Unit Root Tests, Persistence, and the Unemployment Rate of the U.S. States

Abstract

In this paper, we investigate the existence of infrequent shocks and the degree of persistence of U.S. state unemployment over the period 1976‐2004. We first apply individual Lagrange multiplier (LM) unit root tests and fail to reject the hysteresis hypothesis in forty states. When two changes in level are incorporated, we again fail to reject the hysteresis hypothesis in forty states. Since individual unit root tests normally lack power, we employ the recently developed panel LM unit root tests with up to two changes in level. Only in this case are we able to reject the joint unit root hypothesis in favor of regime stationarity. Computation of half‐lives through impulse‐response functions indicates the high degree of persistence of U.S. state unemployment. These results contrast with the common belief among scholars that U.S. state unemployment is closer to the natural rate paradigm than to the hysteresis paradigm.