1 retirement and pension arrangements in britain recent trends and prospects for the future james...

1

Retirement and pension arrangements in Britain

Recent trends and prospects for the future

James Banks and Richard BlundellInstitute 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

5

10

15

20

1841 1881 1921 1961 2001

men women

Additional years life expectancy, 1841-2001UK Males and Females aged 65

Source: GAD

5

Additional years life-expectancy at age 65, Men

10

11

12

13

14

15

16

17

18

1960 1970 1980 1990 2000

Year

Add

ition

al Y

ears

ove

r 65 Japan

France

US

Germany

UK

6

• Fertility rates have fallen too• By 2020 more than 50% aged over 20 will

be older than 50


7

0%

25%

50%

75%

100%

1981 1991 2002 2011 2021 2031 2041 2051

80+

70-79

60-69

50-59

20-49

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

8

ILO activity rates, 60-64 year old men

0

20

40

60

80

100

UK Swe Spain Neth Italy Ger Can Jap US

1960 2000

Source: World Bank, 2001, www.worldbank.org/pensions/statistics

Employment rates have fallen too:

9

ILO activity rates, 55-59 year old men

0

20

40

60

80

100

UK Swe Spain Neth Italy Ger Can Jap US

1960 2000

Source: World Bank, 2001, www.worldbank.org/pensions/statistics

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 2051.

• Just two years later it revised this prediction up to 16.8 million.


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.


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


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

0 2 4 6 8 10

SERPS with original formula,retirement ages and inheritance

rights

SERPS after the 1995 Act

State Second Pension after the1999 Act

Average annual real increase

Source: Emmerson and Johnson (2002)

Future Costs of State Pensions

16

State pension entitlements over time:

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

2055

2060

Per

cent

age

of a

vera

ge e

arni

ngs

at r

etire

men

t

Bas ic State Pens ion SERPSSERPS + Second State Pens ion Pens ion Credit Guarantee

17

• Best deal for those reaching the state pension age between 1995 and 2005.

• 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

0

5,000

10,000

15,000

20,000

25,000

1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001

Num

ber

of m

embe

rs (t

hou

sand

s)

Em ployer DC

Hybrid

Em ployer DB

Public DB

Individual DC

SERPS

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


• 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 30-63 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?


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


• 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


• 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


• 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, 1968-2003

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

1968

1975

1977

1979

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

Year

Pro

port

ion

LFS 60-64 Men FES 60-64 Men

31

UK male employment rates, 1968-2003

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

19

68

19

75

19

77

19

79

19

83

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

Year

Pro

po

rtio

n



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?


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.


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.


36

Proportion in ELSA reporting ‘low’ chances of employment after age 60, by pension type

0

10

20

30

40

None Other DefinedContribution

DefinedBenefit

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 50-54 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.


38

• ELSA also highlights the two different groups of those who leave employment early in the UK.


39

Retirement, labour market inactivity and wealth

0.0

20.0

40.0

60.0

80.0

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

Perc

en

tag

e

Retired Other inactive

50 to 54 55 to 59 60 to 64

Age group / wealth quintileSource: 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 BlundellInstitute for Fiscal Studies and University College London

British Association, Festival of Science, 2004

44

Estimates of wealth effects and pension plan effects on subjective probability of working at age 60;

Parameter Coefficient Std. Error Renter -6.84 2.70 ‘Wealth’ quintile 2 -4.83 2.63 ‘Wealth’ quintile 3 -1.30 2.65 ‘Wealth’ quintile 4 -7.06 2.66 ‘Wealth’ quintile 5 -8.24 2.70 Employer DB plan -3.54 2.05 Employer DC plan 4.33 2.53 Individual DC plan 7.62 2.33 Other controls include age dummies, gender, marital status, health and education. ELSA respondents aged 50-54.

45

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

0%

20%

40%

60%

80%

100%

50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69

Age

Working

Long-term sick / disabled

Retired

Unemployed

46

Expectations of working; Inactive individuals by age, gender and health status

0

5

10

15

20

25

50 to 59 60 to 64 50 to 54 55 to 59

Ave

rag

e e

xpe

catio

n o

f re

turn

ing

to w

ork

(%

ch

an

ce) excellent/very good/good fair/poor

Men Women

Source: ELSA

47

0%

25%

50%

75%

100%

1981 1991 2002 2011 2021 2031 2041 2051

80+

70-79

60-69

50-59

20-49

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

48

Ageing of the EU population

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0

ItalySweden

GreeceBelgium

Spain

GermanyFrance

UKAustria

Portugal

FinlandDenmark

LuxembourgNetherlands

Ireland

Percentage

2000 2040

Source: EUROSTAT

Old age dependency ratio

49

0

200

400

600

800

1,000

Poores

t 2 3 4

Riches

t

Poores

t 2 3 4

Riches

t

£ p

er w

eek,

200

1-02

pri

ces

Benefit income Occupational pensionInvestment income EarningsOther income

Source: DWP

1979 2001-02

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, 2002.

0.0

10.0

20.0

30.0

40.0

50.0

50-54 55-59 60-64 all 60-64Age

Per

cent

age

Defined Benefit Defined Contribution Other

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

0.0

005

.001

.0015

.002

.0025

Density

0 250 500 750 1000 1250 1500Net weekly earnings

None/dk DC

DB

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

52

Figure 3: Longevity expectations by age

How like ly do you think it is that you w ill live to be 75 or m ore?(very like ly, like ly, unlike ly, very unlike ly, don't know )

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64

Age (5 year bands)

Per

cent

age

of s

ampl

e

Very likelyDon't knowVery unlikely

Source: BHPS 2001 micro data

530 5 10 15 20 25

AustriaItaly

GreeceFrance

GermanyFinland

DenmarkBelgiumPortugal

SpainSweden

NetherlandsLuxembourg

Ireland

UK

% of national income

2000

2050

Source: Eurostat

Projected public pension costs across Europe

54

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

20-31 32-42 43-53 54-64Age

Su

bst

itutio

n

BSP SERPS

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

55

0

0.25

0.5

0.75

1

50 55 60 65 70

Age

Su

rviv

al

pro

bab

ilit

y UK men

No Occupational Pension

With Occupational Pension

Survival probabilities, by pension status

56

Table 1. Current and past private pension status, by age and gender

Percent currently contributing to: Per cent ever contributed to: None DB DC None DB DC Both Men 50-54 32.69 30.21 37.09 8.68 44.53 28.30 18.49 Men 55-59 45.54 20.44 34.03 11.51 45.44 25.69 17.36 Men 60-64 68.09 9.80 22.11 13.94 50.88 22.86 12.31 Women 50-54 53.38 25.81 20.81 36.08 37.10 19.06 7.77 Women 55-59 65.43 19.36 15.21 39.15 38.55 16.34 5.96

Source: Banks and Casanova, 2003 Calculations from 2002 ELSA microdata

1 retirement and pension arrangements in britain recent trends and prospects for the future james...

Documents