Funding gratefully acknowledged from the ESRC and members of the IFS Retirement
Savings Consortium
© Institute for Fiscal Studies
Saving for retirement in a world with automatic enrolment
Pensions Commission recommended three key policy changes in response to under-saving for retirement
• Earnings indexation of flat rate state pension
• Higher state pension age
• Automatic enrolment
How well prepared will the current working age population be for retirement?
What have we learned over the last 2 years that helps us to answer this question?
Saving for retirement and automatic enrolment: what have we learned?
“triple lock”
What have we learned?
1) Younger generations have lower wealth than older generations did at similar ages.
2) Automatic enrolment massively boosts pension participation. Mostly at very low contribution rates, but some with larger contributions.
3) Automatic enrolment boosts coverage by different amounts for different groups. And also boosts participation among "ineligibles".
4) Many people are not good at estimating their life expectancy.
Saving for retirement and automatic enrolment: what have we learned?
1) Younger generations have lower wealth thanolder generations had at similar ages
Saving for retirement and automatic enrolment: what have we learned?
Median net household wealth by age for different generations
Saving for retirement and automatic enrolment: what have we learned?
0
50
100
150
200
250
300
25 30 35 40 45 50 55 60
Med
ian
net
hous
ehol
d w
ealt
h pe
r ad
ult (
£’00
0, 2
014–
15 p
rice
s)
Age
1960s
Early 1980s
1970s
1950s
Source: IFS calculations using Wealth and Assets Survey. (Figure 5 of Cribb, Hood and Joyce 2016)
The Economic Circumstances of Different Generations © Institute for Fiscal Studies
2) Automatic enrolment massively boosts pension participation.
Mostly at very low contribution rates, but some with larger contributions.
The Economic Circumstances of Different Generations © Institute for Fiscal Studies
Automatic enrolment: how it works
Employers have to:
‒ enrol all eligible employees into a workplace pension
‒ with at least minimum levels of total and employer contributions
Employees can then choose to leave the pension if they wish
Automatic enrolment has been rolled out from the largest employers to the smallest employers
Minimum pension contributions have started to rise
‒ 2% of qualifying earnings (1% from employer) up to March 2018
‒ 5% of qualifying earnings (2% from employer) in 2018-19
‒ 8% of qualifying earnings (3% from employer) from 2019-20
Overall workplace pension participation
0
20
40
60
80
100
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Perc
ent
Source: Crawford, Cribb and Emmerson (2018) using data from the Annual Survey of Hours and Earnings.
Saving for retirement and automatic enrolment: what have we learned?
Private sector
Public sector
Workplace pension membership of eligible employees, by employer size
0
10
20
30
40
50
60
70
80
90
100
2009 2010 2011 2012 2013 2014 2015 2016
Perc
ent i
n a
wor
kpla
ce p
ensi
on 30,000+
6,000-29,999
350 to 5,999
160 to 349
58 to 159
50 to 58
30 to 49
2 to 29
Source: Cribb and Emmerson (forthcoming) using data from the Annual Survey of Hours and Earnings.
Saving for retirement and automatic enrolment: what have we learned?
Effect of automatic enrolment on total pension contributions
0
5
10
15
20
25
30
35
40
45
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
Effe
ct o
f aut
o en
rolm
ent (
ppt)
Total contribution rate (% of earnings)
Source: Figure 3 of Cribb and Emmerson (2016) using data from the Annual Survey of Hours and Earnings.
Saving for retirement and automatic enrolment: what have we learned?
32 ppt increase in % with contributions >1% of earnings
6ppt increase in % with contributions >5% of earnings
2ppt increase in % with contributions >10% of earnings
The Economic Circumstances of Different Generations © Institute for Fiscal Studies
3) Automatic enrolment leads to many more with contributions above the minimums and higher pension
participation of “ineligibles”
Workplace pension membership of eligible employees by age
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
22 to 29 30 to 39 40 to 49 50 to SPA
2012 2015
Saving for retirement and automatic enrolment: what have we learned? Source: IFS calculations using Annual Survey of Hours and Earnings. (Table 3 of Cribb and Emmerson 2016)
Workplace pension participation of eligible employees, by earnings level
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
Lowest 25% Second 25% Third 25% Top 25%
Quartile of earnings distribution of eligible employees
2012 2015
Saving for retirement and automatic enrolment: what have we learned? Source: IFS calculations using Annual Survey of Hours and Earnings. (Table 3 of Cribb and Emmerson 2016)
The Economic Circumstances of Different Generations © Institute for Fiscal Studies
Effect of automatic enrolment on “ineligibles”
0 5 10 15 20 25 30
0 to 2 months’ job tenure
Aged under 22
Aged over state pension age
Under earnings threshold
All not eligible
Effect of automatic enrolment on workplace pension participation (ppt)
Source: Table 5 of Cribb and Emmerson (2016) using data from the Annual Survey of Hours and Earnings.
The Economic Circumstances of Different Generations © Institute for Fiscal Studies
Effect of automatic enrolment on “ineligibles”
0 5 10 15 20 25 30
0 to 2 months’ job tenure
Aged under 22
Aged over state pension age
Under earnings threshold
All not eligible
Effect of automatic enrolment on workplace pension participation (ppt)
Source: Table 5 of Cribb and Emmerson (2016) using data from the Annual Survey of Hours and Earnings.
What will happen as contributions rise?
Increases in minimum contribution rates have started and they will rise again next year
From “mid 2020s” all earnings up to upper threshold to count towards minimum contributions
‒ Also widen age band from 22 to SPA to 18 to SPA
‒ Has potential to benefit people earning low amounts in multiple jobs
Saving for retirement and automatic enrolment: what have we learned?
Higher minimum contributions to workplace pensions
Will more people choose to leave their pension scheme?
Reasons for government to be optimistic:
‒ Many already have contributions in excess of long run minimums
‒ Some (many?) will remain in scheme simply due to inertia
‒ Individual would have to give up a larger employer contribution
‒ Large firm in the USA with 3% employee and 3% employer cont. showed similar opt out rates to UK (Madrian and Shea 2001)
Saving for retirement and automatic enrolment: what have we learned?
Higher minimum contributions to workplace pensions
Reasons for government to be cautious:
‒ Where the higher employer contributions are “coming from”? (lower wages, lower profits, higher prices or higher productivity)
‒ Are households reducing savings elsewhere?
‒ Are the “right” employees opting out?
‒ Will the higher minimum contributions be “high enough”?
Saving for retirement and automatic enrolment: what have we learned?
4) Many people are not good at estimating their life expectancy
Saving for retirement and automatic enrolment: what have we learned?
Life expectancy and saving for retirement
Life expectancy can be an important consideration when planning for retirement
‒ How much to save for retirement?
‒ When will they retire?
But estimating your own life expectancy is very difficult
‒ Unrealistic expectations about chances of survival will make it difficult to plan well for retirement
IFS research has compared “subjective” expectations of surviving to given ages and “objective” measures from life tables
Saving for retirement and automatic enrolment: what have we learned?
Survival expectations and rates for men, born 1930-39
Saving for retirement and automatic enrolment: what have we learned?
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
60 65 70 75 80 85
Prob
. sur
vive
to
at le
ast
that
age
Age of respondent
Age 75
Age 80
Age 90
Age 85
Age 95
Dotted lines: from life tables
Dashed lines: subjective expectations
Survival expectations and rates for men, born 1930-39
Saving for retirement and automatic enrolment: what have we learned?
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
60 65 70 75 80 85
Prob
. sur
vive
to
at le
ast
that
age
Age of respondent
Age 75
Age 80
Age 90
Age 85
Age 95
Dotted lines: from life tables
Dashed lines: subjective expectations
Survival expectations and rates for men, born 1930-39
Saving for retirement and automatic enrolment: what have we learned?
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
60 65 70 75 80 85
Prob
. sur
vive
to
at le
ast
that
age
Age of respondent
Age 75
Age 80
Age 90
Age 85
Age 95
Dotted lines: from life tables
Dashed lines: subjective expectations
Survival expectations and rates for men, born 1930-39
Saving for retirement and automatic enrolment: what have we learned?
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
60 65 70 75 80 85
Prob
. sur
vive
to
at le
ast
that
age
Age of respondent
Age 75
Age 80
Age 90
Age 85
Age 95
Dotted lines: from life tables
Dashed lines: subjective expectations
Survival expectations and rates for men, born 1930-39
Saving for retirement and automatic enrolment: what have we learned?
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
60 65 70 75 80 85
Prob
. sur
vive
to
at le
ast
that
age
Age of respondent
Age 75
Age 80
Age 90
Age 85
Age 95
Dotted lines: from life tables
Dashed lines: subjective expectations
Survival expectations and rates for men, born 1930-39
Saving for retirement and automatic enrolment: what have we learned?
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
60 65 70 75 80 85
Prob
. sur
vive
to
at le
ast
that
age
Age of respondent
Age 75
Age 80
Age 90
Age 85
Age 95
Dotted lines: from life tables
Dashed lines: subjective expectations
Survival expectations and rates for men, born 1930-39
Saving for retirement and automatic enrolment: what have we learned?
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100%
60 65 70 75 80 85
Prob
. sur
vive
to
at le
ast
that
age
Age of respondent
Age 75
Age 80
Age 90
Age 85
Age 95
Dotted lines: from life tables
Dashed lines: subjective expectations
Implications of biased expectations
Saving for retirement and automatic enrolment: what have we learned?
Subjective expectations of survival show systematic biases • Significant survival ‘pessimism’ about surviving until 70s and early 80s • Over optimistic about survival to very oldest ages Concerns about savings for retirement and spending at older ages • Pessimism may mean ‘too low’ saving & ‘too fast’ spending in 60s/70s • Optimism at oldest ages may mean over-reluctance to spend Survival pessimism could explain unpopularity of annuities • Around two thirds of individuals could view an ‘actuarially fairly’ priced
annuity as offering an unfairly low rate
Four things we have learned so far
Saving for retirement and automatic enrolment: what have we learned?
1) Younger generations have lower wealth than older generations did at similar ages.
2) Automatic enrolment massively boosts pension participation. Mostly at very low contribution rates, but some with larger contributions.
3) Automatic enrolment boosts coverage by different amounts for different groups. And also boosts participation among "ineligibles".
4) Many people are not good at estimating their life expectancy.
Four questions for the future
1) Will lower rates of homeownership persist, and what are their implications for retirement resources?
2) How could automatic enrolment work for those working for their own business (i.e. self employed)? And would it be a good thing for them?
3) How high should government set minimum contributions? Will they have to trade off higher contributions with lower participation?
4) To what extent do individuals have the information they need to make good plans for their retirement?
Saving for retirement and automatic enrolment: what have we learned?
What did we know?
Older households, on average, hold a lot of wealth
• Among those aged 55-64 (i.e. on the eve of retirement):
‒ Median household non-pension wealth £250,000
‒ 60% primary housing, 22% financial assets, 11% other property
‒ Very unequal: least wealthy 10% have nothing on average, most wealthy 10% have on average in excess of £1m
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
What did we know?
Older households, on average, hold a lot of wealth
Individuals expect to use many sources to provide retirement income
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Expected sources of money in retirement
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
25 30 35 40 45 50 55 60
Perc
enta
ge o
f ind
ivid
uals
Age
State pension
Private pension
Savings
Primary housing
Future inheritance Other property
Notes: Lines are drawn for 5-yr date-of-birth cohorts. Calculated using Wealth and Assets Survey waves 1-5
Source: ‘The use of wealth in retirement’, IFS Briefing Note 237, Figure 2
What did we know?
Older households, on average, hold a lot of wealth
Individuals expect to use many sources to provide retirement income
Q: How do individuals currently draw on their wealth in retirement?
We studied this question using data from the English Longitudinal Study of Ageing (ELSA)
• Repeated observations of the same individuals over time
• Data covers the period 2002/03 to 2014/15
• Some follow-up interviews with proxies after individuals’ death
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Use of primary housing wealth
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
How common is moving among home owners?
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
0%
2%
4%
6%
8%
10%
12%
Age
Proportion who move over the next two years:
All moves
Non-institutional moves
Source: ‘The use of housing wealth in retirement’, IFS Briefing Note 235, Figures 3 and 4.
Cumulating over retirement suggests…
of owners at age 50:
⅓ would move by age 70 over ½ would move by age 90
weighted by life expectancy
40% would move before death
Are people releasing housing wealth?
Moves out of owner occupation? • 77% of moves (83% of non-institutional moves) are to another owner
occupied property • Current trends suggest 14% of home owners at age 50 would move out of
owner occupation before death
‘Downvaluing’? • On average those who move within owner occupation do release some
housing wealth - amounts (£ and %) increasing with age • Financial situation strongly correlated with ‘downvaluing’
‒ Those with the lowest financial wealth access more housing wealth when move, as do those with greatest housing wealth to income ratio
Taken together, perhaps around two-thirds of moves release some (but not necessarily much) housing wealth
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Implications
For the time being most housing wealth looks set to be bequeathed
Findings suggest individuals are prepared to access housing wealth
Might expect accessing of housing wealth to increase in future
• Greater proportion of working-age expect to downsize than the proportion of current retirees reporting moving for financial reasons
• If future generations have less pension saving then they may be more likely to release wealth, or release more wealth, when they move
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Use of financial wealth
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Trajectory of financial wealth in retirement
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
£0
£5,000
£10,000
£15,000
£20,000
£25,000
£30,000
£35,000
£40,000
68 72 76 80 84 88 92
Med
ian
real
net
fina
ncia
l wea
lth
Average age of cohort
1930-34 1925-29 1920-24
Notes: Sample stays the same within each line. Each point represents data from a particular wave of ELSA, average wealth plotted against
average age for each 5-yr birth cohort.
Source: ‘The use of financial wealth in retirement’, IFS Briefing Note 236, Figure 3
Trajectory of financial wealth in retirement
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
£0
£5,000
£10,000
£15,000
£20,000
£25,000
£30,000
£35,000
£40,000
68 72 76 80 84 88 92
Med
ian
real
net
fina
ncia
l wea
lth
Average age of cohort
1930-34 1925-29 1920-24
Notes: Sample stays the same within each line. Each point represents data from a particular wave of ELSA, average wealth plotted against
average age for each 5-yr birth cohort.
Source: ‘The use of financial wealth in retirement’, IFS Briefing Note 236, Figure 3
31% decline between age
70 and 90
Average remaining life expectancy declines by 75%
Least wealthy half: 13% decline (e.g. from £8,000 at age 70) Most wealthy half: 39% decline (e.g. from £80,000 at age 70)
Implications
Majority of financial wealth looks set to be bequeathed rather than spent in retirement
Lack of drawdown of financial wealth does not necessarily imply current retirees are making “bad” choices
• Likely largely driven by precautionary saving and bequest motives
• But could indicate an adverse effect of low public provision of social care and lack of private insurance
Does not necessarily mean future generations do not need to accumulate so much
• Could IF driven by bequest motives and future generations have smaller desire to leave money to children
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Use of other property wealth
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Prevalence of second home ownership
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
0%
2%
4%
6%
8%
10%
12%
14%
16%
56 60 64 68 72 76 80 84 88 92
Perc
enta
ge o
f ind
ivid
uals
Average age of cohort
1940-44
1935-39
1930-34
1925-29
1920-24
Notes: Calculated using ELSA 2008-09 to 2014-15. Excludes properties worth less than £50,000.
Source: ‘The use of wealth in retirement’, IFS Briefing Note 237, Figure 5
Prevalence of second home ownership
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
0%
2%
4%
6%
8%
10%
12%
14%
16%
56 60 64 68 72 76 80 84 88 92
Perc
enta
ge o
f ind
ivid
uals
Average age of cohort
1950-54
1945-49
1940-44
1935-39
1930-34
1925-29
1920-24
Notes: Calculated using ELSA 2008-09 to 2014-15. Excludes properties worth less than £50,000.
Source: ‘The use of wealth in retirement’, IFS Briefing Note 237, Figure 5
Prevalence of second home ownership
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
0% 1% 2% 3% 4% 5% 6% 7% 8% 9%
10%
56 60 64 68 72 76 80 84 88 92
Perc
enta
ge o
f ind
ivid
uals
Average age of cohort
0% 1% 2% 3% 4% 5% 6% 7% 8% 9%
10%
56 60 64 68 72 76 80 84 88 92 Average age of cohort
Receives income or rent from property: Does not receive income or rent:
Source: ‘The use of wealth in retirement’, IFS Briefing Note 237, Figure 5
Implications
Second homes (even non-income yielding ones) held at ages 70+ are not generally being sold to finance retirement
Differences in prevalence between generations
• With higher level of ownership, selling to fund retirement could be more common
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
End-of-life expenses and bequests
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
End-of-life expenses and bequest behaviour
Do not appear to be large end-of-life expenses that would use up remaining wealth holdings of most individuals • 7% of deceased individuals received assistance with daily activities from a
privately paid employee in the run up to death • 21% stayed in a nursing home in the last two years of life
‒ 32% of whom for 6 months or more • Median out-of-pocket cost for funeral expenses only £1,700 in 2002-03
Bequests are normally made to multiple recipients
Inheritances are likely to be received at relatively older ages • Bequests typically only made when last partner in a couple dies • Majority of cases do not see inheritances given directly to grandchildren
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
What we know now
Retired individuals did not draw on their wealth much over the period 2002-03 to 2014-15
• Two-fifths of home owners at 50 projected to move before death
• Moves rarely motivated by financial reasons, but do on average release some wealth
• Financial wealth (even among the wealthy) drawn down much slower than decline in (expected) remaining life
• Prevalence of second home ownership at ages 70+ pretty stable
Not, in general, big expenses associated with death
Bequests normally made to multiple individuals, and only when no surviving partner
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
Overall implications
Most wealth held by current retirees looks set to be bequeathed • Implications for the level and distribution of resources among current
working age individuals • Though inheritances likely to be (directly) received at relatively older ages
Can this shed light on likely responses to pension freedoms? • Behaviour with respect to main pension resource could be different to
current use of liquid financial wealth • Results likely to provide more reassurance to those worried about
spending too quickly than those worried about spending too slowly
Careful monitoring of how the use of wealth evolves in future important • For the living standards of both retirees and younger generations
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications
11 June 2018
RICS, London
WIFI: RICS Guest network (no password required)
The ins and outs of retirement saving
Panel discussion
Given the current policy context, what are the key challenges for government, the pensions industry and individuals in trying to
ensure decent retirement outcomes?
• John Godfrey Corporate Affairs Director at Legal and General, former head of Downing Street Policy Unit
• Professor John Hills London School of Economics, and former Pensions Commissioner
• Laura Webster Chief Economist and Lead Analyst for Private Pensions, Department for Work and Pensions
© Institute for Fiscal Studies Use of wealth in retirement - evidence and implications