Public Pension Design and Household Retirement Decisions: Cross - PowerPoint PPT Presentation

Public Pension Design and Household Retirement Decisions: Cross ‐ national Comparisons David Knapp (RAND), Jinkook Lee (USC), Maciej Lis (OECD), Drystan Phillips (OECD) Gateway-OECD Pension team Annual Retirement Research Consortium August 1, 2019

Disclosure • This research was supported by a grant from the U.S. Social Security Administration (SSA) as part of the Retirement and Disability Research Consortium (RDRC). The findings and conclusions are solely those of the authors and do not represent the views of SSA, any agency of the federal government, the Michigan Retirement and Disability Research Center, or the OECD. Slide 2

Study Objective • Analyze responsiveness to retirement incentives from social security benefit structure for the primary earner using newly harmonized cross-country data • Extend the approach to estimate additional retirement incentives from auxiliary benefits and benefit taxation It is a work in progress • Motivation • Approach • Key methodological challenges Slide 3

In the Past 35 Years, There Has Been Limited Policy Variation • Old Age and Survivors Insurance (OASI) benefits in the United States have not substantially changed since 1983 • Examining household work and savings behavior under alternative institutional structures will enable policymakers and researchers to learn from other countries’ pension reforms and designs Slide 4

People Respond to Incentives • Economic theory suggests, and past research validates, that household choices respond to key elements of the benefit structure – Benefit eligibility age – Delayed retirement credits • Gruber and Wise (1999) and coauthors demonstrated this consistently internationally – Existing work has not focused on auxiliary benefits, a significant components of US social security Slide 5

People Respond to Incentives…Sometimes • Carman and Hung (2018) find “that while many are aware of spousal and survivor’s benefits, knowledge about eligibility and benefit amounts is relatively low.” • Learning about benefits may occur near retirement – May be too late to respond • Learning may never happen Slide 6

Auxiliary Benefits Can Provide Incentives and Disincentives to Work • Spouse benefits – At full retirement age (FRA), a married beneficiary is entitled to 50% of spouse’s benefit or 100% of her own benefit – If claiming spouse benefit: • Penalty for claiming before FRA • No benefit of delaying after FRA • In households with a dominant earner: – Incentive for the primary earner to work – Disincentive for the secondary earner to work Slide 7

Auxiliary Benefits Can Provide Incentives and Disincentives to Work • Survivor Benefits – At full retirement age (FRA), a survivor is entitled to 100% of deceased’s benefit or 100% of her own benefit – If claiming survivor benefit: • Penalty if beneficiary or the deceased claim before FRA • Benefit to delayed claiming • In households with a dominant earner: – Incentive for the primary earner to work – Disincentive for the secondary earner to work Slide 8

Auxiliary Benefit Example Sin Single in income Prim imar ary-earner o own e entitlement: $ $2000 Seconda dary-earner own e entitlement : : $0 Spou ouse b benefi fit: $1 $1000 000 Total b benefi fit: $3000 $3000 Graphic by Ilia Karasev, Noun Project Slide 9

Auxiliary Benefit Example Spouse Be Spo Benefits Sin Single in income Dua ual i inc ncome Sam ame o e own n Prim imar ary-earner: $2000 $2000 Prim imar ary-earner: $1 $1000 000 ent entit itlement ent Seconda dary-earner: $ $0 Seconda dary-earner: $1 $1000 000 50% gr great eater er Spou ouse b benefi fit: $1 $1000 000 Spou ouse b benefi fit: $0 $0 Total b benefi fit: $3000 $3000 Total b benefi fit: $2000 $2000 ent entit itlement ent Slide 10

Auxiliary Benefit Example Spo Spouse Be Benefits Only in the single income household does the work decision of the Single in Sin income other spouse influence Prim imar ary-earner: $2000 $2000 the incentive to work via Seconda dary-earner: $ $0 spouse benefits Spou ouse b benefi fit: $1 $1000 000 Total b benefi fit: $3000 $3000 Slide 11

Auxiliary Benefit Example Survi vivor B or Benefits 67% 67% 50% 50% Single in Sin income Dua ual i inc ncome Prim imar ary-earner: $2000 $2000 Prim imar ary-earner: $1 $1000 000 Seconda dary-earner: $ $0 Seconda dary-earner: $1 $1000 000 Total b benefi fit: $2000 $2000 Total b benefi fit: $1 $1000 000 Slide 12

Auxiliary Benefit Example Survi vivor B or Benefits The dual income household, if it is forward looking, would 50% 50% Dua ual i inc ncome be incentivized to work Prim imar ary-earner: $1 $1000 000 longer to insure against Seconda dary-earner: $1 $1000 000 lost income in widowhood Total b benefi fit: $1 $1000 000 Slide 13

The U.S. Spouse Benefit Is Rare • The cost of divorce is assumed by the Social Security system – Divorced spouse and survivor benefits • Auxiliary benefits are more limited in other countries – Fixed spouse benefits if spouse is not otherwise entitled (e.g., Korea) – Split entitlements in the case of divorce (e.g., Germany) Slide 14

Survivor Benefits Vary Internationally • Most countries offer survivor benefits • Design can differ significantly – Dependent on the primary earner’s work history (e.g., Italy), – Fixed amount (e.g., Netherlands) Slide 15

Approach • Own and auxiliary benefit incentive measures can be computed separately • Evaluate whether an additional dollar from auxiliary benefits is the same as an additional dollar in a recipient’s own benefits • Compare responsiveness to auxiliary benefit structures across countries – Goal: understand if certain benefit designs exhibit stronger behavioral responses Slide 16

Challenge: Developing Harmonized Cross-country Public Pension Measures • As part of the Gateway to Global Aging, we are developing harmonized public pension measures for the international HRS family of surveys – Documented pension design and history reflecting the policies in place at the time of the surveys – Policy and institutional detail explorer will be available this next year • Pension incentives are intended to reflect forward looking measures of public pension wealth Slide 17

Harmonized Data Requires Common Questions and Methods • Not all surveys ask the same questions • Pensions depend on lifetime work and earnings • Lifetime earnings must be estimated – Use cross-sectional data to estimate age-earnings trajectories – Without administrative earnings history, level of pension entitlement exhibits substantial measurement error (Knapp et al., 2019) – In U.S., pension incentives for men are relatively robust to measurement error • Driven primarily by incentives to delay claiming Slide 18

Cross-sectional Estimates Lead to Bias Earnings Trajectories for Women (a) Men (b) Women 1.2 1.2 Earnings Relative to Last Observed Admin Data based on 1 1 actual experience 0.8 0.8 (35+ Years of 0.6 0.6 earnings) Projections 0.4 0.4 based on potential 0.2 0.2 experience 0 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Years of Experience Years of Experience Slide 19 Figures from Knapp et al. (2019) in Journal of Pension Economics and Finance

Next Steps • Develop a method for estimating female lifecycle earnings that can be consistently implemented across the global HRS surveys • Compute forward-looking pension incentive measures • Evaluate whether an additional dollar from auxiliary benefits is the same as an additional dollar in a recipient’s own benefits in the U.S. • Evaluate household responsiveness to auxiliary benefit design across countries Slide 20