Abstract
In this article, we survey the literature on individual earnings dynamics with a particular focus on allowing for pervasive heterogeneity across individuals. We structure the discussion around ARMA processes with nonlinear trends for each individual. We show that allowing for pervasive and codependent heterogeneity in individual parameters has a major impact on econometric modeling, estimation, and substantive conclusions. We describe an econometric method that is suitable for models with pervasive heterogeneity. We develop a long list of statistics that describe any earnings panel in great detail and provide a demanding set of features of the data for fitting. This list encompasses most moments used in the literature and provides novel statistics based on individual regressions. Finally, we present an empirical illustration using a long Danish panel. Based on this, we provide some conclusions concerning earnings dynamics but emphasize that details will vary according to the sample.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
