1 retirement and pension arrangements in britain recent trends and prospects for the future james...
TRANSCRIPT
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
Pressures on pension systems
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.
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
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
• 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 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?
• 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, 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
LFS 60-64 Men FES 60-64 Men
LFS 55-59 Men FES 55-59 Men
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
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.
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
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