Measuring and forecasting asymmetries in employment cycles with US labor market applications

Abstract

This paper proposes and estimates a particular non-linear model to measure and forecast asymmetry in employment dynamics and their rates of growth for aggregate US employment and six different categories of workers (white males, white females, non-white males, non-white females, non-farm laborers, and professionals). The model, which is an approximation to a partial adjustment model with asymmetric adjustment costs, fits the data better than the standard linear autoregressive employment models, and accurately reproduces the asymmetries for series with non-linear propagation mechanisms.