ONE • Trends in gender relations* employment and pensions As the main source of income in later life, pensions can make the difference between a daily struggle and a period of fulfilment, between merely surviving and actively thriving. At the start of the twenty-first century, life in retirement has lengthened, with people living an average of over 20 years beyond the state pension age. With longer lives, an adequate pension income is more necessary than ever before. Women's savings have to last longer than men's, as they can currently expect to live another 22 years after age 60 compared with 18 for men (OECD, 2001a). There is a widespread sense of crisis and uncertainty over the future of pensions in Britain. While state pensions are apparently in terminal decline due to the policy of successive governments, the risks of reliance on the private sector of pensions are increasingly evident. Plummeting projections for investment returns and a question mark over the continuation of final salary occupational pension schemes follow a saga of mis-selling in personal pensions. Expressing widespread concern, one financial expert concludes that "The much-trumpeted UK model of private pensions funding looks distressingly shaky" (Plender, 2003, p 1). As a result, many people banking on their private pensions are threatened with an impoverished old age, or a much lower income than they expected. Both the government and the private pensions industry urge people to save even more in private pension schemes, to make up the shortfall. However, the ability to do this is gendered, depending on employment and earnings, and also varies ■with stage in the lifecourse. Far-reaching social changes have transformed gender relationships and norms as to partnering and parenting. Major changes include growing expectations of women's equality and financial independence irrespective of marital status; the sexual revolution, giving women greater control over their fertility; and increasing recognition of the importance of women's skills in the workplace and the economy as a whole. At the same time, the expectation of lifelong marriage has declined, with increases in divorce, cohabitation and lone parenthood as all these have become more socially acceptable. These changes offer both opportunities and challenges to individuals and society and have profound implications for women's acquisition of pensions. Yet successive generations encounter (and contribute to) these changes at different stages in their lifecourse, with consequently different effects on their future options. For example, cohorts of women reaching adulthood in the 1960s were able to make choices concerning family planning and employment rarely available to Gender, pensions and the lifecourse those reaching adulthood in the 1940s. A later generation, entering the workforce in the 1980s, benefited from the equality legislation of the 1970s and from the ideas of second wave feminism that permeated society. In the same way, as the British pension system evolves, each generation accumulates pension entitlements under new conditions, applying at particular stages of their life. Whereas the current older generation, and many now in midlife, expected state pensions to form a solid bedrock of retirement income (supplemented if possible by savings and private pensions), that bedrock is crumbling to sand for their successors. Just when increasing numbers of women are becoming entitled to the basic state pension in their own right, its value is falling further below the poverty line - the level of eligibility for means-tested income support. Increasingly, a substantial amount of private pension income (from occupational or personal pensions) will be required to avoid poverty in retirement. Such developments in the structure of the pension system, discussed more fully later in this chapter, may reinforce gender inequalities in prospects for a financially secure retirement, despite women's increasing participation in the labour market. The implications of pension policy are also important for men. Pension penalties - which are being deepened by the most recent developments - are imposed on all those, men or women, whose employment patterns are shaped by caring responsibilities. Because of the long time frame involved in the accumulation of pension entitlements, the retirement income of each generation of women reflects both past reforms in pension arrangements and the changing norms about marriage, motherhood and employment prevailing during that generation's working life. Feminist analysis has barely begun to explore how changing gender roles interact with shifts in pension policy to structure income inequality in later life between men and women and also among women. A political economy approach is used in this book. Thus a key concern is to analyse how British welfare policies alter the distribution of power, income and life chances, creating winners and losers in later life. Specifically, how does pension policy operate to construct life chances differently for men and women, and for different social groups within each gender? Changes over time in gender relations and in the pension system highlight new issues of distributional equity - between men and women, parents and the childless, partnered and lone parents, those who provide informal care and those who do not. The book sets out to explore these issues, using recent research on how employment and private pension acquisition vary according to gender, educational level, class, ethnicity, marital status, parenthood and birth cohort. A major focus is how changes and continuities in the gender division of labour and in patterns of partnering and parenthood shape gender inequalities in pension prospects. In this chapter, the evolution of gender roles, trends in employment and earnings and the gender implications for pension acquisition are considered. Gender-relevant changes in the British pension system since 1940 are outlined, including recent reforms shifting the balance of pension provision towards the private sector. Debate as to the optimum balance of state and private pensions 2 Trends in gender relations, employment and pensions has taken on a new urgency as the inherent risk of pensions based on investment in the stock market becomes increasingly apparent. Gender relations and older women's income The disproportionate share of poverty borne by today's older women reflects the model of gender relations prevailing during much of their earlier lives, one in which a distinct gender division of labour confined most married women to raising a family and homemaking — the male breadwinner—housewife model. Married women of this generation, even those with some hours of paid work, expected to undertake the major share of domestic tasks, including the nurturing and socialisation of children and the care of other family members, freeing men to maximise their earnings and support their family. Before the 1950s, women were expected (or even required) to resign from their jobs on marriage, although later in their lifecourse employment became common. The result can be seen in the work histories of older women. Those who had ever been married had fewer years of employment and substantial periods of part-time, rather than full-time, work relative to never-married women and to men, with consequently low private pension income (Ginn and Arber, 1996). Longitudinal research has confirmed the profound effect of work history in differentiating later life incomes (Bardasi and Jenkins, 2002). Figure I.I: Individual income2 of men and women aged 6S+ by marital status 140 -i----------------------------------------------------------------------------------------------------------- Note:3 median gross income. Source: Ginn and Price (2002) using data from the GHS 1994-96 3 Gender, pensions and the lifecourse As a result, gender differences in older people's income are substantial and three quarters of pensioners claiming means-tested financial support are women. Figure 1.1 shows median individual incomes of men and women over age 65 in the mid-1990s, distinguishing according to marital status. The median, which is the 'middle' value, is a better measure of the average than the arithmetic mean, because it is less influenced by extreme values. Older women's personal income was only about 60% of men's, on average. For example, in the mid-1990s, older women's median income was ^72 per week, compared with ^118 for men. Because older women are more likely than older men to live alone, with all the dis-economies entailed in solo living, these figures underestimate the gender difference in living standards. Older women's personal income varies with marital status, reflecting the way earlier family caring roles reduced their opportunity to contribute and benefit from pension schemes. Figure 1.2: Private pensions of men and women aged 6S+ by marital status a) Percentage receiving a private pension M Men El Women Married Single Marital status b) Median amounts for those with a private pension Widowed Divorced/ separated Married žiliek- Widowed Marital status Source: Ginn and Price (2002) using data from the GHS 1994-96 Divorced/ separated 4 Trends in gender relations, employment and pensions This pattern of income inequality with gender and marital status arises mainly from differential receipt of private (occupational or personal) pensions. Only a third of older women have any private pension income, including widows' pensions based on their deceased husbands' private pensions, and the amounts are less than for men (see Figure 1.2). Among women with some private pension income, the amounts received by married, widowed and divorced women are low compared with single (never married) women. For the remaining two thirds of women, their entire pension income is through the state. In the 1990s, private pensions contributed 25% of older men's personal income but only 11% of women's (Ginn and Arber, 1999). Will the apparent revolution in British women's employment over the past few decades enable working-age women to avoid the pension poverty their mothers and grandmothers are experiencing? What are the prospects for reducing gender inequality in later life income in the future? The next sections of this chapter consider how gender relations in Britain have evolved, and review current gender inequalities in employment and earnings over time. Evolving gender relations Despite changing attitudes towards gender roles and women's increased employment participation, in Britain the bulk of unpaid domestic and caring work is still performed by women (Gershuny et al, 1994; Murgatroyd and Neuburger, 1997; ONS, 2002a), constraining their employment opportunities while boosting those of married men (Gershuny, 1997). Most employed women work a 'second shift' of domestic labour (Hochschild, 1989), bearing the bulk of responsibility for childcare at the expense of their current and future earnings. The reasons for the continuity in the gender division of domestic labour, and whether more radical change can be expected in the future, are matters of debate (Crompton and Harris, 1999). While Hochschild (1989) refers to a 'stalled revolution', noting that changes in the domestic division of labour do not match the increase in women's employment, Gershuny et al (1994) suggest 'lagged adaptation', in which men are merely slow to adjust to new realities by renegotiating domestic responsibilities. Other writers suggest the lack of change in women's share of domestic work is due to the persistence of male material dominance, combined with gender socialisation. An alternative perspective is offered by Hakim (2000) who argues that women's preferences - for building a home rather than a career - are the main reason for their disadvantages in the labour market. However, the majority of women try to maintain both employment and family, their employment behaviour reflecting external constraints as much as prior preferences (McRae, forthcoming). It is not only men's failure to undertake an equal share of domestic work that influences women's opportunities in the labour market. The state's provision (or lack) of welfare services to complement and support family caring is also important in enabling women to engage in full-time employment (O'Connor et al, 1999). In liberal welfare states such as Britain and the US, policy has been contradictory, 5 Gender, pensions and the lifecourse simultaneously urging women into paid employment while providing minimal support. The chronic lack of publicly subsidised childcare, in particular, has tended to confine British mothers to being secondary earners if partnered, and low/no earners if not, as discussed more fully in Chapter Four. Falling fertility rates, which reflect women's difficulties in reconciling motherhood with increasing employment participation (Duncan, 2002), are a major cause of ageing populations. Achieving the policy goal of promoting women's employment and independent pension entitlements without further depressing fertility rates — an outcome important to the sustainability of pension schemes — is considered in Chapter Six, in the context of policies in other EU countries. British women's difficulties in accommodating the dual roles of paid and unpaid work are central to understanding the persistence of the gender gap in hours of employment and earnings. These gender inequalities in the labour market, which lead to women's lower pension income in later life, are described in the next section. Gender inequalities in employment British men and women's overall rates of participation in the labour market are converging. While the economic activity rate of working-age men (16-64) declined from 94 to 79% between 1959 and 2000, that of working-age women (16-59) rose from 47 to 70% over the same period (ONS, 2002b, Chart 4.1). However, much of the increase in women's employment is in part-time work. In 2001, 37% of working-age women were employed full time and 28% part time. The corresponding figures for men were 61% and 5% (ONS, 2002b). The proportions of men and women who were full- and part-time employees in 1986 and 2001 are shown in Figure 1.3. Women's full-time employment grew noticeably between 1986 and 1990 but there has been no increase between 1990 and 2001. Thus the gender convergence in employment participation is deceptive: full-time employment of women remains much lower than men's and may have stalled at under 40%, severely limiting women's earnings and ability to build an adequate pension income. Two factors magnify the effect of the gender differences in full-time employment rates on lifetime earnings. First, women's age profile of employment participation differs from mens. British women's full-time employment rates tend to peak in their twenties, when relatively few have had children, and in their forties, reflecting women's return to longer hours of employment as children become more independent (see Figure 1.4). The timing in the lifecourse of part-time working is relevant to occupational status achieved and to earnings. Among women, part-time work is common in the prime earning years between ages 30-54 (Figure 1.4) when opportunities for wage gains and advancement would otherwise be highest. In contrast, part-time work among men is rare under age 55. Second, the combination of this age profile with the characteristics of part-time jobs in Britain undermines women's capacity to earn pension entitlements. 6 Trends in gender relations, employment and pensions Figure 1.3: Percentage employed, by employment status, men and women aged 16-64 (1986 and 2001) 100 80- 60 40 20 Mi-n D Self-employed/other □ Part-time ■ Full-time n r Women 1986 Men Women 2001 Note: As % of population aged 16-64. Source: ONS (2002b,Table 4.3) Part-time work is generally associated with poorer working conditions, job insecurity, lack of training and career development opportunities and lack of fringe benefits, as well as lower hourly pay. Part-time workers are typically concentrated in small firms, in certain private sector industries which are highly feminised and poorly unionised - all contributing to low pay (Paci et al, 1995). The effects of part-time work tend to be long-lasting and may apply even to well-qualified women. Qualitative research on British women who entered the labour market between 1946 and 1970 showed that the adverse effect of part-time employment on mothers' occupational status (relative to mothers employed full time) applied equally to earlier and later birth cohorts of women, despite the better qualifications held by later cohorts (Jacobs, 1999). There has been a growth in the proportion of'short' part-time jobs (under 20 hours per week) in Britain; such jobs tend to be of very poor quality and to act as a trap rather than a bridge to a better job. The EU Social Chapter obliges member states to set minimum hourly wages (although these are still very low) 7 Gender, pensions and the lifecourse Figure 1.4: Percentage of women employed full and part time by age group 100 20 - -------------1-----------------1-----------------1-----------------1-----------------1------- 20-24 25-29 30-34 35-39 40-44 45-49 Age group Note: Full-time employment defined as 31 + hours. Source: Ginn et al (2001, Figure 2.2), using data from the GHS 1993-94 50-54 55-59 and to ensure part-timers have equal access with full-timers to maternity leave, career break schemes, sick leave and redundancy pay. Britain enacted the necessary legislation in 2000. Some improvements for part-timers can be expected, although it takes time for the effects of legislation to translate into employment practices. Women's gaps in employment, or reduction in hours of paid work, after giving birth are greater in Britain than in those European countries where affordable quality childcare services are widely available 0oshi and Davies, 1992). Women's maternal and marital status also affects their likelihood of being employed full time (as discussed more fully in Chapter Four). Gender inequalities in earnings Since contributions to private pensions are based on earnings, higher earnings over a longer period generally translate into higher pensions, although the precise relationship depends on the type of pension scheme. In final salary occupational pension schemes, earnings after age 40 are especially important to the amount of pension entitlement, as the latter is based on salary in the last few years before retirement, as well as on years of membership. In defined 8 Trends in gender relations, employment and pensions contribution (DC) schemes, contributions are invested on the stock market, building a fund that is annuitised at or soon after retirement. In such schemes, the individual bears the risks of both poor investment choices by the plan provider and of adverse developments on world stock markets as a whole. The fund accumulated at retirement reflects the level and duration of contributions but also the time they have had to grow. Thus contributions made early in the working life are most important. DC pension schemes may be occupational or personal (including stakeholder) pensions. Among all full-time British employees, women's hourly earnings are still only 82% of men's, and among non-manual employees even lower at 70%. Hourly earnings of women employed part time are lower still, only 61% of the hourly earnings of men employed full time (ONS, 2001). Even among full-timers, women tend to work shorter hours than men and their weekly pay was 74% of men's in 1999 (Rake et al, 2000). The earnings of married women, in particular, remain far below men's, both nationally and within couples (Arber and Ginn, 1995). A fifth of employed women work less than eight hours per week, most earning wages too low to allow contributions to either state or private pensions. Those who are low paid tend to remain low paid and there is a high turnover between low pay and non-employment (McKnight et al, 1998). Toynbee (2003) blames women's low pay on the low level of the Minimum Wage (^4.20 per hour in 2002-03) and on the continuing low valuation of what are regarded as women's skills — caring, cleaning, cooking, teaching and nursing. Pay in these occupations is falling further behind that of male-dominated occupations. The gender gap in pay is modified by factors such as women's lifecourse stage, family circumstances, educational level, occupational class and ethnicity. Recognising this diversity is crucial to understanding variation in the extent of women's pension disadvantage relative to men. For example, whereas men's earnings tend to rise with age, suggesting they receive a premium for age and experience, women's earnings decline from age 30, reflecting their interrupted employment patterns and labour market segmentation. For workers in their twenties, the gender gap in full-time hourly wages has diminished since 1976 from about 15% to less than 10%. But the gender gap widens after age 30 and is around 25% among all those aged over 40 (Rake et al,2000,p 46-7). Figure 1.5 shows the gender gap in individual income by age group during the working life. The widening of the gender pay gap with age is likely to reflect the cumulative effects of women's involvement in family caring work and also gendered ageism, in which employers perceive women employees as too old for training or promotion at a younger age than equivalent men (Itzin and Phillipson, 1993). Family responsibilities and especially motherhood magnify the gender pay gap. The 'family gap in pay' — or the wage loss due to the presence of dependent children - for women aged 24-44, after taking account of age, education and other relevant variables, was relatively large in Britain compared with other countries, rising from 8% for one child to 24% for two children (Harkness and Waldfogel, 1999). 9 Gender, pensions and the lifecourse Figure 1.5: Median individual income of men and women by age group 50 - 0 "I---------------1---------------1---------------I---------------1---------------1---------------1---------------1---------------I--------------- 16-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 Age group Source: ONS (2002b, adapted from Chart 5.5), data for 1999 The gender pay gap may diminish for later cohorts, as younger women enter the labour market with better educational qualifications and benefit from improved labour market opportunities. However, neither qualifications nor entry to top occupations guarantee gender equality in pay. Indeed, economists have shown that gender inequality in earnings widens with occupational status and that the relationship between occupational status and earnings is weaker for women than for men in Britain (Sloane, 1990). The extent to which periods of caring for children or older relatives place women at a disadvantage in acquiring pension entidements of their own depends on the structure of the pension system, including the balance of public and private provision and redistributive features in state pension schemes. In the next section, gender-relevant features and reforms in the British pension system are examined. Gender and the changing British pension system The design of British welfare institutions incorporated the assumption that women would either marry and have financial support throughout their life from a husband, or would remain single and childless, pursuing a career in the same way as men (Lister, 1994). State pensions have since been modified to reflect changes in gender relations, but the declining value of state pensions and the increasingly heavy reliance on private pensions has rendered these adaptations ineffective in ensuring women's economic independence in later life. The gender effects of pension policy since 1980 can be seen in the changing ratio of 10 Trends in gender relations, employment and pensions older women's income to older men's. In the mid-1980s older women's median personal income was 71% of men's, declining to 62% in 1993-94 and only 53% in 1998. Women have been less able than men to compensate for declining state pensions through private sector pensions. Women in Britain, if they marry, can obtain retirement income through three main routes, which are conceptually distinct although often combined in practice. These are: 1. sharing a husband's pension income; 2. receiving pensions derived from a husband's (or former husband's) pension contributions (as wives, ex-wives or widows); and 3. receiving pensions based on their own contributions. Relying on the first route is a risky strategy. First, an increasing proportion of marriages end in divorce and divorced women are unlikely to obtain an adequate pension settlement, as discussed in Chapter Four. Second, not all husbands share their pension income equally. Third, husbands are not necessarily successful breadwinners and pension earners, especially if they suffer redundancy or ill-health in midlife, curtailing their pension-earning years. The second route, through pensions derived from a husband's contributions to pension schemes, has the drawback that such pensions are often small. Most marriages end in widowhood for the woman - about half of British women aged over 65 are widows. Widows generally receive only half their deceased husband's occupational pension and may receive a smaller fraction from his personal pension. The third route, in which women acquire pension entitlements based on their own employment, can provide financial security but this is only possible in a women-friendly pension system - one that minimises the pension effects of gaps in employment and periods of low earnings due to caring commitments. A primary concern of this book is to explore to what extent women are able to build adequate state and private pensions of their own - the 'independence model' of pensions - and how pension policy promotes or frustrates this outcome. Because current older women's incomes reflect pension policies in place up to 50 years ago, it is helpful to set out the main changes in pension structure, focusing on gender effects and distinguishing three policy periods between 1940 and 2002: establishment of state and private pension schemes (1940-74); revitalising state pensions (1975-80); and state pension retrenchment with promotion of private pensions (1980 on). State pensions 1940-74 The retirement pension introduced by the 1946 National Insurance Act was originally intended to be "sufficient without further resources to provide the minimum income needed for subsistence" (Beveridge, 1942, p 122), although low enough to encourage voluntary additional saving by breadwinners, mainly men. However, Beveridge's 'subsistence' amounts were not based on objective II Gender, pensions and the lifecourse study of need and were more miserly than Rowntree's 'primary poverty' standard. Compared with social insurance contribution levels in other EU countries, Britain's National Insurance (NI) contributions are relatively low. This is reflected in state spending on individuals aged over 55, which is over 10% of GDP on average for the EU but only 5.5% in Britain (EPC, 2001). Because of the gendered assumptions on which the welfare state was based, a married man's NI contributions buy a basic flat-rate pension for two - his own and a reduced (60%) Category B pension. This is payable for a wife when she is over state pension age (60, but rising to 65 by 2020) provided her husband is aged at least 65. This arrangement gives rise to the inequity that a wife who earns entitlements in her own right through paying NI contributions may receive no more than a stay-at-home wife, even though both may have had similar domestic responsibilities. Widows over state pension age receive a basic pension equal to their deceased husband's (if better than their own entitlement) and divorced women can use their ex-husband's contribution record for the period of the marriage if this improves their own entitlement. Thus the postwar welfare state 'compensates' married women for limited opportunities in the labour market through derived benefits - that is, benefits based on their husband's contributions. Women's lower state pension age of 60, compared with 65 for men, was conceded in 1940 following representations from the National Spinsters' Association and from married men whose wives were several years younger than themselves and therefore ineligible for the Category B pension at the time the husband reached age 65 (Thane, 1978). This reform was deemed a cheaper solution than increasing the level of the basic pension. In most other respects, women were ill-served by the Beveridge scheme. Married women were handicapped by the notorious 'half tesť, whereby those paying contributions for less than half their working life since marriage lost the value of all their contributions. Periods of caring for children, parents or parents-in-law attracted no credits in the basic pension. Married women were further discouraged from building their own pension by the right to opt for reduced National Insurance (NI) contributions which carried no pension entitlement - the Married Woman's Exemption, or 'small stamp'. Many married women did not appreciate that paying a reduced NI contribution disqualified them from receiving a basic state pension of their own. As a result of these gender-biased provisions and women's short employment records, only a quarter of women pensioners in the late 1990s received a basic pension solely through their own contributions and of these only half received the full amount. A further third were married women with a Category B (60%) pension and just over 40% were widows with a Category B (100%) pension (Social Security Committee, 2000, p 168). The value of the pension increased slightly, as a proportion of male manual earnings, from 19.1% in 1948 to 21.5% in 1975 (Johnson and Falkingham, 1992), yet for most of the post-war period it has remained about 10% below the level of means-tested benefits and thus wholly inadequate to live on. From 12 Trends in gender relations, employment and pensions 1961, a small state Graduated Pension was introduced. This extended an earnings-related scheme to those excluded from occupational pensions, but was too little and too late to stem the demand for private provision (Hannah, 1986). Private pensions 1940-74 The low level of state pensions fostered rapid expansion in occupational pension coverage from only 13% of the workforce in 1936 to 47% in 1967, falling slightly to 46% in the 1970s (Hannah, 1986). While other European state's provided substantial state earnings-related pension schemes to most employees, Beveridge's recommendations had "furthered the conditions within social security for the growth of a multi-billion-pound enterprise of private pensions" (Shragge, 1984, p 33). When the Graduated Pension was introduced, occupational pensions meeting certain standards were allowed to contract out, diverting the increase in National Insurance contributions of both employer and employee into the private scheme. The lack of a realistic state second tier pension and the growing importance of occupational pensions were to have serious consequences for British employed women, because of their limited access to occupational pensions. Most occupational pension schemes provide a defined benefit (DB) pension at retirement, based on the individual's years of pension scheme membership and final salary (their average salary in the final few years before retirement). The extra contributions generally made by employers to their occupational pension schemes, above the minimum required to replace the state second tier pension, are effectively a form of deferred wages and make occupational pensions a valuable fringe benefit of employment. British employers' welfare payments, as a proportion of total employee remuneration, doubled between the mid-1960s and the 1980s (Green et al, 1984), improving benefits for the minority of the population belonging to an occupational pension scheme but exacerbating the social division of welfare (Titmuss, 1958). The social cleavage was of gender as well as class, with women's coverage by private pensions well below men's. A major reason is that part-time employees were often excluded from occupational pension schemes. This has been unlawful in the EU since 1986 (following the Bilka-Kaufhaus judgment) and in Britain the 1995 Pensions Act prohibited the exclusion of part-timers from occupational pension schemes. However, small employers in the private sector, for whom women part-timers often work, rarely operate an occupational pension scheme (Ginn and Arber, 1993). The return on contributions in an occupational pension scheme is greatest for those with continuous membership until retirement and earnings that rise with age. Those who leave the scheme early, mainly women with family responsibilities, receive less value for their contributions. Also, women tend to have a flatter earnings profile than men, with a lower final salary and hence a lower pension even for the minority of women whose length of service matches men's. Thus women have both lower coverage rates than men and poorer benefits due to lower earnings, fewer pensionable years and the penalties of 13 Gender, pensions and the //fecourse early leaving (Groves, 1987; Ginn and Arber, 1991,1993,1996). Despite being hailed as a success story, occupational schemes have yielded quite modest pensions. Among the recently retired with an occupational pension, the median amount for a non-married man in 2000 was £81 per week and even less for a comparable woman: £47 per week (House of Commons, 2002). It was not until 1975 that a Labour government introduced a radical new state pension scheme, which addressed a number of the inadequacies of state pension provision and promised a much better pensions deal for women. Revitalising state pensions 1975-80 The 1975 Social Security Benefits Act introduced by Labour heralded major improvements to the British state pension system. First, the basic pension was formally indexed to rises in national earnings or prices, whichever was the higher. This ensured that older people would share in rising national prosperity and signalled that the basic pension would provide a secure foundation on which individuals could build other sources of income in later life. Second, the pension needs of women were explicitly addressed for the first time. The 'half tesť was abolished and the Married Woman's Exemption phased out. Although married women who were already paying reduced NI contributions were allowed to continue after 1978, women who divorced were required to pay the full stamp. Most important, Home Responsibilities Protection (HRP) allowed years of family caring to count towards eligibility for the basic state pension. Those not in paid employment or who earn less than the Lower Earnings Limit (LEL) (about £75 per week in 2002), are excluded from contributing to the NI system. This applies to some two million women each year (McKnight et al, 1998). However, if they are caring for a child aged under 16 (or 18 if in full-time education) they qualify automatically for HRP. Carers of a frail or disabled adult may also be covered by HRP. HRP reduces the number of contribution years required for entitlement to the basic pension so that, provided NI contributions have been paid for at least 20 years, a woman may still qualify for the full amount if the remaining years are covered by HRP It was expected that pensioners would increasingly be floated off means-tested benefits by a rising basic pension and that HRP would eventually ensure that most women would receive the full amount in their own right. By protecting pension rights during caring years, HRP promised to help accommodate women's dual commitments to family and employment. A new State Earnings-Related Pension Scheme (SERPS) replaced the meagre Graduated Pension. Benefits, based on the best 20 years of earnings (revalued at state pension age), would accrue at 1.25% per annum and entitlements were automatically portable across jobs or across gaps in employment — a great advantage for those with interrupted employment patterns, particularly women. Contributions to SERPS through the NI scheme were compulsory for all employees earning above the LEL, unless they contracted out into an occupational pension scheme. 14 Trends in gender relations, employment and pensions The new state pension package, if allowed to mature, would have been redistributive towards the low paid, as well as minimising the adverse effect of women's caring responsibilities on their state pension income. A woman on average female manual earnings could expect (at maturity of SERPS in 1998) a replacement rate of 50% of her earnings from the combination of basic pension and SERPS, while her higher paid non-manual sister could have expected a replacement rate of 43% (Groves, 1991). The benefits of the 1975 legislation were shared across generations, helping both current pensioners and working-age people. The reinvigoration of state pensions was short-lived, however, overtaken by the neo-liberal reforms of the 1980s. Retrenchment of state pensions 1980-2002 With the election of a Conservative government in 1979, an individualistic, competitive ideology gained ascendancy, expressed in privatisation of many aspects of welfare. The ensuing spate of reforms included cutting state pensions and promoting private personal pensions as an alternative to SERPS, a policy that ignored women's lesser ability to acquire private pensions. The main changes since 1980 are briefly outlined. From 1980, the basic state pension has been indexed only to prices, eroding its relative value from 20% of average male earnings in 1980 to around 15% in 2002. It is projected to decline to only 7% by 2050 (DSS, 1998), falling ever further below the level of means-tested benefits. The second tier SERPS was scaled back in the 1986 Social Security Act, so that it would in future provide a maximum pension of 20% (instead of 25%) of revalued earnings. A major change was that the pension was to be based on average earnings over a working life of 44 years for women and 49 years for men, instead of average over the best 20 years. This substantially reduces the amount of SERPS pension for those with periods out of employment or on low earnings, compared with the original formula that would have become mature in 1998. There was a further cut in SERPS for women. Those widowed after October 2002 will inherit only half their deceased husband's SERPS, instead of the whole amount as originally provided in the 1975 Act. The 1995 Pensions Act brought yet another cut to state pensions in the future, affecting women only. Their state pension age will be raised from 60 to 65, phasing in the change from 2010 until 2020. By 2020, women will need 44 years of contributions instead of 39 (including any HRP years) to qualify for a full basic pension. Similarly, for the state second tier pension, average earnings over 49 instead of 44 years will be used to calculate the pension entitlement. These changes will substantially reduce the amount of state pension income for women born after 1950, unless they are able to continue in full-time employment until age 65 (Hutton et al, 1995). In 2002, SERPS was replaced by the State Second Pension (S2P). This differs from SERPS in two ways. First, it will provide a higher pension for the low paid than SERPS, by applying a higher accrual rate to low earnings bands. 15 Gender, pensions and the lifecourse Second, there is some carer protection in S2P, in that mothers of a child up to age six are credited into the scheme as if they were paying contributions on earnings of about -£10,000 pa. Some of those caring for adults may also qualify for carer credits. It is planned that the S2P will become a flat-rate pension by 2007. These reforms potentially benefit women. However, because of the projected decline in value of the basic pension under current policy, the S2P and basic pension combined will only provide a pension below the level at which means-tested benefits are payable (Palkingham and Rake, 2001; Brooks et al, 2002; Falkingham et al, 2002). Thus "the new State Second Pension will make little difference to women as the gains from this will be washed out by the ongoing erosion of the basic state pension" (Evason and Spence, 2002, p 4). The low level of the combined basic and S2P perpetuates the pensions poverty trap, in which additional pensions, savings or earnings may bring no financial gain because of the loss of means-tested benefits. An indication of the extent of state pension cuts planned by the Labour government in 1998 is provided by projections to 2050. The cost of the state pensions (basic and SERPS/S2P) was projected to fall from £34bn to ^26bn (in 1997 earnings terms) and the value of the combined state pensions to fall from 37 to 20% of average male earnings in 2050 (PPG, 1998). Spending on state pensions was projected to fall from 4.4 to 3.4% of GDP (DSS, 1998). Above-inflation rises in the basic pension in 2001 (+7.4%) and 2002 (+4.1%) slightly modify these projections. The government hoped that the shortfall in older people's incomes would be met by a further expansion of private pensions; their aim was to reverse the balance of pension provision, which was then 60% state and 40% private. Expansion of private pensions 1980-2002 Despite warnings from the government's natural supporters as early as 1985 that personal pensions would bring poorer returns for many employees than SERPS (for example, CBI, 1985), the Conservative government pressed ahead, providing generous financial incentives to encourage individuals to opt out of SERPS into the new Appropriate Personal Pensions. These personal pensions are individual portable defined contribution (DC) accounts whose fund must be annuitised (converted to an annual income for life) at or during retirement. A major drawback, played down at that time, is that investment risk is individualised, shared neither with the workforce as a whole nor even with members of an employer's scheme. Some five million individuals took the bait, providing lucrative business for the personal pensions industry but dubious benefits to contributors. Some companies set up group pension schemes on a DC basis, reducing administration costs and sharing investment risk among a larger group. Personal pensions provide poor value for most women and for the low paid (Davies and Ward, 1992; Wane, 1995; Ginn and Arber, 2000a). In any DC scheme, contributions made early in the working life have a disproportionate 16 Trends in gender relations, employment and pensions a effect on the fund at retirement; yet women's ability to contribute is often very limited until their children have become independent. To make matters worse, the same size of fund buys an annuity for a woman that is about 10% less than man's, due to actuarial calculations reflecting women's greater average longevity. The charging structure, with flat-rate fees paid predominantly in the first year of membership, penalises the low paid and those with breaks in contributions, making personal pensions a very poor option for women who have gaps in full-time employment when their children are young. Charges for administration, investment management and annuitisation may reduce the value of contributions by 45%, according to experts (Murthi et al, 2001). An estimated 30-40% of personal pension account holders find that charges actually exceed the amount they have contributed (Disney and Johnson, 1997). Many employees succumbed to high-pressure sales techniques and opted out of an advantageous occupational pension scheme into a personal pension — the mis-selling scandal (Ward, 1996). The issue of proliferating pension choices and new risks is considered more fully in Chapter Two. Stakeholder pensions (SHPs) were introduced in 2001 partly as a result of the tarnished image of personal pensions. Employers with over five employees who do not operate an approved occupational pension scheme are required to offer an SHP to their employees. SHPs are a more heavily regulated form of personal pension, provided by the same companies that provide personal pensions but intended for individuals with modest to average incomes. SHPs have the advantage of a cap on administration charges at 1% per annum of the fund and a feature which is particularly helpful to women is that gaps in SHP contributions incur no extra charges. Contributions of up to £3,600 per annum can be made to an SHP on behalf of a non-earner. However, this will only help women -with a husband (or other relative) who is well-off, generous and accepts women's need for their own independent pension entitlement. Drawbacks are that SHPs share with other personal pensions the risk of poor investment returns. Moreover, hidden fees (such as dealing charges) are not capped so that costs may still be high relative to individuals' contributions (Wynn, 2001). If the S2P becomes flat-rate by 2007 as planned, SHPs will be the only earnings-related pensions available for the moderately paid who lack access to an occupational pension. The policy of encouraging contracting out of state second tier pensions through advantageous rebates is costly, withdrawing financial resources from the National Insurance (NI) system and hence reducing the scope to improve the state pensions that are the mainstay of older women's income. Although the Conservative government claimed that opting out of SERPS would reduce public spending, the net cost of the financial incentives to opt out, from 1988 to 1993, was estimated as £6,000 million (1988 prices) (NAO, 1990). Every pound paid in rebate on a personal pension incurs a net cost to the NI Fund of 22p (PPG, 1998) and this subsidy applies equally to SHPs. A further loss of public resources that is largely invisible arises from tax relief on private (occupational and personal) pension contributions by employees and employers. Tax relief on private pensions is long-standing but thft^meuft&T^ Fakulta sociálních studií JoštovalO ---- —^^ 602 00 BRNO ]7(^> Gender, pensions and the lifecourse of tax lost in this way grew dramatically under the Thatcher administration, from £1.2 billion in 1979 to £8.2 billion in 1991 (Wilkinson, 1993), reflecting the spread of personal pensions, and continued to rise (Sinfield, 2000). By 2000 such tax spending (not counting the £1.5-2.2 billion tax forgone on lump sums from pension schemes) had risen to £13.7 billion in that year, equivalent to over 40% of state spending on the basic NI pension (Sinfield, 2000). Tax spending on private pension incentives is highly regressive, with half the benefit received by the top 10% of taxpayers and a quarter by the top 2.5% (Agulnik and Le Grand, 1998). Pensions tax relief and rebates cost Britain nearly 3% of GDP in forgone revenue (Adema, 2000), a huge expenditure that could otherwise be spent on improved state pensions. A gender audit of tax spending has not been carried out but it is likely that men are the chief beneficiaries, while women have been harder hit than men by the cutbacks in state pensions. The main elements of the British pension system in 2002 are shown diagrammatically in Figure 1.6. For a comprehensive account, focusing on the relationship between state and private pensions, see Liu (1999). Employees must make NI contributions to both the basic state pension and a second tier pension, but may choose to contract out of S2P into one of a variety of private schemes, including occupational and personal pensions. However, the choices are more limited for those lacking access to an occupational pension scheme. As noted above, final salary occupational pension schemes (DB) are the most advantageous, due to the employer's contribution, which is generally higher Figure 1.6: Outline structure of the British pension regime (2002) Third tier, voluntary Second tier, mandatory First tier, mandatory11 Additional voluntary contributions (AVCs) (DC) or Free standing AVCs (DC) or Pe rsonal pension (DC ) _...... SER.PS, replaced Fir al Mo ney Personal by State salary purchase pension, Second or occupational or occupational or can be Pension from pension pension St akeholder April 2002 (DB) (DC) pension (DB) (DC) Basic state pension Note:'Those earning below the Lower Earnings Limit (LEL) pay no National Insurance contributions, but may acquire credits through Home Responsibilities Protection; those earning between the LEL and the Lower Earnings Threshold (LET) acquire NI credits; those earning above the LET must pay NI contributions. For the self-employed, second tier pensions are not compulsory. 18 Trends in gender relations, employment and pensions than in an occupational money purchase (DC) or personal pension scheme. Those with sufficient income may opt to make additional voluntary contributions (AVCs) to an individually arranged personal pension, to a company-sponsored AVC scheme or to an individually arranged free standing AVC. Analysts from academic, political and actuarial backgrounds have questioned whether the government's pension policy is politically sustainable. In the final chapter, the major concerns expressed — about pensioner poverty, savings disincentives due to means-testing and failures in the private pensions sector — are considered, along with policy alternatives. Summary This chapter has outlined British trends in gender relations and in women's employment and earnings, tracing the parallel changes in the pension system that will influence gender differences in pension income in the future. The remainder of the book is organised as follows. ChapterTwo is concerned with pension choices and analyses gender and class inequalities in the ability to contribute to the expanding range of types of private pension, using data from the General Household Survey (GHS). The characteristics of employees who have opted for personal pensions or remained in the state pension scheme are examined, distinguishing between those with access to an occupational pension and those without, and considering the implications for later life income. Little is known about the pension prospects of men and women from minority ethnic groups and how they are affected by disadvantage and discrimination in the labour market, as well as different cultural norms concerning women's employment. Chapter Three uses data from the Family Resources Survey to analyse variation in employment and private pension scheme membership according to gender and specific ethnic group. Chapter Four examines how the impact of childrearing on participation in paid employment and private pension scheme membership differs according to women's partnership status, using data from the GHS. Recent legislation permitting pension sharing is considered in relation to the pension needs of divorced and cohabiting men and women. Chapter Five also uses GHS data, focusing on differentiation among women in their employment and private pension coverage, according to their educational qualifications. It critically examines the thesis that graduate mothers can largely avoid the adverse impact of childrearing on future pension income, in contrast to less skilled women. Chapter Six turns to the European Union, comparing trends in fertility and women's employment, and the differing extent to which EU pension systems are adequate and adapted to the needs of those with caring commitments. Finally, Chapter Seven summarises the gender implications of British pension policy, examines the concerns of a range of critics and discusses the main alternative strategies proposed, focusing on the gender implications of each.