Abstract

What is the source of labor unions' power? A large literature argues that union membership powerfully affects political behavior, in turn making unions an influential interest group. However, union members may not receive, may ignore, and may deprioritize their union's political messages. Previous research faced empirical challenges with causal inference, limited sample sizes, and limited outcomes. I marshal data from 14 panel surveys from 1956 to 2022 with 49,638 respondents to help better understand the causal effects of union membership. I use difference-in-difference designs to evaluate whether gaining union membership causes voters to liberalize their attitudes and participate in politics. I do not find any evidence that union membership substantively affects political behavior, and I show that we can reject modest effects in many cases. I also show that a prominent previous result is not robust. These results suggest a reconsideration of how labor unions influence politics.