Political Segregation at Work

• In Event: Social Networks and Political Sorting

Abstract

Existing research has documented the effects of partisan segregation for contact between groups, discriminatory behavior and local political participation in the US. However, previous work has almost exclusively focused on residential segregation. We argue that workplaces may hold independent explanatory power in the study of sorting based on political preferences and identity. In this paper, we investigate the effect that worker-firm ideological misalignment has on workers' decisions to exit ideologically incompatible jobs. We use a novel source of survey data collected from service sector workers employed at 166 unique firms. These data measure worker partisanship and intention to leave their jobs at baseline in 2020 as well as contain panel measures of actual job exit gauged X – Y months after baseline. Uniquely, these worker-level micro-data are linked at the firm-level, allowing us to construct firm-level measures of ideology, which we do by recording the proportion of campaign donations to Democratic/Republican U.S. House, Senate and presidential candidates and PACs made by corporate managers and executive board members by firm. Using the timing of corporate spending activity releases as mandated by FEC filing deadlines as an instrumental variable, we are able to estimate whether worker exit is caused by firm-level ideological mismatch.

Authors

• Kiara Hernandez, Harvard University

• Daniel Schneider