TWO Choice.and risk in pensions? gender and class inequalities Pension providers and governments alike warn that individuals need to save more through private pensions. Most economists attribute lack of saving to irrational behaviour described as 'myopia' — an inappropriately short time horizon. Yet individuals often have sound reasons for reluctance to save additional amounts towards pensions (Rowlingson, 2002). These include a rational response to uncertainties, especially the risk of a poor return on investments. The mis-selling of personal pensions and the theft and misuse of occupational funds are indications of the way pension risks and costs are increasingly passed on to individuals (Ward, 1996; Peggs, 2000; Ring, 2002). Meanwhile state pensions are shrinking as governments seek to reduce state spending on pensions, as part of a more general rolling back of the welfare state (Aldridge, 1998). Confronted with unsatisfactory state pensions and risky private provision, even individuals who can afford to are understandably reluctant to save for their old age in this way. Widening choice, growing complexity and uncertainty about the future of pensions means an increasing risk of making decisions that do not provide the best value for money. When politicians and pensions experts admit they find the British pension system hard to grasp, it is no wonder that individuals are bemused by pension scheme rules, and that many feel alienated or misled by 'information' from the state or the purveyors of private pensions. For those with interrupted and unpredictable employment trajectories, choosing an optimal pension strategy becomes well-nigh impossible. This chapter first considers the issue of pension choice in the context of uncertainty about risks in the various types of second tier pensions. Gender and class inequalities in pension coverage are then analysed using data from the General Household Survey (GHS). The characteristics of employees who opted for a personal pension or who remained in the state pension scheme are examined, distinguishing between those who could have belonged to an occupational pension and those who lacked access to such a scheme. The chapter finally considers the issue of mis-selling of personal pension plans to men and women for whom this was unlikely to be the best option. Choice, information and risk Pension choices were limited during the post-war period before the 1980s. Where employers offered an occupational pension scheme, membership was 21 Gender, pensions and the lifecourse compulsory. The main social divisions were between employees with or without an occupational pension scheme (although the quality of these varied) and between employees and the self-employed. For most employees, individual advice was not important; information about pensions, such as from a trade union or employer, would apply to most of a company's employees. However, married women's long-standing right to opt for the 'small stamp' in National Insurance led to many making a decision they regretted at retirement, having not realised that they would forfeit a basic state pension in their own right. This exemplifies the 'downside' of choice. Choices in pensions have proliferated since the 1980s. Pension saving in the first and second tiers of pension is compulsory for most workers (see Chapter One, Figure 1.6), with a number of options — whether to contract out of the state scheme into an occupational or personal pension, whether to make additional contributions above the compulsory level, whether to invest any surplus income into housing or other forms of saving or whether to use all income for immediate needs. Decision making is beset by multiple unpredictabilities. These include not only the future performance of the stock market as a whole, of specific pensions and savings plans and of the housing market, but also the individual's future employment, earnings trajectory, retirement timing and family circumstances. Pension choices are made in the context of an ideological and moral climate constructed by policy makers, commercial interests and the media. Since the mid-1980s, Conservative governments engaged in a'rhetoric of responsibility' (Smart, 1999) in which contributing to a private pension was portrayed as responsible behaviour while paying into state pensions was not. This ideological position, which is in sharp contrast to the rest of Europe, was associated with an emphasis on consumer power and choice, the latter being linked to notions of good and evil and right and wrong (Phillips, 1998). The morally inferior status assigned by this rhetorical framing to those relying solely on state pensions affects mainly older women (see Chapter One). Such dependence on state pensions may seem to disqualify recipients from full citizenship. It "withdraws their status as adults [and] may promote the insidious idea that, like children in Victorian Britain, they may be seen but not heard" (Mann, 2001, p 138). The potentially disempowering effects of social disapproval are magnified for those older people who depend on means-tested benefits, again mainly women. The Labour government elected in 1997, despite having condemned Conservative pensions policy while in opposition, continues to imply that making provision for retirement income through the private sector is morally superior, encouraging individuals to contract out of the state second tier pension through financial incentives (see Chapter One). People's understanding of second-tier pension schemes is limited (Williams and Field, 1993). Occupational pensions are often taken up without much forethought or knowledge and the complexity of written material about pension schemes is a deterrent (Field and Farrant, 1993). Contributors to personal pensions similarly lack understanding about them (Williams and Field, 1993) 22 Choice and risk in pensions and advice about the implications of women's interrupted employment patterns has been lacking (Davies and Ward, 1992). Women are particularly prone to feel uninformed about pensions. Hawkes and Garman (1995) found that women were three times more likely than men to say they had not joined an available occupational pension scheme because they knew too little about pensions or had not given enough thought to the matter. Almost 40% of contributors to a personal pension thought they would receive a guaranteed amount, unaware that their pension was linked to stock market performance. Employees also lacked understanding about the State Earnings-Related Pension Scheme (SERPS). Thus 42% of full-time employees who were not contracted out into a private pension scheme thought they were not contributing to SERPS, despite this being the default option (Hawkes and Garman, 1995). With an increasing range of options, especially the introduction of personal pensions in 1988, the chance of making substantial losses due to a particular pension decision has escalated. Large numbers of people, against their best interests, opted out of sound occupational pension schemes into a personal pension which was likely to provide a lower pension — "a clear case of poorly-informed buyers confronted with sales staff with a powerful incentive to sell" (Mann, 2001, p 133). This mis-selling scandal, with compensation to investors estimated to cost ,£13.5 billion (Jones, 2000) and the high-profile fraud of Robert Maxwell that robbed Mirror Group employees of their occupational pension fund, began to reveal the shaky nature of the pensions promise offered by private sector providers. Investigations revealed multiple flaws (Goode, 1993), and prompted a search, which does not seem to have been very successful, for regulatory mechanisms to prevent a repetition and restore confidence. Neither the endemic nature of private pension risk nor the inability of experts to provide advice to protect against such risk was fully appreciated by the public in the early 1990s. Despite a proliferation of regulatory bodies concerned with the conduct of the private pensions industry, the increased speed of trading stocks, with globalisation of finance markets and computerised processing, have all reduced the feasibility of regulation (Mann, 2001). Because pension funds occupy a grey area in terms of ownership, fund managers are enabled to act primarily in the interests of their shareholders, producing a poor deal for pension scheme members (Blackburn, 2002). In Britain, the collapse of Equitable Life's guarantees after 1999 showed that even an ancient and respected institution (founded in 1762) can make mistakes, while the implosion of Enron in the US in 2002 demonstrated (if this were necessary) that neither the integrity of top executives nor the independence of auditors can be assumed. These events highlight the irrelevance of financial education of consumers in the face of boardroom incompetence or fraud. Warnings about the effect of ageing populations on the future sustainability of state Pay-As-You-Go pension schemes have been rife but the equivalent threat to the viability of private funded pensions has been largely ignored. Indeed, a persuasive message of governments and private pension providers has 23 Gender, pensions and the lifecourse been that a personal pension fund is inviolate, owned and controlled by the contributor. The risk of loss in value was downplayed and despite the mantra (in small print) that share prices can go down as well as up, few took this warning seriously during the bull market of the late 1980s and early 1990s. Those belonging to occupational pension schemes were also blissfully unaware of the level of risk until the trickle of closures and reductions in projected benefits became a river in 2002 (see Chapter Seven). Shock at the pensions crisis might have been less if debate on pension privatisation in the 1980s and 1990s had been more balanced. Challenging the dominant expert discourse and assumptions has been less evident in pensions than in medicine and law, enabling fund managers, actuaries and advisers to marginalise the voices of pensioners and contributing members (Mann, 2001). Policy makers have failed to take seriously the arguments for maintaining robust state pension provision. The only organised resistance to the creeping individualisation of risk in pensions has come from the trade unions, first in successfully demanding a larger rise in the basic pension at the 2000 Labour Party conference and more recently in arguing for protection of final salary occupational pension rights and for employers who operate an occupational pension scheme to be compelled to contribute 10% of payroll (TUC, 2002). This latter proposal could, however, increase the stampede away from occupational pensions, leaving most employees with a choice between a stakeholder pension and the State Second Pension. The reductions in state pensions outlined in Chapter One - the decline in value of the basic pensions, numerous cuts in SERPS and a rising state pension age - are a collective loss imposed initially by Conservative governments in the 1980s and 1990s. Private pension losses are more likely to be experienced as a personal misfortune or as resulting from a failure of the individual's financial judgement. The idea that each individual is responsible for pension decisions that turned out badly is reinforced by the calls for increased financial education, implying that if individuals had been better informed they could have avoided or minimised their loss. This is true only in the sense that resistance to pension privatisation might have been more widespread had the voting public realised the risks; but having been hustled onto the Titanic of pension privatisation, with inadequate state pension lifeboats, few could avoid the disaster entirely. The emphasis on pension choices and financial education not only facilitates blaming the victims but also ignores the fact that vulnerable social groups have less advantageous choices available to them. This applies particularly to women. Due to the low level of the basic state pension and the brief time in which SERPS has operated, private pensions are a major source of gender and class inequality of retirement income (Arber and Ginn, 1991; Ginn and Arber, 1991, 1999). This is increasingly so as the basic pension declines, reducing its redistributive effects. The next section examines gender differences in private pension coverage, among those of working age. 24 Choke and risk in pensions The gender gap in private pensions The gender difference in private pension coverage among employees is stark, mainly because a higher proportion of women work part time. The reason follow membership of occupational pension schemes among part-timers differs according to the sector of industry (Ginn and Arber, 1993). Women employed part time in the private sector tend to work for an employer who does not operate a scheme; in 1987, 71% of women part-timers in this sector gave this as their reason for not belonging to an occupational pension scheme and part-timers are still concentrated in jobs where no occupational pension scheme is offered. In the public sector, where such schemes are more common, it was legal in the past for an occupational pension scheme's rules to exclude part-timers from membership. In 1987, 44% of women part-timers working in the public sector who were not members gave this as their reason (Ginn and Arber, 1993). European Court judgments deeming exclusion of part-timers to be indirect discrimination against women had some effect and in Britain it has been illegal since 1995 to discriminate against part-time employees in terms of access to occupational pensions. In 1995, 30% of British occupational pension schemes in the private sector of industry and 8% in the public sector still excluded some part-timers, the hours limit for eligibility varying among schemes (NAPF, 1996). Part-timers' occupational pension coverage has since increased. While this may help women in the future, many midlife and older women have lost the chance of accruing occupational pension entitlements for those periods of the lifecourse when they worked part time to accommodate the demands of childrearing. Even when women do join an occupational pension scheme they tend to derive less benefit from it than men for several reasons. First, many women leave their employment and hence the pension scheme for family reasons, long before normal retirement age. Their pension rights may then be 'preserved' in the scheme for later payment. Although a degree of inflation proofing is now required for preserved pensions, the entitlement will be much less than if the member had remained in the scheme and benefited from the rise in earnings. A large minority of early leavers receive no pension at all, having withdrawn their own contributions after one or two years' membership and forfeited the value of their employer's contributions. Thus over 30% of women aged 60-74 in 1994 who had joined an occupational pension were never able to draw a pension from it, compared to only 14% of men (Disney et al, 1997). Second, even women who stay in an occupational pension scheme until retirement receive less due to the gender pay gap and to periods of part-time employment. Third, women's tendency to have a flatter earnings profile with age means they gain less from a final salary pension scheme, in which the amount of pension depends on earnings in the last few years. Women's lower earnings relative to men also reduce their ability to make AVCs to an occupational pension scheme above the compulsory minimum (Price and Ginn, 2003). 25 Gender, pensions and the lifecourse Despite the wider access to private pension coverage offered by personal pensions, the gender gap remains substantial. In the mid-1990s, just over half of adults aged 20-59 were contributing to some form of private pension, 64% of men but only 38% of women (Ginn and Arber, 2000a; and see Table 2.1a). A third of adults contributed to an occupational pension scheme and 19% to a personal pension. Among employees, 81% of men but only 56% of women contributed to a private pension. Just over half (52%) belonged to an occupational pension scheme, 61% of men and 42% of women (see Table 2.1b). Only 23% of women part-timers were members. Personal pensions were held by 17% of employees, 20% of men and 14% of women. Thus among employees the gender gap in occupational pensions is replicated in personal pensions, with women's coverage only about 70% of men's. The gender difference among all adults is wider because of women's lower employment participation rates. Considering only gender differences in private pension coverage masks important variations among women. Later chapters explore variation according to ethnicity (Chapter Three) and partnership and maternal status (Chapters Four and Five). The next section examines class differences in men's and women's private pension coverage. Table 2.1: Percentage contributing to a private pension, women and men aged 20-59 a) All adults All Men Women b) Employees All Men Women FT PT All Has private pension 51 64 38 Employee,occupational 32 40 25 pension Employee, personal II 13 9 pension Self-employed, personal 6 9 2 pension Not employed, 2 2 2 personal pension 69 81 72 34 56 52 61 56 23 42 17 20 16 II 14 No private pension 49 36 62 Employee 19 12 27 Self-employed 4 6 Not employed 26 18 Column % 100 100 100 N= 24,069 1 1,756 12,313 31 19 28 66 44 31 19 28 66 44 3 32 100 100 100 100 100 15,056 7,603 7,453 4,260 3,175 Note: FT (full-time) employment is defined here as 31+ hours/week. Source: Ginn and Arber (2000a), using data from the GHS 1993-94 combined 26 Choice and risk in pensions Class differences in private pensions Whereas men and those previously in middle-class occupations generally have occupational pensions to cushion them from cuts in state pensions, this is less so for men and women in manual occupations. The class bias of occupational pensions has been well documented (Sinfield, 1978; Hannah, 1986; Arber, 1989; Ginn and Arber, 1991, 1993), and is additional to the effects of earnings, hours of work, sector of employment, and job duration. Among employees aged 20-59, manual workers in 1987 were far less likely to belong to a scheme than non-manual, even after controlling for all these factors as well as age group (Ginn andArber, 1993). The effect of occupational class was even more marked for women than for men. For example, the odds of membership for a woman in an unskilled manual occupation were only a fifth of the odds for a woman working in a professional occupation or as a manager in a large organisation and only a quarter of the odds for a woman in a routine non-manual occupation. One important reason for class inequality in occupational pension scheme membership is that certain types of employers, mainly larger organisations, are more likely to offer such a scheme. With the introduction of personal pensions, however, access to a private pension became theoretically available to all employees. Figure 2.1 shows differences in coverage by occupational and personal pensions among employees, according to socioeconomic category (SEC), gender and hours of work. Figure 2.1: Private pension contributions of employees by gender, socioeconomic category and hours of work of women 100 90 80 70 60 1 » 50 40 30 20 10 0 G Personal pension B Occupational pension Men Women full-time" Women part-time 111 12 3 4 5 12 3 4 5 6 Socioeconomic category 12 3 4 5 6 Note: For key to socioeconomic categories, see Table 2.2;a3l + hours per week. Source: Ginn andArber (1999), using data from the GHS 1993-94 combined 27 Gender, pensions and the lifecourse The small proportion of women employed as professionals and managers in large organisations (SEC 1) retained much of their advantage in private pension coverage even if they worked part time. Part-timers in this occupational group had a pension coverage rate (71%) exceeding that of women employed full time in routine non-manual occupations (SEC 3) or in manual occupations (SECs 4, 5 and 6). However, high occupational status is rare among part-timers. Only 13% of women employed as professionals and managers (SEC 1) worked part time, while half of women in routine non-manual occupations (SEC 3) did so. The class gradient in private pensions coverage was mainly due to occupational pensions, reflecting their differential availability, and the decline in coverage with socioeconomic group was steeper for women than for men. Personal pension coverage, in contrast, was more evenly distributed among socioeconomic groups, reflecting wider availability of this type of pension. However, personal pensions benefit those in higher socioeconomic groups and full-timers most. For part-timers and the low paid, predominantly women, the advisability of choosing a personal pension (including stakeholder pensions) is doubtful, especially if they cannot be sure of many years of high earnings in the future. Computer simulation of a range of hypothetical employment trajectories of women has illustrated this point (Falkingham and Rake, 2001). These authors estimate that a person retiring in 2050 would need to have earned above women's average wages and have been employed full time for 47 years in order to obtain a combined basic pension and stakeholder pension above the level of means-tested benefits - a tall order. Their calculations for the stakeholder pension were made before the 2001 fall in investment returns and hence should be regarded as optimistic. In the context of the shift in the public—private mix of pensions, it is important to understand which social groups opted to contribute to the different second tier pension schemes. The remainder of the chapter draws on research (Ginn and Arber, 2000a) focusing on how pension choices have been exercised by employees. Pension choices made by employees Personal pensions, introduced in 1988, were intended for employees who might benefit from a private pension, yet were unable to join an occupational pension scheme — either because their employer did not operate a scheme or because they were ineligible under the scheme's rules, as was the often the case for part-timers. Half a million employees were expected to sign up for a personal pension but due to the over-generous financial incentives offered in the late 1980s, five million were contributing to a personal pension by 1993 (DSS, 1994, p 9). Personal pension coverage among employees rose between 1988 and 1994 from 8 to 13% for women and from 15 to 22% for men; by 1995, coverage of working age employees was 28% for full-time men, 22% for full-time women and 11% for part-time women (ONS, 1997). An unexpectedly large number of occupational pension scheme members were persuaded to 28 Choice and risk in pensions switch to a personal pension in circumstances likely to provide poorer benefits - the notorious mis-selling of personal pensions, while others opted out of the State Earnings-Related Pension Scheme (SERPS). In order to understand what characteristics were associated with the likelihood of employees rejecting an occupational for a personal pension or choosing a personal pension in preference to SERPS, data from two years of the General Household Survey (GHS) combined, 1993 and 1994, were analysed (Ginn and Arber, 2000a). Employees'pension options are summarised graphically in Figure 2.2, which shows the proportions of employees taking each pathway. The 63% with access to an occupational pension scheme had three options: to belong to their employer's scheme (52%), to reject it for a personal pension (5%), or reject it and make no private pension contributions (6%). The 37% of employees lacking access had two alternatives: to contribute to a personal pension (12%) or not (25%). The majority of employees with no private pension arrangements must contribute to SERPS, although some, mainly women employed part-time, were paid too little to contribute to National Insurance. Figure 2.2: Pension arrangements of British employees aged 20-59 Employees 15,069 Access to occupational pension scheme 9,457 (63%) Excluded from occupational pension scheme 5,612(37%) Member of occupational pension scheme 7,772 Rejects occupational - pension for personal pension 770 Rejects occupational - pension stays in SERPS" 920 Excluded, contributes to personal pension 1,776 Excluded, stays in SERPS» 3,825 52% 5% 6% 12% 25% Note: "A minority currently paid no SERPS contributions because earnings were too low. Source: Ginn and Arber (2000a), using data from the GHS for 1993-94 combined 29 Gender, pensions and the lifecourse Three mam groups of employees can be distinguished: • Members: the 52% who belonged to an occupational pension scheme • Rejectors: the 11% who could belong to an occupational pension scheme but chose not to • Excluded: the 37% who could not join an occupational pension scheme. The characteristics of these three groups of employees are shown in Table 2.2. As expected, both access to an occupational pension scheme and membership were associated with being male, being aged over 30 (Table 2.2a) and with higher socioeconomic category (SEC) (Table 2.2b). Among employees with access to an occupational pension scheme, analysis showed that women, especially those working part time, were more likely than men to be rejectors (see Figure 2.3a). For men and for women employed full time, there was a near-linear relationship between age group and proportion of rejectors, while among women part-timers, the proportion of rejectors was high in all age groups at around 40%, only falling below 30% among those in Table 2.2: Percentage who were Members, Rejectors and Excluded, by age group and socioeconomic category, men and women employees aged 20-59 Men Women Women (full- time) (part-time) y ° if »