Works councils, labor productivity and plant heterogeneity: evidence from quantile regressions

Using quantile regressions and a rich cross section data set for German manufacturing plants, this paper reports that the impact of works councils on labor productivity varies along the conditional distribution of value added per employee. It emerges that the positive and statistically significant effect of works council presence estimated by ordinary least squares now vanishes for large parts of this distribution. According to our results, such an effect can only be detected in plants at the top end of the conditional productivity distribution – plants that can be considered ‘over achievers.’ We would speculate that this might be because only highly competent managers can cooperate with a works council in a way that much enhances productivity.