Works Councils and the Productivity Impact of Direct Employee Participation

This paper measures the productivity impact of management-led participative establishment practices. On the basis of a representative German establishment data set, the IAB establishment panel, the study finds that the presence of team-work, a reduction of hierarchies and autonomous work groups in 1997 significantly increases average establishment productivity in 1997 – 2000. An endogeneous switching regression model takes the endogeneity of work councils into account and shows that the productivity effect can only be measured in establishments with works councils, i.e. employee induced participation. The estimation strategy controls for unobserved time invariant establishment heterogeneity by using a two-step system GMM panel regression approach. It simultaneously controls for endogeneity of participative work organization by using instrument variable regressions. Key-words: employee participation, works council, establishment productivity, panel regression JEL-codes: C23, D24, J50, M12 * I want to thank the Institut für Arbeitsmarktund Berufsforschung, Nürnberg and especially Holger Alda for providing me with the possibility to analyse the IAB establishment data. In addition, I am also grateful to Bernhard Boockmann, Tobias Hagen, Thomas Hempell, Friedhelm Pfeiffer, Petri Rouvinen, Michela Vecchi, Guy Vernon, and Elke Wolf for helpful comments.