Politics in Tax Office: Complicated Effects of Contributing to a Politician on Tax Avoidance Using Regression Discontinuity Design

Abstract

Using a regression discontinuity design in a sample of close U.S. elections, this paper investigates how firms’ tax avoidance is affected by four types of political contribution. With hand-collected datasets, we compare firms donating to marginally winning candidates and firms donating to marginally losing candidates in special and general elections of federal congress, gubernatorial elections, and general elections of state congresses. We find that: 1) only large firms gain tax benefits from donating to federal congressmen; 2) donating to state governor candidates only helps large firms avoid tax, while 3) donation to state congressional election winners of the losing party may even hurt. Our results reveal the complicatedness of U.S. political contribution’s effects on firm tax avoidance.