Community Policing and Public Trust: A Field Experiment in Uganda

Context

Community-oriented policing — designed to improve police-citizen relationships by increasing the frequency and quality of interactions between police and citizens — has been touted as a way to build trust and increase collaboration between officers and citizens. This study evaluates a homegrown model of community policing in Uganda, a country in which police-community relations have long been strained by political bias and excessive use of force against civilians.

Study Design

The model is explicitly designed to create opportunities for more positive, mutually respectful interactions between civilians and police officers. A working group composed of police officers and local NGO representatives decided on the specific components of the program: town hall meetings, door-to-door visits, and the formation of Community Watch Teams (a group of residents who better understand the police, report crimes, and improve police-citizen communication). The researchers evaluated the impact of this program by randomizing a subset of 72 relatively rural and high-crime police stations and posts across all regions of Uganda to adopt this model.

Results and Policy Lessons

Results show that the program had little to no effect on the prevalence of crime, citizens’ perceptions of personal safety, or perceptions of the police force’s intentions, capacity, or responsiveness.

However, community policing did increase the frequency of interactions between civilians and the police. Contrary to researcher expectations, there is also evidence of a modest increase in the incidence of “unofficial payments” made to police officers. Specifically, 9.8% of respondents in treatment villages reported having made an informal payment to police officers in the past six months, compared to 7.3% of respondents in control villages. This suggests that the program either inadvertently exacerbated petty corruption, or (more benignly) demonstrated improvement in citizens’ understanding of what does and does not constitute an unofficial payment under Ugandan law (increasing the reporting, rather than the true incidence of payments). This more benign interpretation would be consistent with related results showing that the program increased Ugandans’ understanding of the criminal justice system, but the COVID-19 pandemic prevented follow-up qualitative work that could further illuminate this finding.

These results underscore the importance of careful monitoring, robust incentives, and sanctions to minimize the risk of misconduct. Otherwise, community policing may only reinforce existing patterns of behavior and do little to improve police-community relations.

Read the working paper here, and their results alongside the results from a coordinated multi-country study here.