The Effect of Neighborhood Unemployment on Individual Employment Probability in Germany

Thomas K. Bauer1, Michael Fertig2, Matthias Vorell 1
1RWI, Germany, 2ISG, Germany

Using a unique dataset for Germany which links individual longitudinal data from the SOEP to regional data from the federal employment agency for the years 2000 to 2004, we estimate binary discrete choice models to evaluate the impact of neighborhood characteristics, especially neighborhood unemployment, on individual employment decisions. The panel setup and richness of the data allows us, at least partly, to overcome some of the identification problems which are present in this strand of literature (see Manski, 1993). Using an instrumental variable approach to control for selection effects and area-fixed effects to control for correlated effects, we observe a significant negative impact of neighborhood unemployment on individual employment probability. The results are robust to various specification, model changes and the inclusion of individual fixed effects.

View full paper