True versus spurious state dependence in firm performance

This paper analyzes the persistence of firms’ exporting behavior in a panel of German manufacturing firms using dynamic binary choice models. We distinguish between true and spurious state dependence in exports and apply fixed effects methods that allow us to verify the robustness of our results to critical assumptions on firms’ initial export status. We find robust evidence of state dependence in the current export status of firms. We also document an important role of unobserved permanent firm heterogeneity (spurious state dependence) and quantify the relative importance of different export determinants. Our results, which are consistent with the findings of previous studies on firms in developing countries and in the United States, show the presence of important sunk costs in export market entry and a depreciation of knowledge and experience in export markets. This paper benefitted substantially from three anonymous referees and Bernd Fitzenberger, one of the editors of this journal. The authors wish to thank the Centre for European Economic Research (ZEW), Mannheim, for its hospitality during the time this study was carried out. Both authors gratefully acknowledge helpful comments received at a workshop hosted by the Centre for Applied Microeconometrics, University of Copenhagen workshop, especially those from Bo Honoré. Useful discussions with Georg Licht also lead to improvements of this paper. We thank Ken Chay for sharing Gauss codes used in parts of the paper. Lastly, we would like to thank Bettina Peters, a member of the ZEW team that compiles the data set used in this study, for data guidance and numerous helpful comments. The activities of the Centre for Applied Microeconometrics are financed by a grant from the Danish National Research Foundation.