"Social partners (trade unions and employers' associations) and their representativeness can shape labour institutions and economic and social outcomes in many countries. In this paper, we argue that, when examining social partners' representativeness, it is important to consider both affiliation rates and dissimilarity measures. The latter concerns the extent to which affiliated and non-affiliated firms or workers are distributed similarly across relevant dimensions, including firm size. In our analysis of the European Company Survey, we find that affiliation density and dissimilarity measures correlate positively across countries, particularly in the case of employers' associations in which we focus. This result also holds across employers' associations when we use more detailed, firm population data for Portugal. We conclude that higher affiliation densities do not necessarily correspond to more representative social partners as they can involve greater dissimilarity between affiliated and non-affiliated firms."

Support: Digital