We estimate dynamic effects of works councils on labour productivity using newly available information from West German establishment panel data. Conditioning on plant fixed effects and control variables, we find negative productivity effects during the first five years after council introduction but a steady and substantial increase in the councils’ productivity effect thereafter. Our findings support a causal interpretation for the positive correlation between council existence and plant productivity that has been frequently reported in previous studies.

Unions are an important indicator of various measures of firm performance inAnglo-Saxon countries. The same is true for the German analogue of the workplace union, the works council. Using German establishment data, I examine the impact of works councils on further training. I employ pooled logit and count-data models to analyse the further training activity and intensity of German firms. Because the treatment variable may suffer from endogeneity, I also adopt linear and nonlinear instrumental variable techniques. The analysis reveals a positive impact of works councils on firm-provided training and provides slightly weaker evidence for firm-size differentials in workplace representation. I conclude that enhanced management-employee relations foster the training efforts of firms.