Pension Reforms and the Allocation of Retirement Saving

1
Pension Reforms and the Allocation of Retirement
Saving

• Renata Bottazzi
• University of Bologna, IFS and CHILD
• Tullio Jappelli
• University of Naples Federico II, CSEF and CEPR
• Mario Padula
• University Ca Foscari of Venice and CSEF
• Prepared for the Annual Conference on Financial
  Security in Retirement
• 18-19 September 2008

2
Motivation

• Assess people awareness of retirement outcome
  innovations.
• What do people know about their pensions?
• Provide the anatomy of the offset between social
  security and private wealth.
• If social security wealth falls, do people
  increase more financial or real wealth?
• Study the demand for targeted retirement
  products.
• Why are Italian pension funds still small?

3
The framework

• We exploit a decade of pension reforms,
• use data on subjective probabilities on
  retirement outcomes,
• and look at several components of private wealth,
  including financial (risky and safe) and real
  (business and housing) wealth.

4
Main results

• Large revision of pension expectations, but many
  individuals have not completely updated their
  expectations yet.
• Financial and real wealth have increased
  following the reforms, but the increase is more
  pronounced for real assets and housing in
  particular.
• No effect on the propensity to hold targeted
  saving plans.

5
Offset and portfolio choice

• Standard life-cycle framework if social security
  wealth falls, private wealth should increase
  accordingly.
• In a complete market world, the reduction of
  social security wealth should not affect
  portfolio rules.
• With uninsurable income risk, borrowing (and
  short sale) constraints, portfolio rules become a
  function of age and wealth.

6
Background literature

• Estimate social security by
• using current and projected legislation on
  pension eligibility Gale (1998), Gruber and Wise
  (1999), Attanasio and Brugiavini (2003),
  Attanasio and Rohwedder (2003)
• using subjective expectations of retirement age
  and benefits Bernheim (1990), Gustman and
  Steinmeier (2001), Bottazzi, Jappelli and Padula
  (2006)
• The effect of the 1992-95 reform
• Miniaci and Weber (1999) 1993 consumption drop
  partly due to the 1992 pension reform
• Attanasio and Brugiavini (2002) offset between
  private saving and pension wealth (coefficient of
  -0.3)

7
The Italian pension reforms

• Three main reforms (1992, 1995, 1997)
• Features
• retirement age and minimum years of contributions
  for pension eligibility
• abolishment of seniority pensions (if started
  working after 1995)
• indexation of pension benefits to prices instead
  of wages
• less generous pension award formulas

8
The Italian pension reforms

• The eligibility rules and the pension award
  formula change
• according to the years of contribution at the end
  of 1992
• Three groups of workers
• Old, more than 18 years of contribution as of
  31/12/1995
• Middleaged, less than 18 years of contribution
  as of 31/12/1995
• Young, enter labor market in 1996

9
The Italian pension reforms retirement age
10
The Italian pension reforms the pension award
formula
11
Data

• Survey of Household Income and Wealth (SHIW)
  representative of Italian population
• Subjective expectations on retirement age and
  replacement rate
• Retirement age - all survey years
• - When do you expect to retire?
• Replacement rate - years 1989, 1991, 2004, 2006
  Consider the moment when you will retire.
  Setting your final monthly income before
  retirement equal to 100, what do you expect your
  first monthly pension to be?

12
Expected pension wealth at retirement

• Use subjective expectations on retirement age and
  replacement rate to construct the ratio of
  expected pension wealth at retirement to earnings
  (evaluated at each survey yeart)

expected retirement age
expected replacement rate
P()survival probability (by age and gender,
before and after the reform) ggrowth rate
rinterest rate
13
Social security wealth
14
Expectation error distribution of social security
wealth before and after the reform
15
Trends in financial and real wealth
16
Offset between pension wealth and private wealth
the estimating equation
Financial (Real) wealth -to-income ratio
Year, cohort, and employment dummies and their
interaction, dummies, age and education of the
hh, area of residence
Social security wealth-to-income ratio

• SSWY - weighted sum of both partners SSWY
• - Gale adjustment factor
• - Instrument with the statutory SSWY
• - Sample split Informed (Expectation error less
  than the median) vs. Uninformed

17
Offset between pension wealth and private wealth
the results
18
The anatomy of the offset
19
The anatomy of the offsetInformed vs. Uninformed
20
Summary

• A reduction in social security wealth equivalent
  to 1 years income brings about an increase in
  financial wealth of just below 1 months income,
• and an increase in real assets of about 6 times
  monthly income.
• The effect is larger for the Informed and for
  housing wealth.
• Other possible channels through which retirement
  saving increases

21
Targeted retirement saving products
22
Conclusions

• Large revision of expected social security
  wealth.
• Larger effects of reform for real, smaller for
  financial wealth.
• Information on pension outcomes is still
  important,
• But the effect of pension reforms on the demand
  of targeted retirement saving products is small

23
Implications

• Improving the dissemination of information about
  pension rights,
• but increasing awareness of pension reforms might
  not be sufficient to prompt households to
  increase private wealth.
• Pension reforms dont seem to have diminished the
  propensity to invest in real estate.
• Annuity markets are still at an infant stage.