Raising household saving
Thomas F. Crossley, Carl Emmerson and Andrew Leicester
Launch event, British Academy, London, 22 February 2012
© Institute for Fiscal Studies
Overview
• Concern that a significant number of individuals might not be
saving enough for future needs
• What does economic theory tell us about why policymakers
might or might not be concerned about low saving?
• Assess the evidence on a number of possible policy responses:
– financial incentives
– information, education & training
– choice architecture
– social marketing
• Conclusions
© Institute for Fiscal Studies
1946
1948
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
Household saving rate (%)
Aggregate household saving rate
14
12
10
8
6
4
2
0
-2
-4
-6
Year
© Institute for Fiscal Studies
Notes & source: see Figure 1.1.
Why do people save – and why don‟t they? (1/2)
• Standard economic models suggest that individuals save when:
– income is high
– needs are low
– returns are high
• Might expect to see lower saving rates among those with
apparently low current income
– income might be temporarily low
– income might be mis-measured
• Passive and active saving
• Low returns for low income households could mean that low
saving might be privately optimal
© Institute for Fiscal Studies
Why do people save – and why don‟t they? (2/2)
• Individuals can only be expected to do the best they can with the
information that is available to them
• Role for better information
– avoid choice over-load
– how it is provided
– education to help individuals understand and use information
• Behavioural perspectives
– bounded rationality
– mental accounting
– loss aversion and reference points
– time inconsistency and self-control
© Institute for Fiscal Studies
Evaluation challenges
• Does an intervention boost saving?
– example: introduction of new savings account
• Two questions to ask of any study
– has an appropriate outcome been measured?
– has a credible counterfactual been estimated?
• If new saving is this only a short-run impact or does it lead to an
enduring impact on saving?
© Institute for Fiscal Studies
Financial incentives
• Financial incentive to save at all, and in different forms, affected
by the tax, tax credit and benefit system
• Clear evidence these affect the form in which savings are held
© Institute for Fiscal Studies
Responsible teenagers?
% of employees with a second-tier pension choosing to
contract out into a personal or stakeholder pension
70%
60%
All
Percentage
50%
40%
Under 20
30%
50 and over
20%
10%
Year
© Institute for Fiscal Studies
Notes & source: see Figure 3.2.
2003–04
2002–03
2001–02
2000–01
1999–00
1998–99
1997–98
1996–97
1995–96
1994–95
1993–94
1992–93
1991–92
1990–91
1989–90
1988–89
1987–88
0%
Financial incentives
• Financial incentive to save at all, and in different forms, affected
by the tax, tax credit and benefit system
• Clear evidence these affect the form in which savings are held
• Less clear is whether overall saving is increased
– majority of funds in tax-favoured accounts not new saving, but some
might represent new saving
• Those expecting to receive means-tested support in retirement
can have particularly weak financial incentives to save
– lack of evidence on whether this does lead to lower private saving
– further research could be fruitful, but difficult to identify who might be
affected by such policies
• Matched saving accounts attempt to target marginal saver
– limited evidence of impact on total saving
– no evidence on whether impact endures
© Institute for Fiscal Studies
Financial education
• No good evidence that adults‟ financial education raises savings
• Does educating children on financial matters affect adult saving?
– need long-term data on adult saving and wealth to assess
– with accurate information on financial education received at school
– and random variation in the provision of such education
• Closest is US study by Bernheim et al. (2001)
– survey of 2,000 adults aged 30 to 49
– variation in „consumer education‟ mandates at high school
– exposure increases saving rate and wealth holdings
• Policy could be informed by workplace financial training
– retirement planning seminars help raise retirement saving
– may also have small spillover effects
© Institute for Fiscal Studies
Providing information
• Little evidence on whether information alone affects savings
• Most compelling study suggests not (Choi et al. 2011)
– 689 workers in a firm not using full employer 401(k) match limits
– age and tenure meant no penalty for immediate withdrawal
– contributing to the limit raises wealth at potentially no cost
– half given this information, half not
– those treated raise contributions 0.1% more than those not
• Not just what information is provided but also how might matter
– simplified information might improve investment decisions
– evidence somewhat mixed
© Institute for Fiscal Studies
Choice architecture: changing default options
• Large US evidence base on „opt-in‟ defaults for pensions saving
• Some empirical regularities: the default matters!
1. Large increase in participation
2. Many workers stick to default contribution rate
3. The default investment fund is widely chosen
© Institute for Fiscal Studies
Madrian and Shea (2001)
%
Pensions saving behaviour 3–15 months after hire, single firm
Groups hired just before and just after move to opt-in default
100
90
80
70
60
50
40
30
20
10
0
Participation
© Institute for Fiscal Studies
3% (default)
contribution
Opt-out default
Opt-in default
6%+
contribution
Default fund
only
Choice architecture: changing default options
• Large US evidence base on „opt-in‟ defaults for pensions saving
• Some empirical regularities: the default matters!
1. Large increase in participation
2. Many workers stick to default contribution rate
3. The default investment fund is widely chosen
• Overall effect on total saving not clear
– some people will start to save, but others may save less
• „Active decisions‟ – make people choose
– smaller impact on uptake but less effect on contributions
• Default people into future increases in retirement saving?
– Thaler and Bernartzi (2004) “Save More Tomorrow”
– large rise in contributions but sometimes low take-up
© Institute for Fiscal Studies
Choice architecture: framing
• Presentation of savings options may affect choices (Saez 2009)
– offer random one-off inducement to open a retirement savings fund
– „match‟ 50% of contribution or „rebate‟ 33% - economically the same
– take-up 3.3% if no inducement, 6.4% for rebate, 10.2% for match
• Laboratory evidence that framing matters for:
– portfolio of retirement savings
– willingness to investing in risky assets
• Field evidence for these effects not clear
• Framing could be built into evaluation design for future policy
© Institute for Fiscal Studies
Social marketing
• Draw on techniques from marketing to promote social goals
• Kotler and Zaltman (1971) summarise key features
1. Identify target population
2. Understand barriers to behaviour change
3. Design, test and modify specific intervention based on barriers
• Application to retirement saving by Lusardi et al. (2009)
1. Target: new employees at a US firm, opening retirement account
2. Barriers: lack of information and complex online enrolment form
3. Intervention: „planning aid‟ leaflet breaking down application
process into simple steps and providing information
→ 41% of treatment group enrolled within 2 months, 28% control
• Government to help fund and evaluate different schemes?
© Institute for Fiscal Studies
Final thoughts and conclusions
• Significant gaps in the evidence base remain
– though obviously some individual high quality studies
• The main limitations are:
1. A lack of randomised variation to give credible counterfactuals
2. A failure to focus on comprehensive savings/wealth outcomes
3. A focus on short-run, not long-run outcomes
4. In general, a lack of UK-specific evidence
• An opportunity for policy makers
– growing recognition of need for gold-standard randomised trials
• Trials not always possible or appropriate
– need to develop models of behaviour change validated by evidence
– crucial in assessing possible impact of new proposals
© Institute for Fiscal Studies
Raising household saving
Thomas F. Crossley, Carl Emmerson and Andrew Leicester
Launch event, British Academy, London, 22 February 2012
© Institute for Fiscal Studies