Union membership density and wages: The role of worker, firm, and job-title heterogeneity

We examine the association between union density and wages in Portugal where just 10 percent of all workers are union members but nine-tenths of them are covered by collective agreements. Using a unique dataset on workers, firms, and collective bargaining agreements, we examine the union density wage gap in total monthly wages and its sources – namely, worker, firm, and job-title or ‘occupational’ heterogeneity – using the Gelbach decomposition. The most important source of the mark-up associated with union density is the firm fixed effect, reflecting the differing wage policies of more and less unionized workplaces, which explains two-thirds of the wage gap. Next in importance is the job-title fixed effect, capturing occupational heterogeneity across industries. It makes up one-third of the gap, the inference being that the unobserved skills of workers contribute at most only trivially to the union density wage gap. In a separate analysis based on disaggregations of the total wage, it is also found that employers can in part offset the impact of the bargaining power of unions on wages through firm-specific wage arrangements in the form of the wage cushion. Finally, union density is shown to be associated with a modest reduction in wage inequality as the union density wage gap is highest among low-wage workers. This result is driven by the job-title fixed effect, low-wage workers benefiting more from being placed in higher paying ‘occupations.’