Is justice blind? Evidence from federal corruption convictions

Abstract

Are federal prosecutors influenced by partisan political concerns? We examine that question by analyzing 40 years of federal corruption convictions at the state and federal district levels. Our key finding is that state-level federal corruption convictions fall by roughly 9% in years when a state’s governor belongs to the same party as the president who appointed local US Attorneys, a measure of state-federal political alignment. The result is robust to controls for the state political environment, election cycles, party tenure in the executive branch, public sector employment, federal aid to states, a state’s electoral importance, and the changes in Honest Services law, the statutory basis for many federal corruption cases. Our results are consistent with a significant level of partisan prosecutorial bias on the part of US Attorneys. In a placebo test, we find no evidence that state-federal political alignment affects the total number of federal criminal convictions. That finding provides support for the mechanism that we propose, namely the partisan exercise of prosecutorial discretion, rather than the partisan allocation of prosecutorial resources across federal districts.