Cognitive Ability, Union Membership, and Voter Turnout

Abstract

Labor unions are said to in uence elections and public policy by increasing their members’ electoral turnout. But existing research likely overestimates the turnout effect of union membership by ignoring sorting in the labor market. In the presence of a union wage premium, both membership and turnout are shaped by the same (unobserved) factors, such as cognitive ability. To disentangle the union effect from positive selection, we use unique data from the U.S. National Longitudinal Survey of Youth. It allows us to specify a latent factor potential outcome model with matching on both observable and unobservable individual characteristics. We find that about one-third of the observed union turnout effect is due to selection, more than what previous studies suggest.