Police departments across the United States tend to be more Republican-leaning than the communities they serve, according to anew study published in the American Journal of Political Science. However, despite this political imbalance, the research found little evidence that Republican and Democratic officers behave differently in similar on-the-job situations. While political affiliation appears to have a limited impact on officer behavior, the study reaffirmed previous findings that race and ethnicity are significant factors influencing how officers conduct their duties.
The question of how political beliefs might shape policing has become increasingly important in the United States. Public opinion on law enforcement is sharply divided along partisan lines. Surveys consistently show that Republicans express much higher levels of trust in the police and are less likely to believe that police treat different groups unfairly or that police killings are a significant problem.
In fact, political party identification is a stronger predictor of someone’s views on policing than even their race or general political ideology. This partisan divide in public opinion raises concerns about whether these political differences extend into law enforcement itself. Police officers are granted considerable power and discretion in their daily work, deciding when and how to enforce laws.
This power, coupled with the strong partisan disagreements about policing in society, leads to fundamental questions about fairness and representation in law enforcement. Specifically, researchers wanted to understand the political makeup of police forces, whether this makeup reflects the communities they serve, and if an officer’s political affiliation influences their behavior on the street.
“Policing has become a locus of partisan strife in the United States, with Democrats and Republicans disagreeing strongly on the nature and extent of problems in policing and whether and how policing should be reformed. Given these divisions, my team and I had a number of questions about the partisan affiliations of the people who decide how policing is done every day–officers themselves,” explained study author Jonathan Mummolo, an associate professor at Princeton University.
For their study, the researchers undertook a massive data collection effort. They sought to gather personnel records from the 100 largest local police agencies in the United States. These agencies were identified based on the number of officers primarily engaged in patrol duties, as these are the officers who have the most direct interactions with the public. The team successfully obtained records from 99 of these 100 agencies, representing over a third of all local law enforcement officers nationwide.
Gathering this data was a complex task, involving requests for public records, data-sharing agreements, and searching publicly available information sources. The records included officer names, and in many cases, job titles, which helped distinguish sworn officers from civilian employees within the police departments.
Once they had the officer rosters, Mummolo and his colleagues merged this data with voter registration records. This allowed them to estimate the political party affiliation, voting history, age, and household income of individual officers. To ensure accuracy in matching officers to voter records, they focused on individuals residing in or near the counties where they worked. For information on race and gender of officers at the agency level, they primarily used data from federal surveys of law enforcement agencies.
For a more detailed analysis of officer behavior, the study focused on two major city police departments: Chicago and Houston. These cities were chosen for different reasons. Chicago has made significant efforts to diversify its police force and has been a focus of discussions about police practices. Houston, in contrast, has a more politically balanced police force, with roughly equal numbers of Republican and Democratic officers.
For these two cities, the researchers obtained detailed records of officer shift assignments and enforcement actions, including stops, arrests, and uses of force. This detailed data allowed them to compare the behavior of officers with different political affiliations when they were working in similar places, at similar times, and performing similar tasks. By comparing officers working under essentially the same conditions, the researchers could more reliably assess whether political affiliation was related to differences in their policing actions.
The study’s initial findings confirmed that, on average, police forces are more Republican than the general population in their jurisdictions. Across the 99 agencies studied, about 32% of officers were estimated to be Republican, compared to only 14% of voting-age civilians in those same areas. Officers were also found to be more likely to be white, vote more often, and have higher household incomes than the civilians they serve.
However, this pattern was not uniform across all agencies. Some police departments closely mirrored the political and racial makeup of their communities, while others showed significant differences. In Chicago and Houston, the analysis of officer assignments revealed that even within specific districts or divisions, police officers were consistently more Republican than the residents of those areas.
Despite these clear political imbalances, the analysis of officer behavior in Chicago and Houston yielded surprising results regarding political partisanship. When comparing Republican and Democratic officers working in similar circumstances, the researchers found very few statistically significant differences in their rates of stops, arrests, or use of force. While some minor differences were observed initially, these largely disappeared after accounting for the fact that many comparisons were being made. This suggests that, in general, an officer’s political party affiliation does not strongly predict how they will behave on the job.
“Given how politically polarized the issue of policing has become, it is somewhat surprising that officers’ partisan identities do not translate into detectable differences in behavior,” Mummolo told PsyPost. “However, there are many reasons different officers might converge on common behaviors (for example, they go through similar training and face similar professional incentives).”
In contrast to the limited impact of political affiliation, the study reaffirmed previous research highlighting the influence of race and ethnicity on officer behavior. Consistent with earlier findings, Black and Hispanic officers in Chicago were found to make fewer stops and arrests compared to white officers in similar situations. Black officers also used force less often in both Chicago and Houston.
“Police agencies tilt heavily Republican relative to the civilians in their jurisdictions, on average,” Mummolo explained. “Officers are also more likely to be White, more likely to vote, and have higher household incomes than local civilians. We establish this using a large dataset on 99 of the 100 largest police agencies in the U.S., which covers more than one third of local law enforcement agents in the country.”
“However, when facing similar circumstances, we find very little evidence that officers of different political affiliations behave differently on the job — they make similar numbers of stops and arrests and use force at similar rates. We also replicate our earlier work showing that officer racial and ethnic groups exhibit stark differences in these behaviors.”
Interestingly, when looking at officers of the same race, one notable partisan difference emerged: White Democratic officers in Chicago made more arrests for violent crimes than White Republican officers. However, this was one of the few instances where a partisan difference in behavior was observed, and it was in a direction contrary to what might be expected based on general Republican preferences for more punitive law enforcement.
The researchers acknowledge that their study has some limitations. The behavioral analysis was focused on only two cities, Chicago and Houston, and it is possible that patterns might differ in other locations. Furthermore, while the study controlled for many factors, it cannot definitively rule out all other possible explanations for the observed patterns. Future research could expand this type of analysis to more cities and explore other aspects of policing, such as community perceptions of police and the long-term effects of different policing styles on crime rates and public trust.
“There are roughly 18,000 police agencies in the United States,” Mummolo noted. “While this paper assembles data on the largest agencies, there are many places left to examine. There may be other places where officer partisanship maps closely to behavior on the job. We will need to collect more data to find out.”
Despite these limitations, the study provides the most comprehensive examination to date of political affiliation within law enforcement and its relationship to officer behavior. The findings suggest that while police forces tend to be more Republican than their communities, these political differences do not automatically translate into distinct patterns of officer behavior on the ground. Instead, factors like race and ethnicity appear to play a more significant role in shaping how officers perform their duties.
“We hope our study provides a template for studying police officers and their behavior,” Mummolo explained. “Though difficult to collect, there is a wealth of data being collected inside police agencies that can help us answer longstanding policy questions. In addition, we believe our method for comparing officers facing common circumstances (using detailed records of officer shift assignments and enforcement activities) can be used for other purposes, such as developing improved early warning systems to flag officers who are exhibiting outlying behavior relative to their peers. We have work in progress on this and related questions.”
“The scale of the data collection in this paper is unusual relative to typical studies in the field,” Mummolo added. “That was only possible because of the large team of talented co-authors on this paper.”
The study, “Political diversity in U.S. police agencies,” was authored by Bocar Ba, Haosen Ge, Jacob Kaplan, Dean Knox, Mayya Komisarchik, Gregory Lanzalotto, Rei Mariman, Jonathan Mummolo, Roman Rivera, and Michelle Torres.
