Uncertainty, Persistence, and Heterogeneity: A Panel Data Perspective

The purpose of this paper is to review newly developed identification and estimation tools that are relevant for the analysis of dynamic dependence structures of income risk. I present an application to nonlinear permanent–transitory models of household income using data from the Panel Study of Income Dynamics (PSID), but the empirical approach is more generally applicable. Household income processes are of interest because the size of shocks, the nature of their persistence, and cross-household heterogeneity are all important to understand how income inequality varies with age and cohort and how it translates into consumption inequality. I argue that going from an econometrics of autocovariances to an econometrics of flexible distributions is feasible and has the potential to reveal richer aspects of risk—for example, nonlinear persistence of unusual shocks.