Hysteresis in Unemployment: Evidence from Norwegian Counties

Abstract

The paper tests hysteresis effects in unemployment using aggregate and panel data for Norway. While tests using aggregate or county—specific time series do not reject the null of unit root hysteresis, the panel tests firmly reject the null. When a one—time structural break is incorporated, the unit root hypothesis is rejected (or nearly so) in most counties. All results reveal a high degree of unemployment persistence, but the speed of adjustment is much higher when we allow for a change in mean.