Worker satisfaction and worker representation : the jury is still out

This paper investigates the relationship between worker job satisfaction and workplace representation, to include works councils as well as local union agencies. The paper marks a clear shift away from the traditional focus on union membership per se because its sample of EU nations have industrial relations systems that diverge markedly from those of Anglophone countries. Our dataset comprises two waves of the European Working Conditions Survey (EWCS). Pooled cross-section data indicate that workers in establishments with workplace representation have less job satisfaction than their counterparts in plants without formal representation. We proceed to upgrade these findings of conditional correlation by constructing a pseudo-panel with cohort fixed effects to take account of unobserved worker heterogeneity. Causality issues are directly tackled using an endogenous treatment effects model to address the possible endogeneity of worker representation. A persistence of our central finding leads us to conclude that, despite the recent evidence of a turnaround in the association between job satisfaction and unionism, it would be premature to conclude that this result can be generalized to continental European nations.