Financing Retirement: Compulsion or Persuasion

1
Financing Retirement Compulsion or Persuasion?
Watson Wyatt Lecture

• John Piggott
• March 29 2004

2
Demographic Change

• Longevity
• Male life expectancy is over 75 in most OECD
  countries, an overall increase of over 12 per
  cent since 1960
• In Australia, life expectancy at 65 is over 20
  years for women, UK not much less.
• For a 65 and 60 year old male/female couple, 60
  chance one will live past 90 (based on Australian
  life actuarial tables)
• Changing Labour Habits
• Participation rate for males aged 60-64 have
  fallen to between 10 and 50 in the G7, from
  over 70 in 1970
• Fertility
• Has fallen to sub-replacement levels
• Result Aged Dependency ratios set to increase
  from 23 today to 42 in 2030.

3
Population aged 65 and over, by region (millions)
Source World Bank (1994)
4
Does voluntary retirement provision work?

• There is significant empirical evidence for
  undersaving
• If the government is concerned with
  undersaving, then one response is compulsory
  private provisioning publicly mandated forced
  saving
• New paper in US suggests drop in consumption
  even when retirement occurs as expected
• - between 7 and 11 reduction.
• Source Haider and Stephens 2004

5
Components of Retirement Provision
Universal
SAFETY NET
Targeted
PAYG
Publicly provided
Funded
COMPULSORY EMPLOYMENT RELATED
Privately managed
Privately provided
Publicly managed
Employment related
VOLUNTARY SAVING
Tax preferred
Other
Non tax preferred (private saving)
6
Components of Retirement Provision
PAYG
Publicly provided
Funded
COMPULSORY EMPLOYMENT RELATED
Privately managed
Privately provided
Publicly managed
7
Pension Funding Crisis?
Given the predominance of unfunded retirement
schemes (almost half of all mandatory pension
plans), demographic change has resulted in a
funding shortfall.
Source Palacios and Pallares-Miralles (2000)
8
Pension Funding CrisisStatistics
Source Disney (2000) and The (Australian)
Treasury (2002)
9
Solutions

• Given the funding crisis, governments can
  undertake
• Parametric reform
• Substantial increases in taxation to fund current
  benefits
• maintain current system but reduce benefits
• Structural reform (substantially increased
  private provision)
• Compulsion?
• Persuasion?

10
World Banks Views
The World Bank advocates a multipillar approach.
The public pillar, funded by payroll taxes or
general revenues, focuses on redistribution. The
second pillar is a funded
system, where individuals' mandatory
contributions are saved and invested to pay their
future pensions. This generally does not
redistribute. And finally, mandatory pensions
should be small enough to allow growth of a third
pillar, voluntary savings, allowing individuals
to choose how to allocate their income over their
lifetime.
- World Bank Website
11
Components of Retirement Provision
PAYG
Publicly provided
Funded
COMPULSORY EMPLOYMENT RELATED
Privately managed
Privately provided
Publicly managed
12
Recent International Growth of mandatory DC
schemes
1981 Chile 1985 Switzerland
Netherlands 1986-92 Australia 1993
Argentina Columbia, Denmark, Peru 1996
Uruguay 1998 Hungary Kazakhstan Bolivia
Mexico 1999 El Salvador Poland 2000 Hong
Kong Sweden 2001 Latvia 2003 Dominican
Republic
439 million people potentially affected since
1980 India has moved towards compulsory DC
schemes for state employees China is playing with
the idea
Not strictly compulsory
13
International ExperienceChile

• Transfer to full privatisation in 1981
• Compulsory for all workers 95 coverage
• 10 of wages invested in private accumulation
  accounts
• Regulated investment choice
• Government guaranteed annual returns (in range)
• Administered in private sector by AFPs
• Indexed annuity or phased withdrawal only, up to
  a ceiling

14
International ExperienceSingapore

• Mandatory saving in accumulation accounts 98
  coverage

• Publicly administered Central Provident Fund set
  up in 1955
• Crediting rate very conservative
• Fund used for other purposes, esp housing
• 30 - 40 mandatory contribution to a ceiling
  split equally between employer and employee
• Replacement rate only 20

15
International ExperienceAustralia

• 1986 Accord compulsory saving arose as part of
  union wage deal
• Superannuation Guarantee Act 1992
• 9 of earnings, phased in to 2002 employer
  mandate
• Around 90 of Australias workforce is covered
• Administered in the private sector
• No restrictions on portfolios or payouts

16
Assets
17
Asset Composition
18
Contributions and Benefits
19
Returns
20
Outcomes

• The story so far
• Average income of working person 50 to 55 is
  52 000 compared to 16 000 for retired
  person.
• The average superannuation balance for those
  retired and aged 50 to 55 is 24 000, overall
  average for those aged 50 to 69 is 83 000 and
  the proportion of retirees with more than 100
  000 is around 20.
• The family home by far remains the largest
  component of household wealth average home
  equity is 240 000
• Lump sums used to pay off mortgage at retirement
• Estimates from AMP.NATSEM Income and Wealth
  Report No. 7.

21
PitfallsDesign

• Getting the arithmetic right
• x contribution rate gt what is x?
• x of what?
• x goes where?
• Earnings rates?
• Net of taxes, charges (regulatory costs, asset
  controls)
• Integrating benefits with public programs to
  achieve adequate longevity insurance

22
PitfallsSubstitution

• People could offset compulsory saving by reducing
  saving elsewhere or by borrowing
• Liquidity constraints (or debt aversion) and
  mental accounting are inhibitors to full
  substitution
• Evidence
• Australia 62 of compulsory saving is
  additional household saving rate increased by 2
  percentage points Connolly and Kohler 2004
• Chile national saving increased by more than 2
  percentage points Coronado 1997

23
PitfallsAdministrative Costs

• Both economies of scale in administration and
  incentive structures matter.
• Significant variability of charges both across
  and within different regimes
• Size and extent of costs significantly affect
  long term accumulations, reducing annual returns
  by up to 2 percentage points, and final
  accumulations by up to 35
• Small annual charge differences (as of assets)
  can almost double the final gross charge
• However, higher fees may be worth it

24
International Charge Ratios in Funded Pension
Schemes
Source Whitehouse 2000
25
Pitfalls Governance and Mandates

• Plan Governance
• ________________________
• Corporate Trustee
• Who pays? ________________________
• Employee OK OK
• Employer Not OK OK

26
PitfallsTaxation

• What to tax?
• Contributions, Investment Earnings or Final
  Benefits?
• Australia taxes all three, Singapore none
• Earnings taxation can distort funds investment
  choices
• Contributions tax can discourage additional
  employee contributions, and are less easily
  integrated into the progressive tax system
• Relationship with other taxes?
• Favourable taxation of other assets can affect
  efficacy of compulsory program (housing etc)

27
Fees and taxes adding insult to injury?
Source Bateman and Piggott (2001)
28
Pitfalls Benefit Design
U.S. Health and Retirement Study 1992-2000, with
over 12 000 respondents
69 of those who rely on DB pensions are happy
with their retirement only 54 without annuity
income After 10 years of retirement, those with
annuity incomes are 45 more likely to be very
satisfied Retirees with annuity income streams
are 30 more likely to have no depression symptoms
Source Panis (2003)
29
PitfallsPolitics

• Asset base very tempting to either
• Tax
• Return to accountholders prior to preservation
  age
• Use for purposes other than retirement provision
• Weight of pensions taxation front loaded
• then later, more tax?

30
Lessons from Behavioural Finance
Could Too Much Choice in fact be
discouraging? Two exotic jam tasting booths set
up near supermarket
One with 6 and the other with 24 different
jams 60 of passers-by were attracted to the
second booth However, only 3 made a purchase,
versus 30! Student Essays, choice of 6 topics
versus 30 produced more and better quality essays
than students facing the longer list
Source Sethi-Iyengar (2003)
31
Lessons from Behavioural Finance
What impact do co-workers have on investment
decisions? Employees from some university
departments randomly invited to retirement
seminar
Attendance rate of employees not receiving
invitation was three times higher than those from
departments in which no-one received the letter
Moreover, 401(k) plan sign-up rate for these
employees were 20 higher, even higher than that
of those who received an invitation!
Source Duflo and Saez (2003)
32
Lessons from Behavioural Finance
Default choice and naïve diversification
rational investors? Contributors exhibit a 1/n
heuristic rather than modern (rational) portfolio
theory rules
Choice of funds offered has a strong influence on
asset allocation TWA 401(k) contributors invest
75 in stocks, University of California staff
invest 34 in stocks Simply replacing a bond fund
with equity fund can increase allocation to
equities by 3.67 Participation in 401(k) plans
can more than double if employees are
automatically enrolled
Source Benartzi and Thaler (2001)
33
Take-aways

• Compulsion wont work well on its own
• Starting point important
• Private behavioural response
• Administrative costs and industry positioning
• Political hazard
• But it does have some positives
• Long term, provides a way of coping with
  demographic imbalance
• Short term, may help economy-wide asset
  allocation
• In mature mandated systems, evidence of higher
  saving

34
Take-aways

• What lessons are there?
• Program design must be compatible with incentives
• Regulation, incentives and industry structure
  must all combine to keeping admin costs low
• The arithmetic must work
• There is potential for behavioural finance
  techniques to be harnessed to buttress compulsion
• Benefit income streams should be encouraged

35
John Piggott March 2004 Email
j.piggott_at_unsw.edu.au