Gender Arrangements and Pension Systems in Britain and Germany: Tracing change over five decades

This paper studies the modernisation of gender arrangements and the restructuring of pension systems in the United Kingdom and Germany since the 1950s. We firstly aim to pinpoint the time when pension programmes were apt components of the ‘strong breadwinner model’. Secondly, we explore the assumption that pension systems are tools of stratification, by comparing the ways in which the constraints and incentives of these pension systems have been in line with typical life courses of women. Our paper argues that the constraints and incentives of pensions have altered quite significantly over time and questions whether they have appropriately been characterised as components of strong breadwinner models over the long term. In the UK the pension system only supported the strong breadwinner model until the mid-1970s, whilst the German system never fully supported it. In addition, it is shown that the impact of pensions on women’s behaviour is relatively limited. At times, women’s lives were in accordance with the male breadwinner model, and they suffered high poverty risks despite having potential access to a more modern pension regime; during other periods, their employment choices were at odds with the strong directives issued by pension International Journal of Ageing and Later Life 68 regulations to stay at home. This demonstrates the importance of taking other factors, such as cultural influences and other societal institutions into account when exploring the impact of social policies on citizens’ lives; but it also poses the question of whether pensions are really important building blocks of the breadwinner model.


Introduction
Since the late 1960s, one of the major transformations experienced by European societies has been a change in their gender arrangements from the traditional housewife marriage towards modernised forms.This modernisation process fundamentally changed societies; it had implications for the organisation of private households, the labour market, and cultural norms.Certainly, it also brought with it challenges for states, and welfare states in particular, which came under pressure to restructure social protection away from housewife-based systems towards individualised systems.
This paper aims to study the modernisation in the United Kingdom and Germany of gender arrangements on the one hand and of social policies on the other hand, using pensions as an example.However, rather than analysing the implications of both processes for contemporary society, as is the case in a range of studies, our paper will assess, over five decades, the restructuring of pension systems in Britain and Germany, and by tracing their changing constraints and incentives for women's lives, we aim to pinpoint the time when both pensions systems where components of the 'strong breadwinner model' to which, according to the literature, both countries' welfare states belong.Secondly, we will compare the way in which the constraints and incentives of these pension regimes have been in line with typical life courses of women.
The adoption of such a perspective is not common in comparative studies of the welfare state.Since the early 1990s, this research has been dominated by the regime approach.This literature aims to capture typical features of national social policies and their impact on society; on this basis, various welfare state models have been developed.While this approach has been successful in highlighting differences between countries, one weakness is that it is conceptually constrained regarding the incorporation of policy change (Cochrane & Doogan 1993;Duncan 1995;Offe 1993;Taylor-Gooby 1996).As a consequence, regimes have acquired a static nature.This is not a new observation, yet little research has been carried out that systematically evaluates the transient nature of social policies over time.The current paper aims to make a contribution to knowledge in this area.In our case studies of pensions presented below, we argue that if we trace changes in this core policy area over a longer time period, it becomes apparent that the incentives, constraints and the redistributive potential of this policy also alter quite significantly.This observation potentially questions the long-term validity of welfare regime types, given that pensions are normally seen as building blocks of regimes.
Our paper diverges from the regime literature in another respect.One of the achievements of this literature is to highlight the impact of typical formations of the welfare state on stratification (Esping-Andersen 1990; Orloff 1993; Daly 2000).Social policies are shown to constrain citizens' preferences, cement patterns of social inequalities and slow social change.This perspective focuses on the power of the welfare state over citizens' life choices and opportunities, yet it does not explore its restrictions, for example by looking at the way in which citizens make use of such policies or by analysing how citizens' lives change social policies.In the following, we explore the limitations of the constraining nature of pension programmes by asking to what extent social actors actually lived lives in line with social policy incentives.To this end, in our case study below, we will use three typical female biographies, each illustrative of a different period since the 1950s (see table 3) and we will ask how they related to the constraints and incentive structures of German and British pensions policies at these different points.Through this we will show that the stratifying impact of pensions is looser than is generally assumed in the regime literature.
Overall, we argue firstly that both these pension systems have not strongly enforced the traditional division of labour -with the exception of the UK until 1975.The German system has been less constraining and offered women greater autonomy than has been assumed by regime the-ory because potentially it supported employment as well as marriage from the start, and later also integrated care work.It is therefore typical of a modified, and not a strong breadwinner model.The British pension regime had all of the features of the strong breadwinner model until the mid-1970s, before which it had prescribed gender inequality; since then, however, this label is no longer appropriate.The modernised system, valid until the mid-1980s, afforded women some autonomy and recognised care work.The subsequent retrenchment of 1986 led to a consistently low level of protection for substantial groups of the population, it was thus highly commodifying, undermining a strong breadwinner model.Since the millennium, the British pension system has only supported a moderate breadwinner model.
Secondly, we show that women's typical biographies are not always in line with the constraints and incentives of pension systems.In Germany, women did not make use of the opportunities to accrue independent rights until the early 1980s, and, despite the more inclusive system, they still lead more traditional lives than British women today, with the exception of East Germany.By the same token, in the United Kingdom, women had entered the labour market in greater numbers since the late 1960s, regardless of being constrained by their authoritarian pension system.Moreover, despite the consistently low level of protection for substantial groups of the population during the second half of the 1980s and throughout the 1990s, we do not see that the traditional gender arrangement was abandoned altogether.Instead, the male-breadwinner/parttime carer model became dominant.
Our paper is structured as follows: In the first section, we discuss the usefulness of the welfare regime approach for analysing the relationship of gender modernisation and social policy from a cross-national perspective, and we explain why we have chosen British and German pension policies for our case study.
In section two, we outline the similarities and differences of the gender arrangements in both countries since the 1950s, and show how they have been modernised.
In section three, criteria for the cross-national analysis of pension systems from a gender perspective will be introduced.We then analyse the restructuring of pension systems from the 1950s until today by assessing their constraints and incentives with regard to the strong breadwinner model and by comparing these with typical biographies of women.
Section 4 discusses our findings.
'Welfare regime' approaches and social change Public pensions were among the few policy areas Esping-Andersen selected in order to build his seminal welfare state typology (1990).
According to this study, Germany typifies the conservative welfare regime because the state reproduces the existing structure of social inequality amongst the employed; social rights are income and status-related and family policies promote unpaid work.Germany's Bismarckian pension system is a perfect example for this type of welfare state.The United Kingdom comes close to the liberal category because social rights are means-tested and comparatively low, which tends to cause sociostructural polarization, and private welfare is subsidised.Again, the Beveridgean pension system with its low level benefits is typical for this type.
Esping-Andersen's welfare regime approach was criticised from a gender perspective.Researchers argued that it did not systematically consider the family and the unpaid work that women perform in it, a point that Esping-Andersen conceded in his later work (1999).Feminist researchers therefore argued that citizenship needed to be 're-gendered' from the perspective of women (Lister 2000).Against this background an alternative typology of welfare regimes was developed, that was firstly sensitive towards the typical social risks that women face in modern society, and secondly towards the role of social policy in enforcing structural inequalities between women and men (for overviews see Daly 2000; Siim 2000; Bambra 2004).Most prominently, Lewis (1992) and Lewis and Ostner (1994) classified welfare states into strong, moderate and weak male breadwinner states, depending on the degree of support given to the family model that assumes employment for men and home-making for women.In this typology, West Germany and the United Kingdom both came out as 'strong male breadwinner states'.In the context of this classification pensions were not considered in any empirical detail, but they were included in the income support systems capable of assuming women's dependency on men, for example by granting derived rights or by allowing wives to forfeit their social insurance contributions during working life.More recently Daly (2000) found that, through their transfer and tax systems and service provision, Germany and the United Kingdom's welfare states suggest employment as the norm for men, but not for women.However, where the German state actively supports the traditional division of labour through subsidies across classes while neglecting services, in Britain we find a passive toleration of gendered inequality (Daly 2000). 1 Thus, on the basis of the regime literature on gender, by choosing the United Kingdom and Germany we have selected two welfare states as case studies that have given support to the traditional division of labour in the past.
As explained above, our first aim in this paper is to explore the validity of this characterisation over time.This question has gained significance in the light of many of the challenges of post-industrial society; it has been asked, for example, whether it is still appropriate to speak of the strong breadwinner model today (Bleses 2003;Lewis 2001;Seeleib-Kaiser 2002), or whether economic globalisation eradicates the differences between regime types (Scharpf 2000).Such analyses aim to assess whether these models are outdated.What has not yet occurred to any great extent is a retrospective appraisal of the changing nature of regime types over time.However, given almost continuous social policy reform, it seems just as appropriate to ask at what point the regime types became valid and how they changed.
We decided to pursue this question by isolating one policy areapensions -and by assessing the extent to which this supports the traditional division of labour.Ideally, for an appraisal of a regime, our study would need to take all its major policies into account.However, this would be an undertaking far too large for the scope of this paper.Still, we believe that our study of pensions can cast some light on breadwinner models as a whole.Firstly, pensions are an important part of the benefit and income maintenance system that is a core area of regime analysis.
Secondly, regime analysis is based on the implicit assumption that "most of the key policies will indeed reflect a similar approach to issues of public welfare" (Kasza 2002).This means that if the findings for one policy area do not support the character of the regime, this weakens its overall conceptual strength.
Our second main question refers to the relationship of social change and social policy, or, in our case, gender modernisation and pension reform.In this regard, the regime literature often assumes that national welfare states impose constraints on citizens and direct their behaviour through incentives.Social differentiation is explained by state structure.For example, it is common in the comparison of welfare state policies to treat women's labour force participation rates as a direct result of social policies; accordingly, national profiles of female employment are often treated as an indicator of welfare state differences.Accordingly, a low labour force participation rate and a high part-time employment rate of women are seen as the outcome of a "conservative welfare regime" in Esping-Andersen 's approach (1990, 1999) and as the result of policies typical for a "strong male breadwinner model" in the approach of Lewis (1992) and Lewis & Ostner (1994).This perspective pays too little attention to the complex interplay of factors that influence women's behaviour.It does not focus on agency, and is less interested in societal dynamics that are independent of the state.Yet social policies, while undeniably powerful stratification instruments (e.g.Goodin et al 1999), only have the effect that regime analysis would lead us to expect to a limited degree.It has been shown, for example, that welfare regimes are not necessarily in line with the development paths of women's labour force participation; specifically, Norway and the Netherlands, whose welfare states Esping-Andersen classified as social-democratic, were long characterized by relatively low levels of women's employment (Ellingsaeter 1999, Leira 2002).Due to such inconsistencies, assumptions about the way in which social policies affect the social actions of women have been criticised for being insufficiently complex and thus in need of further exploration (Bang et al. 2000;Daly 2000;Duncan 1995, Pfau-Effinger 2004a).
We argue that other factors besides welfare state policies contribute to explaining the decisions women make with regard to employment and the family.In particular, the cultural context of societies needs to be taken seriously, as well as the impact of other central institutions such as the labour market.Due to these factors, 'structural lags' (Riley & Riley 1994, Kahn 1994;Foner 1994) between women's economic behaviour and the restrictions and incentives of welfare state policies are possible.Such structural lags characterise a situation in which policies and practices adapted to the behaviour of earlier generations continue to be enforced, even though the behaviour of later generations has changed.
In the following section, we will analyse the changes of female labour market behaviour in the context of the respective gender arrangements since the 1950s and consider their interrelation with pension policies.
The modernisation of gender arrangements in Germany and the United Kingdom Modern societies arrange the lives of women and men in typical ways (Pfau-Effinger 1998, 2004a).A society's gender arrangement consists of cultural values and models, and of the regulation of the main fields of women and men's work through institutions.It is defined by the prime societal spheres in which childcare and elderly care take place and by the social groups that inhabit these spheres as carers.Furthermore, the gender division of labour, the power relation on which it is based, and the degree to which the relations of spouses are based on dependency or autonomy determines the nature of the arrangement (see figure 1).An arrangement is the historical outcome of negotiations, conflicts and compromises between social actors with differing power, and can be renegotiated and changed by social actors.The cultural values underlying the gender division of labour and the ways in which women and men organise their biographies, as well as the values suggesting what their main sources of income and social security should be, are central elements of the gender culture and are of central significance for women's and men's old age.Modern welfare states refer to such dominant cultural values and models in the population, either by reflecting them, or under certain conditions, contributing to the emergence of new cultural values (see Pfau-Effinger 2005a).
The gender arrangements in Germany and Britain have possessed similar central features, and these have changed in comparable ways since the second half of the 20 th century, mainly due to similarities in the ments on the basis of different cultural models of the family, see Pfau-Effinger 1998;2004a;b.cultural basis of the two gender arrangements. 3In the Western Europe of the 1950s and 1960s, the housewife marriage was the main cultural basis of this arrangement.This model was based on the premise of a fundamental separation of the 'public' and 'private' spheres, and a corresponding location for the two genders: the husband's proper work was in the 'public' sphere, while the housewife was responsible for the private household and childcare; her financial security relied upon his income.This model is linked with the cultural construction of 'childhood', according to which children need special care and comprehensive individual tutelage of the mother in the private household.
In countries with a very strong housewife model during the 1950s and 1960s, this model was increasingly substituted by a modernised version based on the integration of both partners into employment (Pfau-Effinger 2005b).This new gender arrangement was either mainly based on a male breadwinner/part-time carer model of the family (for example, this was true, for the United Kingdom and Germany) or on a more egalitarian dual breadwinner/state carer model (for example, in Norway). 4 It is characterized by a fundamental cultural transformation, and exhibits a relatively high degree of change.
In Britain and Germany, a process that stretched from the end of the 1960s to the end of the 1990s considerably weakened the central position of the traditional family model as the cultural basis for the gender arrangement.This transformation was initiated principally towards the end of the 1960s, when a fundamental contradiction at the cultural level had come to a head: the incongruity between the cultural construct of autonomous and equal citizens of modern industrial societies on the one hand, and the construct of the inequality and dependence of the house-3 For a theoretical framework to explain cross-national similarities and differences with respect to the historical roots of cultural family models see Pfau-Effinger 2004b.4 In those societies where the housewife model of the family did not play a particularly strong role in history, e.g.Finland, Denmark and France, the gender arrangement developed along a different development path towards a dual breadwinner/institutionalised carer model, see Pfau-Effinger 2005b.
wife model on the other.In addition, at this point, alternative gender relationship models emerged on the level of gender culture, made possible by the newly forming international and national feminist movements that seized upon these contradictions (Pfau-Effinger & Geissler 2002).As a consequence, the old housewife model of the male breadwinner family as the dominant cultural image was increasingly replaced by the full-time male breadwinner/female 'part-time carer model '. 5 This type rests essentially on the vision of full integration of women and men into paid economic activity.At the same time, however, it is expected that women, as mothers, may interrupt their economic activity for a few years, after which they combine employment and responsibility for childcare through part-time work, until their children are no longer considered to require particular care.The new cultural models for family and gender relations are characterized in both countries by the expectation that the mother should be employed, but also by the idea that 'private' childhood should still play an important role in family life.This cultural change in the gender arrangement was interrelated with general processes of cultural change, which led to an overall positive reassessment of the value of individual autonomy (Beck 1992).However, the new cultural 'male breadwinner/female part-time carer model' is still contradictory at its core.The financial dependence assumed by a woman who cares for her own children stands in contrast to the high cultural esteem enjoyed by autonomous financial security.This problem has been reduced by welfare state policies, which have established new 'social rights to provide care' (Knijn & Kremer 1997).
In accordance with these cultural transformations, the main features of the gendered division of labour have changed in Britain and Germany.In both societies, women's labour force participation has increased considerably.Today, in both countries a sequence of employment interruptions and part-time work is the norm (Fagan et al. 1999 Despite these overriding similarities between Britain and Germany, there are also country-specific differences.As table 1 shows, the labour force participation rate of women and its increase was higher in the UK than in West Germany from the 1970s, and the gap between the two countries in this regard remained until 2001.1960 1965 1970 1975 1980 1984 1990 1994 1998  Source: http://www1.oecd.org/scripts/cde/default.asp;From 1991: reunified Germany, before: W-Germany. In West Germany, the church and religious ideas were influential for the introduction of the housewife model of the family into welfare state policies, while this was less true for the United Kingdom (Lewis & Ostner 1994).In addition, during the expansionary period of the British welfare state in the 1960s and 1970s, more jobs for women were created in the public sector and in the service industry in general, contributing to higher employment participation rates of British women than West German women throughout the period (Meyer 1997; Table 1).These were accompanied by a higher degree of individualisation in the UK than in West Germany with regard to divorce rates, single parenthood, cohabitation and births outside marriage (Castles 2003;Hantrais 1999).Furthermore, attitudes in Germany are slightly more conservative than in Britain.According to the data of the ISSP ( 2002), the share of respondents who support a traditional gender division of labour is higher in West Germany compared with Britain, as is the proportion of those who be-lieve that the employment of mothers of pre-school children is detrimental for their children's well-being (see table 2).German women; they are more likely to search for a job, and their overall economic activity rate is higher. 6o sum up, it should be highlighted that since the 1950s, the lives of women in the context of both societies have undergone a substantial transformation.During the 1950s and 1960s, the gender culture affirmed the breadwinner model in both countries, and the transformation that has taken place since meant that the belief in the appropriateness of the traditional gendered division of labour was shaken.However, it should also be kept in mind that the transition towards the full-time male breadwinner/female part-time carer model that has taken place retains the personal dependency of women as prime carers in the home as an integral part of the gender culture.
How do these changes relate to social policies?
The modernisation of pension policy in Germany and the United Kingdom from a gender perspective It was argued above that more attention should be paid to the long-term changes in the relationship between the evolution of social policy and gender modernisation.In the following section, we will analyse how pension reforms in Britain and Germany since the 1950s have coincided, sometimes in contradictory ways, with the changing dominant cultural models of the family and of women's behaviour between family and employment.Based on our characterisation above, we seek to outline the direction of social change and policy reform since the 1950s.

Criteria for analysing policy change and social trends
Our main purpose is firstly to characterise the constraints and incentives of both pension systems over time, i.e. to delineate the options afforded to women by the system as well as the boundaries, and to assess the degree to which they support the 'male breadwinner model', which as-sumes employment and personal autonomy as the norm for men, and economic inactivity and personal dependency, mostly through marriage, as the norm for women.To do this, we will apply the following criteria.

Strong enforcement of the breadwinner model through pensions
In its strongest form, a pension system enforces the male breadwinner model if married women have limited or no opportunities to claim independent pension rights, and instead are expected to claim derived rights based on marriage status only, thus becoming dependent on their partners for protection in old age.Such a system contradicts the value of personal autonomy.It is usually connected with the principle of fidelity in relation to the male breadwinner beyond his death, which also means that rights accrued in one relationship are terminated as soon as a woman re-marries.Typical policies for this form are derived rights for widows, couples pensions (accrued by the breadwinner), means-tested pensions based on household income, and barriers to rights accrual for typical female employment patterns, such as part-time work.

Moderate enforcement of the breadwinner model through pensions
A moderate form of enforcement exists if becoming dependent on a partner for protection after retirement through derived rights is one option for married women, but if it is also possible for them to generate independent rights.However, these independent entitlements are unlikely to generate sufficient benefits to guarantee financial autonomy for typical female employment and care biographies, making dependence on the partner in retirement likely.Characteristic policies are partly the same as above, e.g.derived rights for widows, couples pensions generated by the breadwinner, means-tested pensions based on household income, but it is also possible to generate individual, employment-related rights, independent of working hours or type of employment contract.In addition, care responsibilities can be rewarded by pension rights in this system.

No enforcement of the breadwinner model through autonomous pensions
A pension system does not enforce the traditional division of labour if all individuals are assumed to generate independent rights sufficient to be autonomous.Pensions generated in such a modern system must be above the poverty line, because otherwise they still render women dependent on their partners.This system either adopts a citizenship pension, or acknowledges all employment activities together with care times as a sufficient basis for a pension above the poverty line.Widow's pensions and married couples pensions do not exist in this type, and neither does means-testing based on household income.

No enforcement of the breadwinner model through insufficient pensions
Finally, a system does not enforce the traditional division of labour, if, regardless of its type of rights accrual, the pension level spells poverty for many and thus pushes men and women towards the labour market.Typical policies in this system are a state pension below the poverty line and no compulsory coverage by second or third pillar pensions.
*** These criteria will be applied across public pensions, employer co-financed, occupational pensions, and personal, market-based pensions.This broad perspective is the reason why we speak of national 'pension systems' in our analysis -the term reflects our assumption that the significance and role of each pillar is influenced by the role played by the other pillars and that they have to be assessed in conjunction in order to gauge the constraints and incentives for citizens in a comprehensive fashion.For example, in Britain, occupational and personal pensions developed due to a low level of public pensions, and today, they are critical contributors to the social security of retirees.In Germany, by contrast, occupational and private savings for a long time remained fairly insignificant because they were crowded out by the high replacement rates of public pensions (Bonoli 2003).
In the following, we will compare social trends with pension reform, in order to assess whether women have made decisions that followed the constraints and incentives set by these social policies.In methodological terms, we will do this by assuming typical female biographies for different periods.These are built on our discussion in section two, and are as follows: We assume that the traditional housewife model was dominant in Germany until the beginning of the 1980s, and in the UK until the beginning of the 1970s (see table 3).It was followed by the breadwinner-part-time worker model, dominant in both countries until the late 1990s and followed by more individualised household structures, which in Germany are more likely in the new Länder [former FRG states].We will ask whether the constraints and incentives of the pension systems are in line with these typical biographies (see appendix for a schematic overview).After the Second World War, social policy makers had to decide whether to continue the tradition of the Bismarckian pension system that had survived the Third Reich, or whether to replace it with a more universal, citizenship-based pension.In West Germany, the latter had little chance because of the breaking up of the Allied Control Council, the onset of the Cold War, and the political dominance of the Christian Democrats (CDU) after 1948.Thus the German pension system continued its tradition (Hentschel 1983;Kulawik 1992;Baldwin 1990).In the 1950s, the main characteristic of this system was a very low level of benefits for all recipients.A widow's pension existed, worth 60% of her husband's entitlements (Hentschel 1983).This pension level was widely recognised as being insufficient, and in order to address this poverty threat, a major overhaul of the system was initiated.In 1957, the Christian Democratic government introduced a pension calculation formula, which in altered form is still used today.It took into account the individual's wage level, length of employment, a general wage level, and it rose by a fixed rate.This formula guaranteed a pension for many full-time workers of about 70% of previous net earnings, but there was no supplement for breadwinners, i.e. for pensioners with dependants.However, spouses received a widow's pension at 60% of this income.There was no minimum pension, as the CDU had refused to introduce such a threshold.Nevertheless, from then on, the majority of pensions were no longer far below the social assistance level (Hentschel 1983;Schmidt 1998).
From the 1970s onwards, the improvement of individualised social security for women was a recurring issue in the political debate on pensions.It was introduced into the discussion by the feminist movement and supported by collective actors such as unions, welfare associations and political parties.Its establishment as a party-political aim was helped by a decision of the Constitutional Court in 1975, which obliged the legislator to make the pension system more sensitive to women's social risks (Gerhard 1995;Meyer 1998a).At the same time, gender equality was a contested issue and the Social Democratic (SPD) government of the 1970s was ambivalent about women's position in society and expressed their worries about the destabilizing effect on children of mothers' employment (Opielka 2002).
The first reform to improve protection for vulnerable groups was undertaken in 1972 by the Social Democratic government, which increased the pension entitlements that low-waged workers could accrue (Hentschel 1983;Schmidt 1998:96).In 1976, a reform of the marriage law stipulated that after divorce, both partners could claim an equal share of pension entitlements acquired during marriage.This equal split was based on the idea that women and men contribute equally to the family income, through waged work and housework (Veil 2003).A further modernising step was taken in 1985, when the Christian Democrats decided that independent care-related rights would be granted.The reform was widely recognised as a major innovation, because for the first time, informal family care work was put on an equal footing with paid employment (Schmidt 1998), which meant an extension of the definition of work and the adoption of the 'right to provide care' (Knijn & Kremer 1998).The one-year initially granted was successively extended to three, in 1989, 1992, and 1999, and calculated on a higher level.In addition, during the 1990s, the entitlements for low wage earners improved further, women's retirement age was extended to 65, and early retirement possibilities were phased out (Meyer 1998b).Moreover, with the introduction of elderly care insurance in 1996, entitlements to pensions for carers of relatives were provided.This extension occurred alongside a general reduction of the highest possible pension level, in the context of a discourse of fiscal crisis and the threat of an ageing society (Meyer 1998a).Private pensions never played a great part in the German system until 2001.We therefore only need to follow the development of these pillars after the millennium.
In what way did this system support the male breadwinner model and to what extent were the constraints and incentives in line with the typical biographies of women?
Our analysis shows that the German pension system never strongly supported the male breadwinner model because from the start, it recognised not just marriage but also employment as a basis for rights accrual for married women.In its earliest phase, the pension system was strongly commodifying for both genders of all social classes because of the very low level of employment-related rights and the even lower level of rights derived through marriage.The general poverty risk was high, but it was particularly high for widows.The system therefore did not suggest personal dependency as a secure option for married women and was therefore not typical of a breadwinner model.From 1957, this situation improved because rights increased significantly, yet there was never a higher pension for claimants with a dependant spouse, which could have acted as disincentive for women to be employed.From 1957 onwards, women's independent rights were continuously strengthened, through the improved recognition of part-time and low-paid work, and from the mid-1980s of care responsibilities.This system suggested to women dependency on a partner as only one option amongst others, such as independent employment and care work, typical for a moderate breadwinner model.Nevertheless, until the millennium, the German pension system was never fully modernised as it kept the widow's pension and never adopted a guaranteed minimum.
If we compare these constraints and incentives with the typical female biographies assumed in table 3, perhaps the most interesting fact is that until the late 1970s, the German pension system was more modern than women's lives, because despite the possibility to generate protection through employment as well as marriage, the majority of women became housewives.Adult women therefore benefited from the reforms of 1957 and 1972 predominantly through the improved status of their husbands, but with a high risk of poverty remaining especially for working class women relying on a low male wage, because of the absence of a minimum pension.Thus, working class widows in particular, and possibly also couples, continued to be at risk of poverty until the late 1970s, because contrary to the incentives of the system, they lived their lives in accordance with the male breadwinner model.These incentives were perhaps not effective because the gender culture was still traditional and married women expected to be economically inactive.However, we also have to take into account other German family policies which actively directed women towards a traditional role until the 1970s (Moeller 1983).By the late 1970s, however, the gender culture had modernised and it had become more common for married women to combine housework with part-time employment. 7The introduction of care-related rights from 1985 followed this social change.

Pension reform in Germany since 1999
After winning the elections in 1998, the coalition government of the Social Democratic and Green parties immediately started work on a pension reform.The government faced very strong pressure to ensure the financial sustainability of the pension system.Since the early 1990s, with the expense of unification soaring and unemployment figures increasing, cost containment had become a central issue.The reserves of the PAYG [Pay 7 As a consequence of the increase in the labour force participation rates of women, the number of women who received independent rights increased, from nearly 1 million in 1960 to nearly 7 million in 1998, and the number of those who receive survivor's pensions increased from about 2.7 million in 1960 to about 4.2 million in 1998 in West Germany (Veil 2003).
As You Go] system fell below the required levels on repeated occasions, despite earlier reforms.After election victory, the Red-Green government had to act immediately.In this context, a reduction of benefits was almost inevitable, and the Minister of Social Affairs soon released a draft for a reform designed to achieve this aim.This was also able to rely on a new type of public discourse that had developed during the 1990s, which stressed the importance of individual self-reliance, the role of the individual as a 'consumer' and of the need for reciprocity in the relationship between contributions and outcome.In this situation, the modernising family ideology displayed by both the SPD and the Green Party on earlier occasions did not disappear completely, but it had less policy relevance.A further extension of care times was therefore not high on the agenda (Anderson & Meyer 2006).The 'Pension Reform Act 2001' meant a paradigm shift for the system in different respects (Anderson & Meyer 2003, 2004 Langen 2001).Firstly, the explicit goal of status maintenance was abandoned; to control spending, the highest possible public pension level was cut to 64 per cent of last net earnings, and an additional cut of the derived widow's pension was made. 8The latter reduction is compensated for by the extension of care-related pension rights for widows and widowers with children (Standfest 2001).
Secondly, to counterbalance overall losses, two main forms of voluntary pensions subsidised by the state were introduced.Since 2002, employers have been obliged to offer employees the opportunity to pay some of their wages into a funded pension scheme.The employers do not have to make contributions, but are allowed to do so and indeed often do.Alternatively, citizens can choose to save into a personal pension scheme.Initially, women were disadvantaged with regard to personal schemes, because, on the basis of their higher life expectancy, private 8 The new law states that pensions will not sink below 67% of the last net wages until 2030.This is indeed the case, but because calculations will be based on a different formula in 2030, 67% will then only be the equivalent of 64% of today's proportion.
insurers offered different conditions for women and men (Veil 2003).However, after some political pressure from feminist politicians of the Red-Green coalition, trade unions and the European Parliament, which in 2003 agreed on a guideline for unisex conditions in relation to contributions to pension systems (EU Guideline 86/379 EWG), the German government introduced unisex conditions for pension systems, which have taken effect in 2006 (Der Spiegel, 27.4.2004).Occupational and personal pensions are subsidised by the state, occupational pensions through tax breaks and a social insurance contribution rebate, and personal pensions through tax breaks or a flat-rate state sum in addition to a flat-rate sum per child for either parent.Because of tax progression, higher earners are better off with the tax break for personal pensions savings, while lower earners receive the flat rate support.
The third element of the reform was to make it easier for citizens to claim a means-tested pension.This is a variant of social assistance and is for the same amount, with the means test applied to a couple's income and assets, provided they live in one household (BSHG §11).The law is innovative because in contrast to social assistance, children are not obliged to support their parents (AVmG 2000:2), and it has become easier to claim the pension.This third element of the reform goes back to a longstanding discussion amongst members of the left wing of the Social Democratic and Green Parties about the introduction of a universal right to a basic pension, which had started in the 1980s, mainly because it was seen as a means to avoid female poverty in old age, and in particular for single women (Opielka 2002).Derived rights for widows do not exist in this context; they are restricted to the first pillar.
How does the revamped system support the male breadwinner model and is it in line with typical biographies?
With regard to constraints and incentives, the changes have further reduced the level of pension rights that can be accrued through marriage status, and they therefore represent modernisation.At the same time, the cuts in the first pillar lead to a general decline in public pensions, and therefore increased the vulnerability of those not covered by any additional schemes.This is the case for around 40 percent of all employees paying social insurance contributions (TNS Infratest 2005a;2005b).Particularly those in this group with lifetime wages below the average, who receive lower entitlements in an earnings-related system, are likely to suffer increased poverty risks. 9Clearly, women are particularly affected.However, the risks are somewhat muted since it has become easier to claim the minimum means-tested pension.
However, a majority of employees do have access to additional insurance.Coverage of the second-pillar occupational pension and to a very small extent the third-pillar pension increased between 2001 and 2004 by 7 percentage points to 59% of all employees paying social insurance.Women's coverage was slightly higher (60%) than men's (59%) in 2004 (TNS Infratest 2005a;2005b).Projected outcomes show that in general, the new system enforces pension differences according to income, with those with higher public pensions likely to generate higher private pensions, and vice versa.However, the results for women indicate that the pension rights that many of them accrue may be high enough to live independent lives, just above the social assistance line. 10 We therefore see the latest reforms as further modernising the pension system, by continuing to cut derived rights.Moreover, the lowering of the highest possible public pension level weakens anybody's capacity to provide, and citizens with lifetime incomes below average wages (including care-related equivalents) will be particularly at risk of poverty after retirement if they do not pay into additional schemes.How likely is 9 A recent study by the OECD (2005), for example, projects that an individual in Germany, who started work in 2002 and retires at 65 with earnings of half of the average will receive a net public pension worth 37% of average wages, which is below conventional poverty thresholds; the figure rises to 54% for citizens with lifetime earnings of 75% average wages, still below a social inclusion line of 60%. 10 The average projected monthly total pension at retirement for women who were between 18 and 34 years old in 2005 and who retire between 2036 and 2052 would be EUR 1454 in nominal terms (TNS Infratest 2005b:  it that additional pension schemes will be able to fill this gap, especially for the typical female biography?Given the absence of compulsion, participation in schemes depends on the individual's decision to save.The above figures show that a substantial proportion of women have taken this decision, which on average is projected to ensure them an autonomous pension.However, an individual with a biography dominated by part-time work and without second or third pillar entitlements can only lead a financially autonomous life if they are single because they can then claim the means-tested pension.Those who live with a partner with higher entitlements would have no choice but to be dependent on the partner.
Considering that the typical biography in 2000 is characterised by longer periods of full-time work, especially in East Germany, and by an increase in divorce rates, we conclude that the current system no longer enforces the breadwinner model for the majority, and enables those with part-time employment biographies to lead autonomous lives if single.However, for many part-time workers, who live in one household with a spouse, the regime still imposes the moderate breadwinner model.

Public Pensions -the first pillar
In 1946, the basic state pension was introduced, as part of the reconstruction of the British welfare state.Ever since, this pension has been flat-rate and below subsistence level (Hannah 1986).Until the mid-seventies, the system treated married women as dependants.They were not required to pay contributions, and their entitlement was supposed to come through their husbands' insurance contributions.Once they turned 60 and upon their husbands' retirement at 65, they would receive 60% of his state pension (Groves 1983;Land 1985).Those wives who did take up employment during their marriage had the 'Married Woman's Option';11 See Ginn (2001) and Evason & Spence (2003) for an overview of pension reforms since 1975 and their impact on women, and Ginn (2003) for an overview of the structure of the British regime in 2002.
i.e. the opportunity to forgo independent entitlements and not to pay social insurance contributions.This was no real 'option' because women who wanted independent rights had to prove an employment career for at least half of their married years -the 'half test' -and would otherwise lose all benefits gathered, including those as a single person (Groves 1983; for an overview : Ginn 2003).As a consequence of such rules, at the beginning of the 1970s 75% of women did not pay into their own social insurance contributions (Land 1985).
This system lasted until the Social Security Act (1975) was passed (Ginn 2003), during a time of 'social reformism' (Randall 1987).This act was on the one hand a reaction by a Labour government to demands for more equality between men and women (Barr & Coulter 1990), and on the other hand an attempt by the government to address the concerns of blue-collar workers in particular.Workers no longer wished to tolerate the very low level of the basic state pension and demanded pensions that would put them on an equal level with those enjoyed by many other white-collar workers who were members of occupational schemes (Baldwin 1990).The Act phased out the 'Married Woman's Option' and abolished the 'half test '. 12As a rule, all employed women with an income above the 'lower earnings limit' were now asked to pay social insurance contributions 13 ; however, all women who had started paying the married women's stamp were allowed to continue to do so (Groves, 1983;Land 1985).To make up for gaps through care in the household, the 'Home Responsibility Protection' was introduced, which for the first time recognised care times as equivalent to paid work, provided the carer gave up employment for care.HRP reduces the number of qualifying years needed for the Basic State Pensions for carers by 16 years per child, yet care times are limited.The claimant has to have paid contributions for at least half of the qualifying years needed for a full pension, which in most cases is twenty years (Land 1985;Pensions Commission 2004).
12 Except for claimants who had attained pensionable age before 1979.
13 Anyone accruing less than 25% of the qualifying years needed for a full state pension would forfeit all their rights, a continuation of a similar restriction in place since the late 1940s.This rule affected women more than men.
However, even with these improvements, the full pension was still below the subsistence level.This problem was addressed by the introduction of the State Earnings Related Pension (SERPS), a second public tier and another milestone of the 1975 Act.As explained above, this aimed to improve the pensions of blue-collar workers, but it also increased women's independent pension entitlements by giving them access to an earnings-related insurance, withheld for so many of them by occupational pensions (see below) (Ginn 2001).Under SERPS, working women with earnings above the lower earnings limit, regardless of marital status, paid full contributions and their pension levels were calculated in the same way as for men, even though they retired five years earlier and lived longer (Hannah 1986;Groves 1983;Pascall 1986).In addition, SERPS granted a widow's pension (Ginn 2001;Baldwin 1990).SERPS, however, was not allowed to mature.In 1986, a Conservative government decided to dismantle the second public tier because they considered it too heavy a burden for future generations.Possible benefits were lowered and contracting out of the state system and into the private sector was heavily encouraged (Ginn 2003;Pierson 1994).

Occupational Pensions -the second pillar
In the United Kingdom, occupational pensions have always complemented the first pillar, and without it, governments would have been under much more pressure to increase public pension coverage.Furthermore, the expansion of occupational pensions was very much supported and regulated by the state.Third pillar pensions still remain weak today (DWP 2004;GAD 2003a;2003b).Occupational pensions are not directly financed through social insurance contributions, but employers and employees paying into officially recognised occupational schemes pay reduced social insurance contributions and get a generous tax rebate (Groves 1987;Sinfield 2000).In return, pension funds are required by law to provide a certain level of benefits.
Occupational pensions first existed in the civil service in the 19 th century (Groves 1987).Fuelled by the low level of the basic state pension and supported by Conservative and Labour governments alike, they increased soon after 1945 and since the 1960s, they have been a central part of insurance against ageing for all those in Britain who had full-time earnings, stable employment careers, and were members of large private companies or the public sector (Baldwin 1990;GAD 2003a;Ginn 2003).The schemes were selective, and did not cover those in small businesses or with less stable contracts.Until the 1990s, part-time workers were much less likely to be members (Groves 1987), not least because pensions were outside the remit of equal pay and sex-discrimination legislation (Groves 1983).However, successive rulings of the European Court of Justice since 1975, and more specifically since 1986, have led to the equalisation of retirement ages for men and women and improved access of part-time workers to occupational pension schemes; these have come into effect increasingly since 1993 (Honeyball & Shaw 1991;Mazey 1998;Neilson 1998).In 2000, the Part-Time Workers Directive was passed by the European Council, which requires a part-time worker to be treated equally to a 'comparable' full-time worker.However, workers with incomes below the lower earnings limit can still be excluded from occupational pension schemes (PMI 2004).
By the mid-nineties, in the private sector, just under half of all male workers and about a quarter of all female workers had access to occupational schemes, whereas in the public sector, all men and about three quarters of women were admitted (GAD 2002: summary).The most common types were the superior defined-benefit schemes (NAPF 2003).Benefits were earnings-related and often based on final years' salaries.Moreover, after the 1975 Social Security Act, employers were obliged by law to integrate survivors' pensions for spouses into their schemes if they wanted state approval (Groves 1987).Before this obligation was implemented, many occupational pension providers already had in place either widows pensions or lump sum agreements that had the aim of providing for dependants, albeit at variable levels (Groves 1987;1983).
How did the first and second pillar of the system support the male breadwinner model and to what extent was this regime in line with the typical biographies of women?
In contrast to the German system, the British pension system between 1946 and 1975 was typical of a 'strong breadwinner model', because for both pillars, it was assumed that adult women would accrue rights through marriage alone, regardless of their employment activities.In addition to personal dependency, due to very low pension levels, the system also enforced poverty for many.This situation improved for some between 1960 and 1975 because occupational pensions expanded, although this did not weaken the strong breadwinner model because as part-time workers most married women were usually excluded from occupational schemes and therefore only had access to such pensions through their husbands.
The reforms of 1975 rapidly transformed the system towards a moderate breadwinner model, because since then it has recognised not only marriage -the derived rights for married women continue to exist today -but also caring responsibilities and employment for pension right accumulation, thus allowing for women to build up independent pension rights through employment and care (Ginn 2001).Occupational pensions, meanwhile, retained their traditional outlook because they continued providing benefits for widows and still refused part-timers as members.
The nature of this system changed yet again in 1986, through the decline of the earnings-related part of the first pillar, and from then until the late 1990s part-time work received much less favourable treatment, and the SERPS widows pension was halved, to take effect in 2020 (Ginn 2001).This meant an increased poverty risk and consequently a high degree of commodification for all citizens dependent on the first pillar alone, i.e. citizens with instable employment biographies, in low-qualified positions in the private sector and in small businesses, as well as their dependants.We therefore conclude that during this period, the first pillar gave no support to the breadwinner model, through granting insufficient pensions.However, this effect was counterbalanced by occupational schemes, and in particular by the improved recognition of women's typical employment patterns through such schemes since the early 1990s.The constraints and incentives of the first and second pillar of the regime together therefore give moderate support to the breadwinner model because they recognise paid work, care and marriage as a basis for rights accrual.
If we compare these developments with women's typical biographies, we again see that lives diverge from the constraints and incentives of the model (cf.Evason & Spence 2003).In particular, while the system still had all the hallmarks of a strong breadwinner type until 1975, in reality married women had started to seek employment in greater num-bers since the early 1970s (see table 1).In this regard, the situation was characterised by a 'structural lag' (Riley & Riley 1994).This demonstrates the limited power of the pension regime in directing people's decisions, and it highlights the role of other societal factors, such as women's desire to be employed, the availability of jobs, generated not least through the expansion of the welfare state, and a changing gender culture which, as we have shown above, since the late 1960s had ceased to support the male breadwinner model unreservedly.
Designed to be modernising, the reforms of 1975 indeed brought the public regime in line with women's typical lives, characterised by marriage, children, and part-time work, because they recognised not only marriage but also caring responsibilities and employment for independent pension rights accumulation.Yet this period of accord only lasted until the cuts of 1986, when the public system became highly commodifying again.From then on, the lives of married women working part-time were only in line with the system's incentives if they had access to occupational schemes; those who did not generated high poverty risks, and for this reason were at odds with the commodifying nature of the public pensions system.

Pension reforms in the United Kingdom since 1997
Public pensions -the first pillar Since 1997, when the New Labour government came to power, a number of reforms to public pensions have been introduced that transform the way in which rights are accrued.Whilst the low level of the Basic State Pension remained linked to prices, the remnants of the State Earnings Related Pension was replaced with the State Second Pension (S2P), which is more generous than SERPS for low-income groups and better acknowledges caring (Ginn 2003).In addition, surviving dependents, including divorcees who are not re-married, can qualify for a part of this pension. 14This means that the hybrid nature of the system, which recog-nises employment and care as well as marriage, remains, with increases for low-paid women.Despite these improvements, the development will not change the fact that public pensions will remain below subsistence level; calculations suggest that in 2040, when S2P matures, many of its recipients will receive a pension below the level of social assistance either when they retire or very soon after (Rake, Falkingham & Evans, 2000, Ginn 2001;2003).Aware of this problem, the government raised the level of the means-tested social assistance above the poverty line for pensioners only to provide a minimum income threshold (Clark 2001).In contrast to Germany, this is above the social assistance level that nonpensioners receive.Due to this means-tested element, for the first time since the Second World War, all citizens at risk are entitled to a public pension at subsistence level.This has similar shortcomings to its German counterpart, such as disincentives to save and take-up problems, and where relevant, it is applied at the level of a couple, whether married or living together.

Occupational Pensions -the second pillar
Against the background of strong public ideological support for occupational pensions by the government, the real strength of occupational pensions declined considerably.Since the mid-nineties, a substantial share of defined benefit schemes in the private sector has been closed to new members (for an overview of the reasons see Bridgen & Meyer 2005).As a consequence of these closures, coverage by salary-related schemes of workers in companies with more than 20 employees declined from 60 per cent to 50 per cent between 1996 and 1998 (DWP 2002), and from 48 per cent to 41 per cent between 2000 and 2003 (DWP 2004).In a substantial share of businesses, defined benefit schemes have been replaced by the less generously defined contribution schemes.However, for some new workers, the transition means that they will not have access to either (DWP 1998;DWP 2004;OPRA 2004).Despite the government's attempts to strengthen third pillar pensions through stakeholder schemes, their takeup has been very low so far (Pensions Commission 2004).
So far, the public sector has been exempt from this trend, but some expect that before long, public sector schemes will come under pressure to scale down because it will be difficult to justify on the grounds of jus-tice why taxes should finance the much more generous public sector pensions, while all private sector workers are struggling to save enough (Financial Times 21 May 2004) In conclusion, the Labour government's reforms have so far contributed to a preliminary stabilisation of the first pillar and destabilisation of the second pillar of the pension regime.The changes to public pensions reflect the higher commitment of social democratic governments to avoiding poverty.In contrast, the collapse of the second pillar can hardly have been intended, given New Labour's statement in the 1998 Green Paper that "Occupational pensions are (...) arguably the biggest welfare success story of the century" (1998 Cm4179).
When assessing the level of support the current system gives to the breadwinner model, of prime importance is that for the first time public benefits afford a guaranteed minimum level of protection below which nobody is allowed to fall.In addition, the first pillar became more sensitive towards care work and low pay.However, at the same time, the likelihood of becoming eligible for the means-tested pension has increased.Many women working in the private sector and their partners will no longer have a final salary scheme; because of their lower income, this makes women particularly likely candidates for the means test, or it enforces their personal dependence.In contrast, public sector employees have benefited from Labour's much increased investment in public services since 1997, which has meant an expansion of jobs, including an increased number of people in privileged salary-related schemes, many of whom are women (GAD 2003a;IDS 2003).As public sector workers, many women can therefore at least maintain, if not improve their status. 15 The system therefore returned to giving support to the moderate breadwinner model for all social backgrounds, firstly because of its protection against poverty and because care and employment as a basis for independent rights accrual continues to be acknowledged.Secondly, however, given that entitlement to the Pension Credit depends on the means test and can therefore be nullified if a woman has a partner with a 15 Yet a minority of women working in the public sector remain without access to occupational schemes.higher income, policy also still potentially supports personal dependency and therefore the traditional division of labour.
Therefore, the regime protects against the increased risk that contemporary women who lead more individualised lives are exposed to, and it alleviates poverty risks of low-income groups.In this sense, it is in line with many women's lives.At the same time, protection is granted on a fairly low level, and this minimum will, for a substantial share of women, come at the cost of reduced second pillar pension rights.Moreover, women in partnerships may still lose their entitlement to an independent income.
The Labour government recognised that the current pension system still faces serious problems.In particular, it acknowledges that second and third pillar provision is too low and that the protection available to women and carers is inferior.Therefore, in May 2006, a White Paper outlined policy changes that will be the basis for legislation in 2007.The reforms will aim to reduce current poverty risks, especially of carers, and to increase savings incentives for all.To this end, plans are in place to turn the basic state pension into a more substantial foundation for citizens.Depending on public finances, its level could be increased through relinking it with earnings, possibly in 2012.Moreover, citizens could qualify for this pension after 30 years, a sharp cut from the current 39 years for women and 44 years for men.In addition, times spent caring for children or adults in need of care would be recognised without the need to spend a minimum amount of qualifying years in employment.In order to increase private savings, and occupational schemes in particular, from 2012, all employees with annual earnings above a low threshold (£5000 in 2006) would be automatically enrolled in workplace based-savings schemes with compulsory employer contributions.However, they would be allowed to opt out.The reforms would be helped to be financed through a gradual increase in the state pension age from 65 to 68, to be fully in effect by 2046 (DWP 2006).These changes would probably make the basic state pension almost universal; together with better access to second pillar pensions, the reforms could increase the number of people entitled to the full basic state pensions, reduce dependency on the means test, and thus further strengthen women's independence after retirement (see Pension Policy Institute 2006:12-22 for a critical review).However, given that the changes are not envisaged to be phased in immediately, and not before the next elections, it is unclear to what extent they will actually happen.

Conclusions
The aim of this paper was to analyse how the modernisation of the gender arrangements since the 1950s has interrelated with changes to the pension systems in the United Kingdom and Germany.More specifically, we wanted to pinpoint the period during which both countries' pension regimes were fitting building blocks of the strong breadwinner model of the welfare state that is often used to characterise both countries.Secondly, we intended to explore the limits of the stratifying potential of such policies, by comparing their constraints and incentives with the typical life courses of women.In conclusion, we would like to highlight two main points.
Firstly, if we assume that social policy strongly supports the male breadwinner model only when there is little opportunity for women to acquire independent rights, then the German pension system never gave strong support to this model.Indeed during the 1950s, when most citizens faced high risks of poverty, we can even say that no such support existed.In the United Kingdom, strong endorsement prevailed for a long time, between 1946 and 1975.Since then, in both countries, pension programmes have recognised marriage, employment and care responsibilities as a basis for pension rights accrual, and despite some deterioration in Britain during the 1980s and 1990s, they have done so until today, thus only moderately contributing to the male breadwinner model.Our findings therefore suggest that pension regimes have changed their constraints and incentives over time, and that since the mid-1970s they have mostly supported the breadwinner model in a moderate way and at times and for some groups not at all.
What are the implications of our findings, then, for breadwinner model typologies?We have shown how one policy area can have very different effects at different points in time.Moreover, even on a fixed date pensions can lead to significantly different outcomes, depending on the social class and household status of a citizen.Given that something similar is likely to be true not only for pensions but for all policies that constitute a model, it becomes much more difficult and methodologically demanding to assess with any accuracy the nature of a welfare state over a longer period of time.This observation supports Kasza's argument that individual social policies can undergo substantial changes over time, and that there is not necessarily internal consistency between the different policies that are used to characterise models (Kasza 2002).Against this background, there are conceptual and empirical reasons to doubt the long-term accuracy of the breadwinner models.However, because our study is limited to one policy area only, in contrast to Kasza (2002), we are reluctant to consider them obsolete altogether (see also Bambra 2004).
Secondly, we found that the stratifying nature of pension policies for typical biographies does not appear to be as powerful as is sometimes assumed in the regime literature.Despite the opportunities they had, women's typical biographies in Germany, for example, followed traditional patterns until the early 1980s, at a high risk of poverty for working class women (see appendix).In Britain, women joined the labour market in significant numbers from the late 1960s, even though such behaviour did not earn them any pension rights.The situation was, in this respect, characterised by a 'structural lag' (Riley & Riley 1994).As Pascall (1999) put it: "UK women joined the labour market, long before there were any policies that supported their roles as employees."(261) Yet in the 1980s, when the pension system re-commodified many of them, they still lived lives in accordance with the full-time male breadwinner/female parttime carer model.Such discrepancies highlight the relative independence of societal modernisation processes from social policy reform, and they throw into sharp relief the significance of gender cultures.However, while we would like to argue that more attention should be paid to such processes, the discrepancies also point to the comparatively greater significance of social policies other than pensions for the male breadwinner model.We assume that policy interventions at the family-starting and child-bearing stage of the life cycle make a greater difference for women's long-term decisions, for example the availability of affordable childcare services, the existence of tax incentives in favour or discouragement of a single-earner household, or the existence of parental leave arrangements that allow for flexibility.When exploring how strongly a welfare state supports the male breadwinner model, it may therefore be useful to pri-oritise certain policies over others.On the basis of our analysis, we argue that pensions should not have priority.

1
For a discussion see Pfau-Effinger 2005 a, b.
). Part-time work 5 For Western Germany, see Geissler and Oechsle 1996; for the United Kingdom, see Fox Harding 1996, Daune-Richard 1998.Even if this model has the greatest significance in these countries today, this does not preclude that other gender-cultural models still exist and influence behaviour.in Britain and Germany generally means half-day employment or less, and the average number of weekly working hours is 18 in Germany and 19 in the UK (OECD 2005).Correspondingly, British and German households are the most likely in the EU 15 to care for children and other dependants at home rather than use paid services (European Commission 1998).

Table 2 .
(Pfau-Effinger & Geissler 2002)e family in Germany and the UK This is also indicated by the attitude data of the ISSP in table 2. The share of respondents who support a traditional gender division of labour and who think that the employment of mothers of pre-school children is detrimental for their children's well-being is much lower in East Germany than in West Germany and the United Kingdom.Even though the labour market in East Germany is still in crisis, women try to pursue continuous full-time employment(Pfau-Effinger & Geissler 2002).East German women with young children are more likely to work full-time than West Source: ISSP 2002: Family and Changing Gender Roles.All of these factors contributed to the differences in the level of labour force participation rates in West Germany and the United Kingdom, which are shown in table 1. Accordingly, these factors also suggest that the gender regime of West Germany is still more traditional than that of the United Kingdom.In East Germany, however, a different cultural model has prevailed until today.The dissolution of the traditional family form was already widespread before re-unification (Busch, Nauck & Nave-Herz 1999) and a dual breadwinner/state carer model was the dominant cultural idea in the former German Democratic Republic (GDR), based on the assumption that all citizens participate in full-time waged work and that children are mainly cared for by the state.