Log in to access the full article.
Abstract This article tests whether places with higher exposure to unionization during the 1940s, due to their pre-existing industrial composition, tended to have larger declines in wage inequality, conditional on local economic and demographic observables and regional trends. We find a strong negative correlation between exposure to unionization and changes in local inequality from 1940-50 and 1940-60. This does not appear to be underpinned by skill-specific sorting of workers or by firms leaving places with high exposure to unionization. We also find that the correlation between exposure to unionization in the 1940s and the change in inequality after 1940 persists in long-difference regressions to the end of the twentieth century.