Why so Unhappy? The Effect of Union Membership on Job Satisfaction

We investigate the effect of union membership on job satisfaction. We account for the endogenous selection induced by the sorting of workers into unionised jobs and use different methodologies to address the question of how the membership decision is related to overall job satisfaction and to satisfaction with pay. We analyse linked employer-employee data from the 1998 British Workplace Employee Relations Survey (WERS) using Propensity Score Matching and Instrumental Variables techniques. Both the linked structure of the data analysed and the estimation strategies employed allow us to deal with the limitations characterising the existing literature. We find that once the endogeneity of membership is accounted for, the unhappiness of union members disappears, suggesting that their reported dissatisfaction stems from individual unobservable attributes rather than from the union status, i.e. unionisation has no causal effect on satisfaction. When satisfaction with pay is considered, the union/non-union satisfaction differential drops considerably, consistently with the existence of wage premia in favour of members.