Dynamic Panel Data Models with Spatial Correlation

This paper presents an overview of recently developed estimation methods for dynamic panel data models with spatial correlation.We discuss the specification, themain assumptions and the implications of themodel. Themost important estimation strategy is the application ofGeneralized Method of Moments (GMM). The focus lies on the derivation of the moment conditions, the estimation of the degree of spatial correlation and the specification of the optimal weighting matrix. Finally we estimate an augmented matching function to analyze the effects of active labour market policy programs in Germany using two different weighting schemes.