Gender, pensions and the lifecourse "If insufficient time and resources are devoted to [unpaid care], productivity will suffer as human resources deteriorate and the social fabric is inadequately maintained" (Himmelweit, 1998, p 7). If the role of mothers in caring for their children is valued as work, this implies protecting pension entitlements during periods of childcare, as is the case in state pensions in Britain and in many other European countries (see Chapter Six). However, the dominance of private pensions and meagre state pensions in Britain means that the proportion of pension entitlements protected in this way is small. Increasing policy emphasis on private pensions reinforces the pension penalties of motherhood. On the other hand, a basic state pension set at an adequate level could minimise such losses for working age women, as well as lifting out of poverty those older women whose pension building has been restricted by family caring commitments. In Chapter Six, the way other EU countries meet the pension needs of those with childcare responsibilities is outlined. 82 SIX .Gender and pensions in the European Union: towards an independence model? Key concerns of policy makers throughout the developed world are to ensure financial sustainability of state pension schemes without sacrificing adequacy of pensions to meet the needs of older people and future pensioners. Population ageing, which is partly due to falling fertility rates, has brought the question of sustainability of pensions into prominence, although whether this constitutes a serious problem, and in which countries, has been subject to debate (Walker, 1990; Mullan, 2000; Street and Ginn, 2001). Since the employment rate in the working age population is also important to the viability of any pension scheme (state or private) it is useful to consider the relationship between women's fertility and their employment. Is there an inevitable trade-off, for women, between social reproduction and economic production? The chapter first considers the social value of unpaid care work and compares women's fertility and their employment across EU countries. Britain's pension system is then compared with those of other EU countries in terms of adequacy, gender inequality of later life income and the treatment of family caring. Finally, the question of moving towards an independence model of pensions is discussed. The gendered lifecourse and social reproduction The research reported in Chapters Four and Five focused on the impact of women's childcare commitments on their employment and pensions in the context of a lack of affordable childcare services in Britain. The government recognise that mothers find childcare a barrier to employment: "Women tell us that a key obstacle to their re-entering the labour market is accessing affordable and good quality childcare" (DWP, 2002, p 117), although it is not only childcare that conflicts with women's employment participation. Women over age 50 are often called upon to provide informal care for ageing parents or parents-in-law or to look after their grandchildren while their daughters or daughters-in-law engage in paid work. The provision of family caring and consequences for employment and pension building are usually seen as problems solely for women. For practical reasons, earlier chapters of this book are based on the assumption that working age carers are usually women, neglecting those men who also devote time to family caring at the expense of their employment, earnings and pension rights. Given 83 Gender, pensions and the lifecourse the increasing fluidity of family forms, it is timely to recall that gender denotes socially constructed roles and identities, and is not merely an alternative, politically correct, term for sex. In western countries a minority of women, mainly those who are childless or who are highly qualified partnered mothers, pursue a typically'masculine'lifecourse in terms of continuity of full-time employment. Equally, a minority of men undertake caring tasks, either as lone fathers or through role reversal in which their partner is the main breadwinner while they interrupt their employment, following a typically 'feminine' lifecourse. Thus gender, in the limited but important sense of the distribution of paid and unpaid work roles over the lifecourse, is becoming less closely linked to sex and represents a spectrum of lifecourse patterns rather than a dichotomy. In the gendered lifecourse, the variation is between a more feminine or a more masculine pattern, not simply between men and women. Unpaid family caring work is often conceived as a private matter of voluntary labour within the family. Yet it is crucial to the welfare and even survival of society — a point which is obvious but has generally been overlooked by economists. First, the care needs of the young, the sick and frail elderly individuals are unlikely to diminish. Parents' investment of time and energy in bearing, nurturing and socialising children, like informal care for older people, is labour-intensive, time-dependent, emotionally demanding and person-specific. Although marketised or statutory services are an important support to parents and other carers, they do not fully replace family care. The value of unpaid domestic labour to the economy (including other domestic work as well as caring) is gaining recognition among statisticians, with plans to measure and record unpaid work in national statistics. For example, the value of women's unpaid domestic work was estimated as £259 billion or 42% of GDP (calculated from Murgatroyd and Neuburger, 1997) — not a trivial contribution and one that reduces women's time for paid employment as well as engagement in civic, political and social life. Second, falling birth rates have serious implications for societies, especially when fertility remains below the replacement rate (2.1 children per woman) over an extended period, an unprecedented development that is occurring in most OECD countries (Coleman, 2000). In the EU as a whole, the average total fertility rate in the 15 EU countries has nearly halved in 33 years, falling from 2.7 in 1960 to 1.4 in 1993. The shortfall in the birth rate, and the decline between successive birth cohorts of women, varies across countries (see Figure 6.1). In all EU countries except Ireland, women born in 1963 did not replace their generation - although the outcome is not yet entirely determined, due to the trend to later childbearing (Beets, 1999). Ireland, with the Mediterranean countries, had the steepest decline in fertility, while Denmark and Sweden's were stable and Luxembourg and Finland even showed a slight increase in fertility between the two cohorts. The reasons for fertility trends over time and for cross-country variation are debated among demographers but are thought to relate less to cash benefits for parenthood than to other factors such as the domestic division of labour and 84 Gender and pensions in the European Union Figure 6.1: Completed fertility, generations of women born in 1950 and S 963 (EU 15) 3,5 2,5 1 D Generation born 1950 s Generation born 1963 ru r» fl EUI5 IRE F S Fl DK UK B P L EL NL A E D Note: For key to the EU 15 countries, please see Table 6.3. Source: Eurostat, Demographic Statistics, cited in Bagavos and Martin (2000) family policies that enable women to combine employment and childrearing (Bagavos and Martin, 2000). Women in highly marketised societies need to earn an independent income and, in the absence of suitable policies to reconcile childrearing and employment, this creates pressure to delay or forgo childbearing: "The risk-averse woman of today will ensure that she is able to support herself and, given the chance of relationship breakdown, will be careful not to put herself in a position of dependency on a man" (McDonald, 2000). However, in countries with social policies to support childrearing — including adequate paid maternity and parental leave, availability of quality, subsidised daycare for children, after-school care, individual taxation and opportunities to shorten hours of work temporarily without changing jobs — the conflict between women's roles in reproduction and in production is less. Finland exemplifies such policies (Taskinen, 2000). It has been suggested that "fertility is higher in the member states where: the caring activities are better shared between men and -women, public caring infrastructure is more developed, part-time jobs are more available, legislation is more family-friendly" (Eurostat, 2001, p 28). In the next section, trends in women's employment are compared among EU countries. Comparing women's employment in EU countries In the EU as a whole, 37% of working-age women are employed full time, with 14% working part time (under 30 hours a week). The figures for men are 67% and 4% respectively (OECD, 2001b). Women's rates of full-time 85 Gender, pensions and the lifecourse employment range from the lowest levels in the Netherlands, Italy and Spain (27-30%), through medium levels in Britain, Ireland, Germany, France and Switzerland (36-41%) to the highest in Denmark and Sweden (55-57%). Full-time employment rates of men and women are shown for eight EU countries, from 1990 to 2000, in Figure 6.2 (and Appendix Table A6.1). In 2000, a relatively high proportion of women (50-60%) were employed full time in Denmark and Sweden. This also applies in Finland (not shown). Italy and the Netherlands had low rates of full-time employment of women, under 30%, while the range was 35-41% in France, Britain, Germany and Ireland. Over the decade, the proportion of women employed full time rose substantially in Denmark, Ireland and the Netherlands, although from a low base in these latter two countries. The increase was very slight in France, Germany and Britain, while no rise was evident in Sweden and Italy. Part-time employment of women rose in all countries except Denmark and Sweden, while that of men rose slightly in seven of the countries (Table A 6.1). Figure 6.2: Percentage employed full time, men and women in eight EU countries (1990-2000) ® Sweden ™ Denmark O France fcs Britain ^ Germany ° Ireland ~-1— Italy * Netherlands Notes: As a % of population aged 15-64 (16-64 in Britain and Sweden). Break in series after 1990 for Sweden, Denmark, Germany and Italy. Source: Calculated from OECD (2000,2001,Tables A and E) 86 Gender and pensions in the European Union Women's employment participation is related to their maternal status but the extent to which women reduce their hours of employment after giving birth varies among countries (Blossfeld and Hakim, 1997). This highlights the role of social structure, especially childcare provision and school hours, in influencing mothers' employment participation. The constraint on mothers' hours of work is more severe in Britain than in those European countries where affordable quality childcare services are widely available (Joshi and Davies, 1992). According to Esping-Andersen (1999), the net cost of childcare for two children under the age of three, as a proportion of average family income, was 9% in France, 11% in Denmark and the US, 16% in Sweden, 28% in the United Kingdom and 39% in Italy. The UK has the lowest level of publicly funded childcare provision for children aged under three in the EU and relatively short paid maternity and parental leave (Esping-Andersen, 2002). The effect of motherhood on employment was demonstrated by Harkness andWaldfogel (1999) using Luxembourg Income Study data for women aged 24-44 in seven countries including Britain, Germany, Finland and Sweden. Mothers' employment participation was substantially less than that of childless women in each country, but the effect of children was greatest in Britain and negligible in Sweden. The 'family gap' was even more striking for full-time employment. For example, 76% of British women without children were employed full time compared with only 26% of women who had children, while the equivalent rates were 72 and 40% in Germany, 75 and 61% in Sweden and 79 and 66% in Finland (Harkness andWaldfogel, 1999). The pay penalty of motherhood also varies among EU countries. Harkness andWaldfogel (1999) found that the 'family gap in pay' for women aged 24-44 (after taking account of age, education and other relevant variables) was substantial in Britain, rising from 8% for one child to 24% for two children. The wage penalty for two children was much less in other countries - 11% in Germany, 3% in Finland and only 1% in Sweden. These results are consistent with the finding of Gornick et al (1997) that social policies in three social democratic welfare states (Denmark, Sweden and Finland) facilitate women's dual roles in reproduction and production, whereas in three Anglophone welfare states (Britain, Australia and the US) women "faced stiff penalties for work absence due to childbirth; without publicly subsidised childcare many had few viable alternatives to full-time maternal care of children" (Gornick et al, 1997, p 64). The quality of part-time jobs and the number of hours typically worked influence the impact of motherhood on pay, and these vary among EU countries (O'Reilly and Fagan, 1998). For example, in 1988, the average weekly hours worked by part-timers were 25 in Sweden but only 17 in Britain (Sainsbury, 1996,Table 5.3). Where hours are longer, part-time employment is less strongly linked to low pay, low occupational status and poor long-term prospects (Cousins, 1994). If part-time employment is possible in professional occupations and is used by employers as a way of retaining valued staff, it is more a bridge than a trap for women. For example, Swedish and Finnish mothers of young children have the statutory right to reduce their working week to 30 hours without loss 87 Gender, pensions and the lifecourse of occupational status or hourly rate of pay. International variation in the quality of part-time jobs suggests that, although market forces tend to push the most vulnerable workers into the poorest jobs, state intervention can improve the quality of part-time employment. Despite British legislation in 2000, high-quality part-time jobs were still the exception rather than the rule in 2002 and hourly pay was much less than for full-timers. Where a lack of family-friendly policies contributes to women typically having reduced lifetime earnings, the pension penalties vary with the structure of a country's pension system, especially the extent to which pensions are tied to earnings over the working life (Ginn and Arber, 1992). A key issue is how far EU pension systems have incorporated arrangements to enable those with family caring commitments to build adequate pension entitlements of their own (an independence model of pension provision) or instead rely mainly on carers receiving entitlements derived from their marital status (a male breadwinner model). The next section compares both the adequacy of EU pensions and their adaptations to fit the dual roles of women's paid and unpaid work - their carer-friendliness. Comparing EU pension systems: adequacy and carer-friendliness Adequacy The European Council at its Laeken meeting in 2001 recorded agreement by member states on 11 objectives for pension systems, under the headings of adequacy, sustainability and modernisation to meet changing societal needs (EC, 2002). In terms of adequacy, member states are enjoined to: 1. ensure that older people are not placed at risk of poverty and can enjoy a decent standard of living; that they share in the economic well-being of their country and can accordingly participate actively in public, social and cultural life; and 2. provide access for all individuals to appropriate pension arrangements, public and/or private, which allow them to earn pension entitlements enabling them to maintain, to a reasonable degree, their living standard after retirement. Older people in Britain are at a higher risk of poverty than in many other EU countries, according to the most recent poverty statistics for the EU. The poverty rate for individuals aged over 65 in 1998, using the EU definition of income below 60% of national median equivalised income, was 21% in Britain, compared with 11% in Germany 8% in Sweden and 7% in the Netherlands (Figure 6.3 and Appendix Table A6.2a). British pensioners' median income was only 78% of that of the population aged under 65. This compares with a relative income for older people of over 90% in Germany, France, the Netherlands, Italy and Luxembourg (Figure 6.3 and Appendix Table A6.2b). 88 Gender and pensions in the European Union Figure 6.3: Poverty rates3 and relative income13 among those aged 65+(EU IS) 100 80- £ 60 40 20 ■ Relative income □ Poverty rate t EU 15 NL L Dl E FIN F UK B A DK P EL IRE Notes:' Percentage aged 65+ with income less than 60% of the median equivalised income of the national population.b Median equivalised income of those aged 65+ as percentage of that of the population aged 0- 64 For key to the EU 15 countries, please see Table 6.3. Source: CEC 2003,Table 2, based on European Community Household Panel Survey data for 1998 Earlier figures produced by Eurostat (2002) showed Britain to have the highest pensioner poverty rate among EU countries at 40%. However, these figures were reworked during 2002 at the insistence of the British government. Not surprisingly, poverty measures are highly controversial. Comparisons of poverty across countries must be treated with caution due to inter-country variation in charges for services such as healthcare and in provision of other non-cash benefits. Statistics on poverty rates are affected by unreliable reporting of incomes and missing income data. Poverty rates are also sensitive to the assumptions made. In particular, the EU poverty measures, by taking combined household income and then assigning each individual an equal share adjusted for household size (that is, equivalising income), effectively underestimate the prevalence of personal poverty among women who live with a partner or other relative. An alternative way of defining poverty (instead of relative to national population average) is in relation to the income required to pay for goods and services at a basic or a modest level. Research in Britain by the Family Budget Unit calculated the weekly income required to provide a basic - Low Cost but Acceptable (LCA) - living standard for people aged 65-74, in 1999 prices (Parker, 2000). The LCA amounts, uprated for earnings rises between 1999 and 2002, are shown in Table 6.1. For those aged over 75, the majority of whom are women, the amounts required would be higher to take account of disability and other health problems. Over two million pensioners (mainly non-married women) are poor enough to receive the MIG and it is estimated that a further half a million more are 89 Gender, pensions and the lifecourse Table 6.1: Comparison of basic state pension and MIG with LCA amounts for people aged 65-74 (2002, £ per week) Full basic pension MIG LCA amounts, LA tenants uprated to 2002 Homeowners Non-married person 75.50 98.15 141.00 115.00 Married couple" 120.80 149.80 210.00 170.00 Note: 'Assuming husband has full pension and wife has 60%. Cohabiting women have no entitlement to a derived pension based on their partner's contribution record, but cohabiting and married couples receive the same MIG. entitled but fail to claim (DSS, 1999). Despite government attempts to increase take-up of the MIG, this has proved an intractable problem (Corden, 1999; Evason et al, 2002). Since the LCA standard is substantially higher than the MIG, many more than 2.5 million pensioners are living below the LCA standard. According to the government's own figures, the proportion of pensioners at risk of poverty in the year 2000 was 31% of women and 25% of men, while the proportion among pensioner women living alone was as high as 46% (DSS, 2000). Whichever measure of poverty is used, it is clear that a substantial proportion of British pensioners, mainly women, lack sufficient income to "participate actively in public, social and cultural life" and to "maintain, to a reasonable degree, their living standard after retirement". Despite using means testing to supplement meagre state pensions, British pension policy fails to satisfy EU objectives on adequacy and performs poorly compared with the EU as a whole. It is debatable whether this situation is acceptable in a rich country. Carer-friendliness Older women bear a disproportionate share of poverty throughout the EU, having lower personal incomes than older men, due mainly to smaller pensions (Dooghe and Appleton, 1995; Walker and Maltby, 1997). However, the magnitude of women's pension disadvantage varies across countries. As Walker and Maltby (1997) note, in the 1990s older women's pension income as a proportion of men's was approximately 66% in Italy, 56% in France and 42% in Germany. In Britain it was about 60%. This does not imply that older women in Germany are poorer than their British counterparts, but that gender inequality is more acute. In Denmark gender inequality of pension income was much less than in the rest of the EU (Walker et al, 1993), due to the dominant role of the residence-based citizen's pension. Pension systems can be compared in terms of carer-friendliness - incorporation of features that help to redistribute towards those whose employment has been restricted by their family caring commitments. These features may be included in any of the three major sources of pension entitlements - state pension 90 Gender and pensions in the European Union entitlements based on the individual's paid employment and unpaid caring work; state pension entitlements derived from relationship to a husband (alive and still married, divorced or deceased); and private pensions, including both occupational (DB or DC) and personal pensions (seeTable 6.2). In contributory state pension schemes, gender differences in access to the pension and likely amount of entitlement depend on a number of design features of the pension scheme, such as whether there are earnings or hours thresholds for access, whether there is some form of carer credit and how the amount of pension is determined (Table 6.2a). Citizen's pensions, although often referred to as Table é.2: Key components of pension systems from a gender perspective a) State pensions (individual entitlement) Access Earnings or hours thresholds for contributions Years threshold for residence-based pension The treatment of years of caring Age for pension qualification Amount Whether and how pensions are linked to earnings The existence of minimum floors and/or ceilings Maximum amount as % of average earnings Duration of contribution/residence period for full pension b) State pensions (entitlements as dependant) Access Married or widowed status required for eligibility Provision for pension splitting on divorce Amount Whether dependant's entitlement equals that of contributing spouse % allocated to wife, if unequal entitlement % 'inherited' by widow Arrangements for divorcees c) Private pensions Access Balance of state and private pension provision Private pensions mandatory or not Earnings or duration of service threshold for eligibility Ease of transfer or preservation Amount Defined contribution or defined benefit Level of employer's mandatory contribution, if any Whether entitlement is inflation-proofed during preservation Whether pension in payment is inflation-proofed Level of tax relief on contributions, if any Widow's pension as % of deceased member's pension Source: Ginn et al (2001, chapter I) 91 Gender, pensions and the lifecourse universal, generally reduce the pension for those without a long period of residence. State pensions derived from a spouse's contribution record vary in the years of marriage required for entitlement, in the proportion that a spouse is entitled to and in the pension rights of divorced spouses (Table 6.2b). Private pensions vary in terms of the extent to which part-timers and the low paid are excluded from access, in arrangements for preserving and transferring entitlements for those who have gaps in employment and job changes, in the level of the contribution required from the employer and whether the pension is a defined amount (Table 6.2c). An important issue for women is the share of state and private pension provision, since state pensions are potentially more adaptable to carers' needs. Whereas private pensions provide only derived benefits to help carers, most state pensions in the EU have included adaptations that provide some protection for carers' own pensions (see Table 6.3 and Leitner, 2001). This development is most welcome, marking progress towards the independence model of pension provision, in which women (and carers generally) are assisted in building an adequate pension of their own, rather than relying on derived pensions. Later in this chapter, the role of derived pensions is considered more fully. Among countries with a flat-rate state pension, Britain and Ireland allow those with periods of caring to accrue entitlements during that period (Table 6.3a), but carer protection has only applied since 2002 in the British second tier pension. The Netherlands and Denmark ensure that those who have spent years providing care receive the same basic pension as those who have been continuously employed, by paying a tax-funded 'citizen's pension' to all those fulfilling residence requirements. These universal pensions are payable to men and women in the Netherlands at age 65 and in Denmark at 67, while state pensions are payable in Ireland at 66 (65 if retired) and at 60 for women and 65 for men in Britain (rising to 65 for women by 2020). Unfortunately, neither carer protection nor a citizen's pension is effective in preventing poverty if the level of the pension is very low, as in Britain. The British basic pension compares very unfavourably with other flat-rate pensions. For example, in 1998 Ireland's basic pension for a lone pensioner was 29% of average industrial earnings (equivalent to ,£91 per week, compared with Britain's J~ß7), and it is being increased faster than prices (Daly, 2001). In the Netherlands the basic pension for a lone pensioner in 2002 was 203 Euros (about £130) per week plus a holiday allowance, and the pension is indexed to the minimum wage. For an example outside the EU, New Zealanders receive a tax-funded citizen's pension at age 65 set at 34% of average net earnings for each married person and 44% for lone pensioners. In 1998 this was equivalent to over £100 per week for a lone pensioner (using purchasing power parities). As a result, in New Zealand men's and women's incomes in retirement are roughly equal (St John and Gran, 2001). One reason New Zealand has the resources to provide this citizen's pension is that there is no tax relief on private pensions, in contrast to Britain. 92 Gender and pensions in the European Union Table 6.3: Allowances for childcare or eldercare in state pension schemes (ELSIS) a) Flat-rate schemes Britain (UK) Home Responsibilities Protection in basic pension for those with children up to 16 (or 18 in full-time education) and for those providing substantial care for adults Ireland (IRE) Homemakers Scheme, as above Denmark (DK) Pension for all residents, tax-funded Netherlands (NL) Pension for all residents, tax-funded b) Earnings-related schemes Finland (Fl) Coverage for recipients of home care allowance Sweden (S) Four years coverage for each child Austria (A) Four years coverage for each child Germany (D) Three years coverage for each child; informal carers covered, benefit depending on hours of care provided France (F) Two years coverage for each child; bonus for mothers of three children Belgium (BE) Two years coverage for each child, based on individual's last wage Portugal (P) Two years coverage for each child Luxembourg (L) Two years coverage for each child Spain (E) One year coverage for each child Italy (I) Six months coverage for each child; one month/year coverage for informal carers Greece (EL) Three to six months coverage for each child DenmarkiATP scheme None, but pension related to hours not earnings Britain; SERPS/S2P schemes None in SERPS, but its replacement - the State Second Pension - includes coverage until the youngest child is aged 6 Source: Adapted from Leitner (2001, tables 4 and 5) Turning to earnings-related state pensions, the 11 EU countries without a basic flat-rate pension all provide some allowance for pension accrual by those with family caring commitments although they vary in generosity (Table 6.3b). In earnings-related schemes, the value of a pension credit depends on the notional contribution rate applied for the period covered, which may be a fraction of national average earnings or of the individual's recent earnings. The British S2P provides carer credits based on notional earnings of about £10,000 pa. Some pension disadvantage due to raising children is likely to remain for those returning to employment after their 'carer coverage' (years when caring allowances apply) runs out, if continuing childcare responsibilities prevent them from obtaining full-time employment in a similar job to their previous one. Another relevant factor is the number of years used in calculating the average earnings on which the pension entitlement is based (Leitner, 2001). Half of the EU countries which have an earnings-related scheme use lifetime earnings 93 Gender, pensions and the lifecourse (Belgium, Germany, Italy, Luxembourg, Sweden, Britain); in these countries periods of no/low pay will tend to reduce the average earnings on which the pension is based, unless those periods are entirely covered by pension credits. The remainder use average earnings during the last year or last few years (Finland, Greece, Spain) or the average in the best years (Austria, France, Portugal). Since women's earnings are not necessarily highest towards the end of their working life (as is often the case for men, especially in non-manual occupations), use of best years is more helpful to women than use of last years. Income replacement rates for those with a typically feminine lifecourse are a useful way of comparing the carer-friendliness of state pensions. Replacement rates at retirement are shown in Table 6.4, for non-married pensioners in five EU countries in the early 1990s (Eurostat, 1993). Those with a short employment history and former wages two thirds of average had net replacement rates ranging from 82% in Denmark to less than a third in Germany and Britain (see Table 6.4, column 4, shaded). In contrast, those with average wages and a full employment record had replacement rates of 60% in Denmark, 77% in Germany and 44% in Britain (Table 6.4, column 2). In terms of replacement rate, the state pension disadvantage due to a typically feminine lifecourse was most severe in Germany and Spain, followed by Britain, while redistribution towards those with such a lifecourse was strongest in Denmark and the Netherlands. Thus state pensions vary across countries in terms of both overall generosity and the extent to which those with family caring responsibilities are disadvantaged. However, all state pensions are more carer-friendly than private pensions. Private pensions, despite subsidies from public funds through tax relief in most countries, include no provision for periods of caring. Hence, although they formally treat men and women equally, these pensions indirectly discriminate against women, who are the main providers of unpaid family care. Individuals lacking continuous full-time employment are inevitably placed at a disadvantage. Private pensions are becoming more widespread in the EU, while Table 6.4: Replacement ratesa for non-married pensioner by qualification period and former wages Full pension Partial pension (20 yrs) Former wages 2/3 100% 200% 2/3 100% 200% Spain 98 97 97 70 73 71 Germany 72 77 63 32 34 28 Denmark 83 60 37 82 59 36 Netherlands 66 49 27 66 49 27 UK 53 44 30 31 28 23 Notes:a Net, at retirement, for those receiving the state retirement pension. Shaded figures represent likely outcomes for women with interrupted employment. Source: Eurostat (1993) 94 Gender and pensions in the European Union social insurance schemes are being cut back, modestly in some countries, severely in others such as Britain (Clasen, 1997). The question of enabling preservation and transfer of occupational pension rights across EU borders for mobile workers was being addressed by an EC consultation in 2003, although the different tax regimes in EU member states present formidable obstacles. The unequal benefits from private pensions according to gender and class have been shown in earlier chapters and in Ginn and Arber (1991, 1999). Therefore a pension system in which state pensions play the major role is essential if progress is to be made towards an independence model of pensions. Towards an Independence model of pensions Across Europe family structures are changing, with declining marriage rates and rising rates of cohabitation, divorce, lone motherhood and births outside marriage (Dykstra, 2003). The trend towards women raising children independently of a partner, which can be seen as indicating greater autonomy of women as they are freed from traditional family bonds, is strongest in the Nordic welfare states (Denmark, Sweden, Norway and Finland), France and Britain, but is relatively weak in the familialistic Mediterranean countries (Esping-Andersen, 1996). The rises in divorce, lone parenthood and cohabitation, although less striking in the EU as a whole than in Britain, are unlikely to be reversed, and the decline of marriage as a lifelong contract makes reliance on a husband for an income in later life an ever more risky strategy for women. The unravelling of the ties between marriage and motherhood makes a reassessment of derived benefits urgent. As argued in Chapter Four, survivor benefits are no longer an effective means, if they ever were, of compensating carers for their pension losses. Such benefits serve mainly as a subsidy to marriage, financed by carers and non-carers alike, even though many married couples are childless. Thus the rationale for providing subsidies to non-contributors purely on the basis of their marital status is becoming increasingly dubious. Risk-pooling and redistribution are important and valuable functions of social insurance. Yet pension scheme subsidies flowing from non-married mothers to childless married women raise questions of both equity and efficiency. In a context of serial partnerships across the lifecourse, administering derived benefits is becoming increasingly complex and inefficient for both state agencies and private pension operators. Depending on derived benefits is increasingly fraught with risk for women in the EU (Jepsen and Meulders, 2002). Yet derived benefits cannot be abolished, or even phased out, until suitable alternative means are in place so that those with a typically feminine lifecourse are able to build adequate pensions of their own. The European Commission, through its Social Protection Committee, urges member states to ensure equal treatment of men and women in pensions legislation, including applying equal actuarial factors to pensions (EC, 2002). Recognising women's pension disadvantages due to their family roles, they 95 Gender, pensions and the lifecourse also advocate pension credits for childrearing and splitting of pension rights between spouses. However, the EC has not yet explicitly endorsed the independence inodel of pensions — often referred to as the individualisation of social security rights — as a desirable objective. Doing so would risk opposition from several member states. Yet such a lead from the EC would support the campaigns by pensioners and women's organisations to improve women's prospects of an adequate independent income in later life. Conclusions Comparison of EU countries shows that social and employment policies, exemplified in the Nordic welfare states, can do much to help reconcile women's dual roles in reproducing society and in paid employment. If a trade-off between women's social reproduction and economic production can be avoided, or at least minimised, the affordability of state pension schemes is improved by their social insurance contributions and taxes. Women's chance of an adequate income of their own in retirement depends on both the overall generosity of their country's state pension schemes and the extent to which schemes are carer-friendly — fitted to the needs of those with caring commitments. Among EU countries, Britain has relatively high poverty rates and gender inequality, while the Netherlands stands out as having low poverty rates that are equal for men and women and high pensioner incomes compared with the population as a whole. Significantly, both these countries have well-developed occupational pensions, but the more favourable outcome in the Netherlands may be attributed to its relatively generous universal pension. Trends in parenting and partnering across the EU point to the need for an independence model of pensions, in which women can receive an adequate income in later life irrespective of their marital status. Yet progress towards phasing out derived pension benefits depends on improving state pensions in terms of their carer-friendliness. In Chapter Seven, I turn to pensions policy, particularly the question of how the tangled British pension system could be better adapted to women's working lives. 96 SEVEN British pension policy: a gender perspective on alternative rescue plans Few outside government deny that the British pension system is now in crisis, although the exact nature of that crisis is variously defined and gender issues are rarely considered. Critical assessments of current policy abound in the evidence submitted to the Work and Pensions Select Committee's inquiry (Work and Pensions Committee, 2002), from pension providers and advisers (over 25 submissions), trades unions including the TUC (10), employer organisations including the Confederation of British Industry (2), pensioner organisations including the National Pensioners Convention (7), major pensioner charities (2), think tanks (4), consumer groups, academics and other individuals. In this chapter the concerns expressed by a range of such commentators are outlined. The reasons why reaching a pensions consensus has proved so difficult are discussed and the main types of rescue plan proposed are considered, focusing on the gender implications of each. Concerns can be grouped into three main areas — pensioner poverty, erosion of saving incentives and the crisis of private pensions. Pensioner poverty Member countries of the EU have agreed that pensions should be adequate to ensure that older people can enjoy a "decent standard of living" (EC, 2002, p 5) but the British government has not said what income this would require. Many experts believe that neither the basic state pension nor the means-tested Minimum Income Guarantee (MIG) is adequate, and that the levels of pensioner poverty in Britain, especially among women, are unacceptable (see Chapter Six). Cuts in state pensions are set to continue, so that wage replacement rates will decline even further in future and the pensions poverty trap will widen. The government's answer is to expand the system of means-tested benefits through a Pension Credit from October 2003. This will introduce a taper into means testing so that 40% of income above the full basic pension is lost instead of 100% as at present, helping those with incomes slightly above the MIG but also drawing over half of pensioners into means-testing in 2003 and an even larger proportion in future. The government claims the Pension Credit will help women but many older women on low incomes will miss out: 97