A Life-Cycle Model of Income and Wealth Distribution

Abstract

Across U.S. households, the distribution of wealth is more skewed and fat-tailed than that of earnings. Benchmark models such as Aiyagari (1994), however, imply a less skewed and fat-tailed distribution for wealth than for earnings, because of their assumptions of (i) stationary earnings processes and (ii) infinite horizon. This paper provides a model based on (i) highly-persistent conditionally-heteroskedastic earnings processes, motivated by empirical evidence, and (ii) realistic, stochastic, and finite life horizons. These two features together generate an endogenous wealth distribution that is more skewed and fat-tailed than earnings. Preferences and earnings processes are specified so as to allow a closed-form solution, which facilitates analysis and calibration.

Keywords: Wealth Distribution, Stochastic Life Horizon, Conditional Heteroskedasticity