This article asks how police union campaign spending shapes local politics and public safety outcomes by assembling a large new dataset on U.S. mayoral campaign finance and linking endorsements to later departmental outcomes. The authors contribute a broad descriptive account of police-union donations across cities and use a within-city research design that accounts for staggered timing to estimate the association between electing police-union-backed mayors and subsequent police resources and performance.
Large multi-city donation dataset
Within-city staggered-treatment design
Estimated increase in police department size after union-backed wins
Data and methods
The authors build a novel dataset covering campaign finance in U.S. mayoral elections: 2,303 races in 1,073 cities from 2001–2019, recording police-union donations and which candidates they backed. To estimate effects they use a within-city design that accounts for heterogeneous treatment effects with staggered timing of union-backed mayoral victories; the abstract also notes attention to treatment heterogeneity when estimating effects.
Main findings
Descriptively, police unions frequently donate to mayoral candidates—especially in mid-sized and large cities—and the donations tend to target likely winners and incumbents without showing clear partisan bias. Using their within-city design, the authors report that electing a police-union-backed mayor is associated with an expansion in police department size. The analysis does not find evidence of improved police performance following those elections; the authors present this as a qualified finding about associations rather than causal certainty.
- Police unions frequently donate to mayoral candidates, concentrating on likely winners and incumbents.
- Election of a police-union-backed mayor is linked to larger police departments but not to measurable performance improvements.
