This document discusses factors that shape wealth inequality in the U.S., specifically examining the roles of longevity and income inequality. It finds that the colossal rise in life expectancy has significantly contributed to rising wealth inequality over time by increasing lifetime wealth and the percentage of individuals at their peak wealth levels. Even without rising income inequality, wealth inequality would still increase due to greater longevity. The paper uses an overlapping generations model to quantify the effects of increased life expectancy on wealth inequality and compares it to the impact of income inequality.

1. What shapes the U.S. wealth distribution?
Longevity vs income inequality
Oliwia Komada Krzysztof Makarski Joanna Tyrowicz Piotr ›och
FAMEjGRAPE FAMEjGRAPE FAMEjGRAPE, IZA FAMEjGRAPE
Warsaw School of Economics University of Regensburg University of Warsaw
Why? Demography matters!
Collosal rise in life expectancy (longevity)
ˆ " lifetime wealth [permanent]
ˆ " % of individuals at peak wealth [transitory]
Wealth ineq. to " even w/o " in income ineq.
This paper
In an OLG model quantify:
ˆ the role of rise in LE for wealth inequality;
ˆ pitch demography against income inequality.
+ Policy experiments.
In the initial steady state
Wealth inequality driven by income risk and life-cycle
savings. Negiligible role of discount factor shocks
and return risk.
Contributions to wealth inequality (initial steady state)
Contribution Gini Top 10% share
Life-cycle 67.21% 59.79%
Idiosyncratic productivity 17.70% 74.01%
Idiosyncratic discount rates 6.08% 2.82%
Idiosyncratic returns 0.00% 0.03%
Model with multiple mechanisms of redistribution
Government collects taxes and issues debt to nance government purchases, operates PAYG DB social security.
Redistribution via:
progressive labor income tax (as in Benabou, 2002)
progressive social security (AIME).
taxes on consumption and capital income are at
govn't purchases do not enter utility
Model with multiple sources of uncertainty
Individuals risk averse, choose consumption and leisure, retire at age 65 and receive pension.
They pay Social Security contributions, labor income, capital income, and consumption taxes.
Uncertainty at all stages of life:
lifetimes with stochastic survival.
earnings due to idiosyncratic productivty shocks.
capital incomes due to idiosyncratic discount rates.
capital incomes due to idiosyncratic returns.
Production: standard Cobb-Douglas function with capital and labor.
Scenarios (calibrated to the US)
Calibration to match USA economy in 1960.
Initial steady state in 1935. Transition with perfect
foresight.
Variance of productivity shocks rises for subsequent
birth cohorts.
Full model features changes in
ˆ Longevity: historical mortality data + UN pro-
jection until 2100.
ˆ Fertility: historical births data + US Census
projection until 2060.
ˆ Technology: TFP growth and labor share.
ˆ Fiscal policy: tax rates, progressivity of labor
income tax, gov't. purchases, debt/GDP.
S1: No growth in life expectancy
ˆ Mortality risk xed at its 1960 level.
ˆ Counterfactual (implied) population structure.
S2: Not seeing growth in life expectancy
ˆ Individuals perceive mortality rate as in 1960.
ˆ Demographic structure as in the data.
S3: No change in income inequality
ˆ Keep std of shocks xed at its 1960.
ˆ Keep persistency of shocks xed at its 1960.
Longevity and wealth inequality Conclusions: change needed?
Rise in longevity is a big part of the rise in wealth
inequality.
Rise in income inequality is equally relevant.
Relatively minor role of changes in tax system (not
shown here).
These forces will continue to operate.
Work in progress: which policy can address
demography-driven wealth inequality?
Acknowledgements
This project was gracefully supported by Natio-
nal Science Centre (grant #2016/22/E/HS4/00129).
The authors received valuable insights from too many
people to mention here (a full list will appear in the
paper). All the errors are with the authors.