Unemployment invariance hypothesis and structural breaks in Poland

Abstract

Abstract We study the long-run relationship between unemployment and the labour force participation rate while accounting for potential non-linearities. We explore two sources of non-linearities: structural breaks (as we allow the relationship to be time-varying) and differences by gender groups. We apply a novel approach, as we extend the cointegration analysis by combining it with structural break tests. The model is applied to Poland, which is a suitable case for study because it allows us to illustrate how this relationship operates in a developed country that was formerly centrally planned. The results qualify previous findings by providing mixed evidence for the invariance, the discouraged worker, and the added worker hypotheses. In particular, in the past, a positive and cyclical relationship – i.e., an added worker effect – could be observed, albeit with differences by gender and age group. For the current regime, we can provide evidence of a discouraged worker effect among Polish men only.