Labor Market Polarization and the Great Divergence: Theory and Evidence

Abstract:

Two of the most important features of advanced labor markets in the past quarter century are labor market polarization and the great divergence. The first of these concerns the growth of jobs in high and low wage categories and the disappearance of middle wage jobs. The second is an explicitly spatial theory about the intensification of development particularly at the high end in large, already developed cities relative to smaller, less developed cities. This paper addresses how the two phenomena are interrelated. The great divergence is typically contemplated in a two factor setting with skill-biased technical change. Labor market polarization is instead considered in an explicitly three-factor setting, specifically rejecting the simpler framework. We develop a theory in which the driving forces of labor market polarization alone give rise to both phenomena, building on Autor and Dorn (2013), Davis and Dingel (2014) and Davis and Dingel (forthcoming). Key to this is that the productivity advantages in large cities are biased toward high skilled tasks, so that a uniform shock to technology leads to labor market polarization with a biased impact on cities of different sizes, giving rise to the great divergence. We examine the model using detailed data for a sample of 117 French cities and find the patterns in the period 1994-2015 accord well with the theory.