Evidence for Multiple Labor Market Segments: An Entropic Analysis of US Earned Income

Abstract

This article revisits the fitting of parametric distributions to earned income data. A new candidate is proposed in line with Camilo Dagum's dictum that candidate distribution should not only be chosen for fit, but that economic content should also play a role. The fit of a simple finite mixture performs as well or better than the widely used generalized beta of the second kind (GB2) and is argued to be easier to interpret economically. Specifically, the good fit is taken as evidence for a finite number of distinct labor market segments with qualitatively different generating mechanisms. It is speculated that this could be reconciled with either modern search-and-match models in which agent and / or firm heterogeneity can lead to multiple equilibria, or with an older theory of labor market segmentation. Regardless, the use of the mixture model addresses one of the central weaknesses of testing the older theory of dual labor markets empirically. The approach taken in this article is also motivated by the work of E. T. Jaynes, the father of maximum entropy approaches to statistical inference, and related to recent work by physicists on the income distribution.