A spatial panel approach to the East German wage curve

Abstract

"This study presents results about wage effects of unemployment in eastern Germany. The standard approach to analyse the wage curve takes no account of the spatial relationship among regions. Ignoring this relationship may seriously bias the results. To investigate this, the East German wage curve is estimated including spatial effects using panel data classified into 114 administrative districts over the period 1993-1999. It extends the analysis of Baltagi, Blien and Wolf (2000) in which spatial effects were not taken into account. To control for the possible endogeneity of the unemployment rate, one of the explanatory variables, instrumental variable methods should be used. To control for spatial dependence, this paper adopts the use of matrix exponentials for spatially transforming the error terms. This transformation has the advantage that the Jacobian term equals zero. Two important empirical results appear. If spatial effects are not taken into account, the unemployment elasticity in Eastern Germany is larger than in Western countries. If spatial effects are added, the unemployment elasticity significantly reduces." (Author's abstract, IAB-Doku) ((en))

Cite article

Elhorst, J., Blien, U. & Wolf, K. (2002): A spatial panel approach to the East German wage curve. In: European Regional Science Association (Hrsg.) (2002): From industry to advanced services : perspectives of European metropolitan regions, p. 1-22.