how does raising women’s retirement age affect labor supply, income, and mortality?
DESCRIPTION
How Does Raising Women’s Retirement Age Affect Labor Supply, Income, and Mortality? Evidence from Switzerland. Rafael Lalive University of Lausanne and CEPR. Stefan Staubli University of Calgary and RAND. Retirement Research Consortium Annual Conference Washington DC, August 2014 - PowerPoint PPT PresentationTRANSCRIPT
How Does Raising Women’s Retirement Age Affect Labor Supply, Income, and Mortality?
Evidence from Switzerland
Rafael LaliveUniversity of
Lausanneand CEPR
Stefan StaubliUniversity of Calgary
and RAND
Retirement Research Consortium Annual Conference
Washington DC, August 2014
NBER (SSA) funding gratefully acknowledged (NB14-09)Lalive and Staubli
Raising Women’s Retirement Age
1 / 20
Research Question
• How does raising women's full retirement age (FRA) affect labor supply, income, and mortality?
• Demographic aging puts financial pressure on social security programs
• Raising the retirement age is the key policy reform to deal with financial pressure• People claim for shorter duration• People stay in labor force longer• Potential concerns: negative effect on health and spillover
effects onto other government programs (e.g. disability insurance)
• Little empirical evidence on impact of large changes to FRA
Raising Women’s Retirement Age
Lalive and Staubli
2 / 20
What We Do
• Swiss FRA before 2001: 62 for women and 65 for men
• We study the effects of raising women’s FRA
• 1997 Law announced increase in women’s FRA via penalty of 3.4% per year:• FRA is 63 if born after Dec. 31, 1938 (i.e. cohorts retiring in
2001)• FRA is 64 if born after Dec. 31, 1941 (i.e. cohorts retiring in
2005)
• Institutional comparison to U.S.• Labor force participation rates of women are very similar• No earnings test in Switzerland: labor supply effects might
be smaller in U.S. context
Raising Women’s Retirement Age
Lalive and Staubli
3 / 20
Generosity of Old-Age Pensions90
100
110
120
130
Inde
xed
bene
fits
(100
=age
62,
bor
n <1
938)
62 63 64 65 66 67 68 6970
AgeBorn <1939 Born 1939−1941Born >1941
Lalive and Staubli
Raising Women’s Retirement Age
4 / 20
Data and Outcome Variables
• Data Sources:• Administrative data on pension contributions and claims• Labor market histories from 1982-2012• 25% random sample of individuals and their spouse
• Sample: women with labor force attachment affected by reforms (born 1939/1942) and not affected by reforms (born in 1938/1941)
• Key variables• Exit age: last age an individual has positive earnings• Claiming age: age individual starts claiming disability/old-age
pension• Mortality: Indicator(died by end of 2012)• Income: amount of benefits and social security wealth (net-
present value of benefits)
Raising Women’s Retirement Age
Lalive and Staubli
5 / 20
Birth Date Regression Discontinuity Design (RDD)
• Estimate effects by exploiting discontinuity in FRA by birth date
• Treatment groups: women born in 1939/1942
• Control groups: women born in 1938/1941
• Perform analysis by month of birth
Raising Women’s Retirement Age
Lalive and Staubli
6 / 20
2001 Change
• FRA 62 ⇒ 63
• Penalty 3.4%
• Informed 4 years in advance
Raising Women’s Retirement Age
Lalive and Staubli
7 / 20
FRA 62⇒63: Women’s Exit Ages59
.560
60.5
6162
62.5
Age
61.5
63
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff
Lalive and Staubli
Raising Women’s Retirement Age
8 / 20
FRA 62⇒63: Women’s Exit Ages59
.560
60.5
61A
ge61
.562
62.5
63
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff
Lalive and Staubli
Raising Women’s Retirement Age
8 / 20
FRA 62⇒63: Women’s Claiming Ages60
.561
61.5
62A
ge62
.563
63.5
64
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff
Lalive and Staubli
Raising Women’s Retirement Age
9 / 20
FRA 62⇒63: Women’s Claiming Ages60
.561
61.5
62A
ge62
.563
63.5
64
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff
Lalive and Staubli
Raising Women’s Retirement Age
9 / 20
FRA 62⇒63: Effects on Women’s Labor Supply Profile
0.2
.4.6
.81
Per
cent
50 55 60 65 70 75Age
1938
1939Lalive and Staubli
Raising Women’s Retirement Age
10 / 20
FRA 62⇒63: RDD Effects on Unemployment Inflow
−4−2
02
4
53 54 55 56 57 58 59 60Age
61 62 63 64 65
Lalive and Staubli
Raising Women’s Retirement Age
11 / 20
FRA 62⇒63: RDD Effects on Disability Inflow−2
02
4
53 54 55 56 57 58 59 60Age
61 62 63 64 65
Lalive and Staubli
Raising Women’s Retirement Age
12 / 20
FRA 62⇒63: Spillover Effect on Exit Age of Spouse?
6262
.563
Age
63
.564
64.5
6565
.5
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff, wife
Lalive and Staubli
Raising Women’s Retirement Age
13 / 20
FRA 62⇒63: Impact on Mortality?0
.05
Pr(
die
by 2
012)
.1.1
5.2
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff
Lalive and Staubli
Raising Women’s Retirement Age
14 / 20
FRA 62⇒63: Change in BenefitsS
wis
s fra
ncs
1600
0
1700
0
1800
0
1900
0
2000
0
2100
0
2200
0
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff
Lalive and Staubli
Raising Women’s Retirement Age
15 / 20
FRA 62⇒63: Change in Social Security WealthS
wis
s fra
ncs
4000
00
420
000
44
0000
4
6000
0
4800
00
500
000
−12 −9 −6 −3 0 3 6 9 12Birth date − cutoff
Lalive and Staubli
Raising Women’s Retirement Age
16 / 20
2005 Change
• FRA 63 ⇒ 64
• Penalty 3.4%
• Informed 8 years in advance
Raising Women’s Retirement Age
Lalive and Staubli
17 / 20
FRA 63⇒64: Effects on Women’s Labor Supply Profile
0.2
.4.6
.81
Per
cent
50 55 60 65 70 75Age
1941
1942Lalive and Staubli
Raising Women’s Retirement Age
18 / 20
Summary of RDD Estimates
FRA 62 to 63 years
FRA 63 to 64 years
A. Labor supply and claiming
effects
Exit age (years)
0.66***
(0.19)0.53**
* (0.08)
0.58***
(0.17)0.55**
* (0.09)
Claiming age (years)B. Health and income effects
Mortality, indicator(died by end of 2012)
0.004(0.011)
-329(219)
-14,489***
(4,939)
0.016*(0.008)
-278(204)
-13,296**
*(4,439)
Social security benefits
Social security wealth
# Observation 11,329 13,912
Lalive and Staubli
Raising Women’s Retirement Age
19 / 20
Summary
• Exploit reform in Switzerland that increased FRA for women⇒Creates sharp discontinuities in financial incentives to claim
benefits
• Main findings:
1. Significant delay in benefit claiming and labor market exit2. No impact on benefits but reduction in social security
wealth3. Some increase in mortality but not very precisely estimated⇒Data update for full population and more recent yearsRaising Women’s Retirement
Age20 /
20Lalive and Staubli