The paper investigates whether unionisation has a spillover effect on wellbeing by comparing non-members in union and non-union workplaces. To this end, it adapts the social custom model of trade unions and goes on to conduct empirical analyses using linked employer-employee data and alternative empirical strategies. The findings in the paper reveal that unionisation does have a spillover effect lowering non-members' job satisfaction. Sub-group analysis based on workplace-level collective bargaining status uncovers that the adverse effect found is specific to establishments that set pay through collective bargaining.