"Previous research has found that the presence of a union at a workplace is an important individual-level determinant of union membership. The present article, drawing on a multilevel analysis of 21 European countries, provides further evidence which nuances the conclusions of previous studies by introducing and testing institutional moderation effects. Thus, in countries with Ghent systems, having a union at workplace is less important, since probability of membership is already very high. Conversely, if there are extension mechanisms for collective agreements there is less incentive to join a trade union, and this is not compensated even when there is an active union at the workplace."

Support: Digital