entitlements. The minimum additional contribution of low income earners who continuously work as full-time employed amounts to €30 per month paid for 30 years continuously just to balance the losses of the last reform (Schmähl 2003: 7). Building up additional entitlements is still much more difficult for women, although Germany has introduced several measures to help mothers with this problem (Kerschbaumer and Veil 2002).

The range for private pension investments is even smaller in France. In total, private pension schemes cover some 2% of old age income (Crevel and Wagner 2003: 58).

In both countries, care credits were introduced into the public scheme. Care credits within the additional occupational scheme (as company arranged child care facilities) depend on tariff arrangements. Within the private scheme, care credits are possible as allowances paid from general revenue, as can be seen in Germany's Riester child allowances. The losses of pension entitlements caused by child rearing is hardly compensated for within these schemes that increase in importance. In general, the shifts in pension compositions tend to be positive for better earners and negative for most women's and mothers’ future pensions (Kerschbaumer and Veil 2000).

Furthermore, gender pension gaps are influenced by different forms of statistical discrimination. Although there are positive developments in European and national law to reduce gender discrimination there are still several measures and factors that cause specific forms of gender discrimination. To name a few:

Private investment: private schemes generally work on unisex principles. For women the gendered life-tables work out very negatively due to the calculations on women's average higher life expectancy, i.e., women's longer period of retirement: to get the same monthly benefits as men they have to pay higher contributions. Germany's preference to combine the Riester pension with additional occupational pension entitlements and to demand a criteria catalogue that includes life-long pension guarantees and same benefits for same contributions shows efforts towards gender-neutral second pillar investments. However, the possibility to invest in private pensions is still rather given for men's average higher income, and the combination of additional occupational entitlements with Riester allowances, which is particularly advantageous for women, is, due to different wage agreements, not (yet) always possible (Kerschbaumer and Veil 2002: 14).

Employment conditions for potential mothers: even if costs concerning motherhood are paid from general revenues (younger) women would have to face discrimination because of less investment in their skills and careers (Barrat and Meurs 2003). Several agreements in labour law have improved mothers’ positions and their right to re-enter their employment after a period of child rearing, and some branches offer better conditions than others which result in segregated labour markets. Although this is worse in other European countries, there are in Germany and France typical female employments that are on average less well-paid than typical male employments (Winter 1999). The deskilling assumptions of parental leave (see above) result in disadvantages for women's reintegration into the labour market. Women tend to re-enter employment in jobs with lower wages also for the same work and with less secure contracts, involving all that that entails for building up pension entitlements (Rudolph 2001: 27). In addition, wage compensations for periods of parental leave are quite low so that most fathers decide not to take parental leave (Drew 2005). Various incentives form part of a vicious circle to maintain the gender roles.

Conditions to join additional occupational schemes: statistical discrimination can also be seen in specific conditions to claim pension rights such as those for German occupational pensions: employees need to be employed by the company for five years and have to be at least 30 years old before leaving the company. Although these conditions were alleviated and both reduced by five years in 1999, they still have a negative impact for mothers in particular. In general, level and participation in additional schemes are highly gendered. In west Germany, only 10% of women and 50% of men received additional occupational pensions and those women who received such pensions have to make do with half of that of men (ASIB 2001: 161).

In France, differences in additional occupational pension entitlements emerge for families with more than two children: depending on which additional scheme one belongs to, the pension raising percentage varies between 5 and more than 10% (Table 8).

Further disadvantages for women's additional occupational pension entitlements do not depend on specific conditions but on related factors, such as the lack of care credits (annuités) within this scheme. The differences in participation and level of additional occupational pension entitlements reflect the gender bias in wages and the gender segregation in labour markets.

Method of calculation: as already pointed out, the labour market participation rate of older people is gendered. If the last years of employment improve the pension entitlements in a non-proportional way, as is the case in the French labour market, women are placed in a secondary position (carrière complete 84% men, 34% women, Observatoire des retraites 2002: 48). In addition, the French calculation basis of the best 25 instead of ten insurance years is disadvantageous for discontinuous work biographies.

In addition, there are other structural discriminations in both systems, particularly in the tax systems. Following their traditional subsidiary family-model, both countries continue to work on non-individualized principles. The tax systems, for instance, favour classical families. In 2003, tax incentives were introduced for French parents to care themselves as well as for temporary interruptions of employment for mothers (Veil 2004: 19). There are also tax incentives that work out positively for mothers, such as tax release for company-internal child care facilities. However, the introduction of the former incentives strengthen the traditional roles of the caring mother and the (paid) working father and give incentives to a one and a half-earner model. In contrast to the purely marriage-oriented German income tax system, the French system calculates on the basis of a family quotient, i.e., each child reduces the part of the income on which tax has to be paid.

The German tax-system works on so-called spouse-splitting (Ehegattensplitting). Allowances and the level of tax-deductible income is derived from the household's income. Spouse-splitting of income-tax works out negatively for women's labour market participation mainly due to the fact that the combination possibilities work out negatively for the lower income which is usually earned by women (Dingeldey and Gottschall 2001: 35; Melzer 2003). Recent developments include shifts of pension taxation from contributions to benefits which could have a levelling effect on gendered pensions (DRV45 2003).

The last factor we analyse here is derived pension entitlements. In both countries women have, on average, relatively low own, i.e., non-derived pension entitlements. In France, within the public pension scheme and the obligatory additional schemes, women receive about half the pension of men (direct entitlements: women receive €593, men €1,200). Their retraite totale, i.e., pensions also including non-contributory insurance periods such as care credits, is €767, that of men is €1,342 (DRV45 2003: 105). Pension entitlements are clearly gendered with levels up to €900 for women and levels above €900 for men (see also NAP [France] 2002, Appendix II: 35f). The German situation is not much different. The net old age income of women in the west is some 48% of that of men (DM 1245 versus DM 2618⁹), in the east it is about 79% of that of men (DM 1322 versus DM 1668, AVID 1996; the differences are even greater in the ASIB report of 2001: 79). In the west of Germany, 38% of all women have entitlements lower than DM 800 (AVID 1996), which corresponds to the high percentage of west-German women who are only entitled to public pension entitlements (43% women, 17% men, see above). In west-Germany, the importance of derived entitlements is very high due to the fact that 21% of women older than 65 do not have their own entitlements at all (ASIB 2001: 110). It is remarkable and ambiguous that from now on such non-pension insured spouses are entitled to Riester allowances without paying their own contributions. While women's pension situation is described as being dual, depending on one's own entitlements and marriage (IGM 2001: 9), these facts even manifest solely dependent situations.

Pension dependency situations are the most critical for divorced women in both countries. While many more women than men in Germany have old age entitlements below DM 1000, the worst position is that of divorced women (the net old age income of 24% of west and 27% of east German divorced women is below DM 1000, AVID 1996; ASIB 2001: 114). Also in France the worst pension position is that of divorced women (called ‘Le drame des femmes divorcées’, Union des Familles 2004). This situation will improve in both countries thanks to growing labour market participation of women (although this growth mainly concerns part-time employment). German mothers work much less than their French counterparts. However, the newly introduced pension splitting could improve their situation. Nevertheless, for this splitting, both partners need to have at least 25 years’ pension insurance. Pension situations for the divorced may have been improved but they are still not satisfactory (ASIB 2001: 94, 97). This voluntary splitting is just possible for couples married after 2002, therefore, its effect is not yet clear. However, contrary to pension splitting by family court, these rights are then unconditional entitlements (Rürup 2003: 76).

For children, specific allowances are added in addition to the above-mentioned child care credits. In France, for families with at least three children, pension allowances are given related to the level of income. This means that the percentage increase (10% in the basic pension, up to 10% in the additional occupational pensions) is asked for the generally higher income of the father which discriminates against the mother (Majnoni D'Intignano 1999). The system of child allowances, including actual family allowances as well as old age allowances, is built up to support large families.

To summarize, most of the measures discussed above cannot be seen as working out only positively or only negatively for women's direct pension entitlements. Most of them may be meant to improve women's old age situation. And some of them do in fact improve their situation. Nevertheless, purely marriage related entitlements were criticized as being far too limited while not having in mind situations of non-married or singles. Non-individualized measures, especially tax-regulations, were criticized for aiming at maintaining the gendered and familialized roles. Although with different nuances, both the French and the German models, reflected in their tax and pension systems, are designed to work out the best for a 1.5-earner household. The different measures form part of the very complex systems that as a whole determine the outcomes of such measures.

As shown above, developments in German and French pension systems are comparable when focusing on gender comparison. Of special importance are retrenchments in the public schemes, shifts in the pension constellation, and incentives for mothers or wives to work part-time or to quit the labour market. The main exceptions within French and German retrenchment policies, and a remarkable new step within European pension systems, is the introduction and development of child care credits. The effects of French and German care credits vary. In France, pension credits as they developed, give incentives for women to quit the labour market whereas pension raising percentages work out the best for the family income or for career mothers because they result in a relatively higher raise than those of mothers with lower income. The German care credits work out the best for mothers who interrupt their labour market participation for three years and who afterwards are in part-time employment or earn less than average income. Both systems put considerable weight on women's ‘traditional’ family role. France, differently from Germany, gives stronger financial support for families with more than two children and hardly any for families with only one child.

In Germany, recent developments in the calculation of pensions and of the newly introduced nursing care insurance gave, in contrast to the past, substantive weight to the parenthood factor. This is a striking new factor for: (1) a system that was criticized for subsidizing childless wedlock (Papier 2001) and (2) for the strictly labour market oriented German pension system. However, the dominance of the labour market in building up pension entitlements remains. This is problematic for German women mainly due to the fact that child care facilities are far from being sufficient and children continue to be taken care of by mothers in particular. Mothers remain in a secondary position in the labour market due to those care responsibilities, yet women in general do stay in a secondary position due to related statistical discrimination. This secondary position in the income level is even more disadvantageous when the constellation of pensions is shifted, and one's ‘own responsibility’ to invest (privately) grows. These developments make it harder for women to build up sufficient entitlements. Other allowances, as for instance the child care allowances within the Riester pensions, improve their entitlements. However, since the Riester entitlements are necessary to maintain the current level of public pensions, the term ‘improvement’ is, in fact, improper. Some developments, in addition, tend to deliver incentives for women to work part-time or even to quit the labour market, as for instance Riester allowances for wedlock.

Within the French system, in which the labour market has the same importance for pension entitlements as in Germany, mothers bear less care responsibility. Although far from sufficient, the coverage of child care facilities is much higher than in Germany, and French policy is much more effective in extending non-parental child care. However, pensions in France are still surprisingly just as gendered as they are in Germany.

On average, the recent pension reforms in Germany and France will entail that the level of benefits will decrease in future years, for men and women. Women, however, thanks to a tendency of increasing labour market participation (even when often part-time) and the pension care credits will partly compensate these foreseen retrenchments. Therefore, we can expect that in the future the gender pension gap will be slightly reduced, in relative terms probably also due to the impaired labour market biographies of men.¹⁰ Nonetheless, it will remain.¹¹ In France, according to Veil (2002: 69), it became even bigger within the last ten years. To adjust seemingly individual disadvantages does not eliminate the countless structural disadvantages: while the years of insurance may be complied with by women thanks to care credits, they will never reach the level of men's pensions. Due to the fact that both countries’ pension systems are strongly linked to wages, and due to the fact that wages and labour markets are gendered, pensions are, per se, gendered.

Both systems introduced measures that are too limited to change this, as for instance care credits when analysing the compatibility of care facilities and work participation requirements. Seemingly gender neutral law, and even ‘women -friendly’ measures, do not at all have to lead to adequate options for both, men and women. The last reforms, in 2001 in Germany and 2003 in France, respectively, again show that the necessity to improve mothers’ pension situation is recognized. Nevertheless, the solutions are problematic: although introducing positive measures for mothers in particular, they are only additions and adjustments to a deeply gendered system. Moreover, they are ambivalent in the sense that they tend to strengthen gender segregation and encourage the development of a 1.5-earner model. As it is, with only adjustments being made to the male norm, the individualization of social rights is far from being implemented in a non-gender discriminatory and equality facilitating manner. In times of individualization and retrenchment this is financially essential for each pensioner's household. However, bringing an end to gender pension gaps is likely to take a considerable amount of effort, and a ‘new gender contract’ (Esping-Andersen 2002) has to be set up.

Patricia Frericks is a political scientist and junior researcher at Utrecht University. Robert Maier is a social scientist and senior researcher at Utrecht University. Willibrord de Graaf is a psychologist and senior researcher at Utrecht University. Together they form the Dutch partner of the European Research Network RESORE (Employee's Resources and Social Rights in Europe).