1 Retirement and pension arrangements in Britain Recent trends and prospects for the future James Banks and Richard Blundell Institute for Fiscal Studies and University College London British Association, Festival of Science, 2004

2 Three related questions: Financial security for the state? Financial security for employer pensions? Financial security for individuals? Underlying theme cohort differences ‘pensions not pensioners’

3 Sizable and largely unexpected increases in longevity The trends for men and women have resulted in a three-four year increase in life expectancy at 65 for men and women in Britain since the 1960s. With a more rapid increase for men in the UK in recent years. Pressures on pension systems

4 Additional years life expectancy, UK Males and Females aged 65 Source: GAD

5 Additional years life-expectancy at age 65, Men

6 Fertility rates have fallen too By 2020 more than 50% aged over 20 will be older than 50 Pressures on pension systems

7 Composition of the population aged 20 and over by broad age group, 1981 to 2051.

8 ILO activity rates, year old men Source: World Bank, 2001, Employment rates have fallen too:

9 ILO activity rates, year old men Source: World Bank, 2001, Employment rates have fallen too:

10 Actuarial forecasts have systematically underestimated longevity. For example, in 2000 the Government ’ s actuarial department predicted that there would be 16.1 million people aged 65 and over in Just two years later it revised this prediction up to 16.8 million. Pressures on pension systems

11 What should we expect of the future? With health improvements likely through the dramatic drop in smoking, these ‘ unexpected ’ increases may well continue. Could actuaries do better? Dora Costa ’ s work on early life effects on life expectancy, …. Whatever the case, unexpected increases in longevity put pressure on all pension systems.

12 The unique form of pension arrangements in Britain mean that these upward revisions in longevity, and general ageing of the population, play out in a relatively complex way. Pressures on pension systems

13 In Britain nearly 70% of pension scheme members are contracted-out of the state earnings related scheme (SERPS/S2P). An unusually high proportion, by international standards. => UK state pension ‘appears’ financially sustainable Pressures on pension systems

14 UK state pension ‘appears’ financially sustainable Tax increases not required, in contrast to rest of EU Achieved by reforms in 1980s and 1990s Large reductions in generosity of state pensions

15 Source: Emmerson and Johnson (2002) Future Costs of State Pensions

16 State pension entitlements over time:

17 Best deal for those reaching the state pension age between 1995 and Uncertainty about future state costs What will happen to other pension provision? Targeting and means-testing MIG/Pensioner credit

18 Until 1988 private pension saving in Britain was mainly held in employer based final-salary schemes or public sector final-salary (DB) schemes. Since 1988 individuals have been able to contract out into private ‘ defined contribution ’ (DC) pension schemes and now more than one third of this group are in such schemes. Pressures on contracted-out pensions

19 UK Second tier pension arrangements

20 What of the shift from DB to DC? the size of the DC pension depends directly on the contributions made, the return achieved on the investments held, and the annuity rate available at retirement.

21 What of the shift from DB to DC? the balance of second-tier state and private provision also matters, the UK is an extreme case. an individual who has been contracted out for their entire working life will only receive the Basic State Pension from the state which will be as low as 10% of average earnings by the year 2025 – contrast to US and Germany.

22 What of the shift from DB to DC? The trade-off between risks becomes much more apparent Longevity risk A DB plan essentially sets the annuity rate at the point when members join the scheme – say around age 30 an individual (in the UK) with a DC plan will need to annuitise their fund somewhere between the ages of 50 and 75 by which time some uncertainty about length of life may already have been resolved =>

23 What of the shift from DB to DC? Adverse selection: Consider individuals below aged 65 - what are the chances that they will live to age 75 on a response scale of ‘ very likely ’, ‘ likely ’, ‘ unlikely ’, ‘ very unlikely ’ or ‘ don ’ t know ’ very likely increases by 10 percentage points from age and don ’ t know/very unlikely fall back - also refer to ageing studies, ELSA, HRS.

24 How well will the annuity market cope with a large cohort retiring with defined contribution pensions? How important is adverse selection? How important is job/pension mobility? What of the shift from DB to DC?

25 What impact on pension incomes and retirement decisions? Even though the pressures from longevity were evident during the 1980s, the rise in equity prices over the 1990s managed to cloud this, at least as far as contracted-out schemes were concerned.

26 What impact on pension incomes and retirement decisions? While returns on the stock market were unusually healthy, company based schemes and individual defined contribution schemes displayed unprecedented growth. Quite generous retirement incomes and retirement windows were afforded despite the increased longevity. Note that around the same time (1998) SERPS reached full maturity.

27 What impact on pension incomes and retirement decisions? This resulted in a ‘ golden generation ’ of retirees, at the end of the last century, who retried on relatively generous pensions despite their increased longevity.

28 What impact on pension incomes and retirement decisions? The recent fall in equity prices and the squeeze on annuity rates has highlighted the pressures in a rather dramatic fashion. At the same time, as we have seen, the state system has become less generous. especially to those on relatively higher levels of earned income.

29 What impact on pension incomes and retirement decisions? This clearly spelled an end to the trend toward early retirement - employment rates among older men haven rise by more than 5 percentage points.

30 UK male employment rates,

31 UK male employment rates,

32 For those individuals in final salary employer (DB) pensions, the pressure on the fund means that firms are unwilling to offer generous early retirement windows. New entrants to firms are typically offered a defined contribution (DC) scheme rather than the company ’ s final salary scheme. What impact on pension incomes and retirement decisions?

33 Given the commitment to those already retired, this put increased pressure on those workers remaining in the company scheme. implicit risks in company DB schemes? For example, the Turner and Newall (T&N) pension scheme – current contributing members might receive 70% less than they were ‘promised’. Issues of governance? What impact on pension incomes and retirement decisions?

34 Those individuals with defined contribution schemes, whose fund was still in the stock market, felt the full impact of the increased individual risk of a DC scheme. In addition falling annuity rates made the income stream, from a now less valuable fund, even less generous. What impact on pension incomes and retirement decisions?

35 In both cases these pressures created an increased inducement not to retire early. This is theoretically strongest in those individuals with mainly DC based pensions. What impact on pension incomes and retirement decisions?

36 Proportion in ELSA reporting ‘low’ chances of employment after age 60, by pension type Men in paid employment, ages 50-54, 2002 ELSA data

37 Research using the English Longitudinal Study of Aging (ELSA) shows that conditioning on wealth, health and education, those aged in (predominantly final salary) employer final salary (DB) plans report on average 10 percentage points lower chances of working at age 60 that those with DC arrangements. What impact on pension incomes and retirement decisions?

38 ELSA also highlights the two different groups of those who leave employment early in the UK. What impact on pension incomes and retirement decisions?

39 Retirement, labour market inactivity and wealth Age group / wealth quintile Source: ELSA

40 Outcomes for future pensioners The poor Low life time income and high replacement rates The rich Unlikely to rely on state provision The middle Mix between state/private, DC/DB matters Marginal incentives will be affected What will adjust, and by how much?

41 Conclusions Individuals are likely to respond by delaying retirement – but incentives are strongest for those on middle earnings, and with DC pensions Savings are likely to respond to changes in retirement incomes, again among those with middle incomes – e.g. the impact of SERPS on savings But pension credit?

42 Conclusions Rising longevity: likely to strengthen trend toward later retirement, may encourage the development of new- style company/occupational pension schemes, in which retirement incomes are less linked to final salaries than in most existing DB schemes, but which offer employees greater protection against longevity risk and unfavourable movements in financial markets than most existing defined contribution schemes.

43 Retirement and pension arrangements in Britain The end of a brief summary James Banks and Richard Blundell Institute for Fiscal Studies and University College London British Association, Festival of Science, 2004

44

45 Patterns of activity and inactivity among 50– 69 year old men in the UK in 2000–01.

46 Expectations of working; Inactive individuals by age, gender and health status Men Women Source: ELSA

47 Composition of the population aged 20 and over by broad age group, 1981 to 2051.

48 Ageing of the EU population Source: EUROSTAT Old age dependency ratio

49 Source: DWP Income sources, by income group; **all individuals aged ?**

50 Percentage of men aged 50 to 64 who report being retired, by pension type and broad age group, Note: ‘Other’ category includes those with both types of pension, and those with unknown type. Source: Banks and Casanova (2003) using 2002 English Longitudinal Study of Ageing.

51 Distribution of income, by current pension type, Employees aged 50 and over, 2002 ELSA data

52 Figure 3: Longevity expectations by age

53 Source: Eurostat Projected public pension costs across Europe

54 Estimated substitutability of private saving for public pension wealth, by type of public pension wealth and broad age band

55 UK men No Occupational Pension With Occupational Pension Survival probabilities, by pension status

56