Revisiting the labor reallocation hypothesis for the european countries

Abstract

We revisit a major question which has sparked diverse perspectives in the labor economics. Specifically we examine the sectoral shift hypothesis, which was formally identified by the seminal work of Lilien, 1982. The hypothesis states that the reallocation of workers across occupations, following various shocks, that create important wage differentials, is the main factor that leads to increased unemployment. We analyze the validity of the hypothesis for 22 European countries and the time period of 1994:4 - 2023:2 using the purged Lilien’s index. We take advantage of the most recent panel analysis advancements, to control for cross sectional dependence, common factors, dynamic bias and parameter heterogeneity, in an attempt to have the most robust results. Moreover we also examine for the presence of non linearities in the effect, using quantile panel analysis. Our results confirm the validity of the hypothesis using linear models, while the quantile analysis offers a more clear view of the non linear effect of labor reallocation on unemployment.