Peer-reviewed article · 2021
Community Policing Does Not Build Citizen Trust in Police or Reduce Crime in the Global South
Graeme Blair, Jeremy M. Weinstein, Fotini Christia, et al. (incl. Nico Ravanilla) · Science 374(6571): eabd3446
In brief
Six coordinated field experiments across the Global South, including the Philippines, testing whether community policing builds citizen trust in police or reduces crime. It did neither.
Abstract
Is it possible to reduce crime without exacerbating adversarial relationships between police and citizens? Community policing is a celebrated reform with that aim, which is now adopted on six continents. However, the evidence base is limited, studying reform components in isolation in a limited set of countries, and remaining largely silent on citizen-police trust. We designed six field experiments with Global South police agencies to study locally designed models of community policing using coordinated measures of crime and the attitudes and behaviors of citizens and police. In a preregistered meta-analysis, we found that these interventions led to mixed implementation, largely failed to improve citizen-police relations, and did not reduce crime. Societies may need to implement structural changes first for incremental police reforms such as community policing to succeed.
Research annotationQuestion, variables, design, findings, and mechanism at a glance▾
- Research question
- Does community policing build citizen trust in police and reduce crime in the Global South?
- Independent variable (X)
- Locally appropriate increases in community policing practices, focused on more frequent beat patrols, community meetings, crime reporting hotlines, and problem-oriented policing. The exact bundle differs by site, since each police agency designed its own intervention.
- Dependent variable (Y)
- Four harmonized families measured across all sites, resolving into eight preregistered primary indices: crime victimization from surveys and from police blotters, perceived future insecurity, overall perceptions of police, police perceptions of citizens, police abuse, crime reporting, crime tips, and police abuse reporting.
- Identification strategy
- Six coordinated cluster-randomized field experiments run in partnership with police agencies in Brazil, Colombia, Liberia, Pakistan, the Philippines, and Uganda, randomizing beats, neighborhoods, districts, communities, barangays, or police stations to a common community policing arm or to control. Each site is estimated as an intent-to-treat effect by linear regression on a treatment indicator, controlling for baseline outcomes and block fixed effects where assignment was blocked, weighted by inverse probability of sampling and assignment, with robust standard errors clustered at the level of assignment. Brazil is the exception, where a randomized encouragement is estimated by instrumental variables. Site estimates are pooled in a preregistered random-effects meta-analysis across eight indices with correction for multiple comparisons. Interventions covered roughly nine million people, and endline surveys reached 18,382 citizens and 874 police officers. In the Philippines site, in Sorsogon Province, the team randomized barangays and ran the longest intervention of the six at 17 months, sequencing weekly citizen-facing foot patrols first and a problem-oriented policing program second.
- Main findings
- The meta-analysis finds no effect on any of the eight primary outcomes and rules out even small effects in either direction, with confidence intervals excluding crime reductions larger than 0.078 standard deviations and increases larger than 0.047 in the victimization surveys. Administrative crime data show no decrease either. Overall perceptions of police shift 0.051 standard deviations, which does not survive adjustment for multiple testing. No site shows an effect on any primary outcome, though three sites move secondary attitudinal measures: perceived police intentions rise in Liberia and Pakistan and perceived police capacity rises in Colombia.
- Mechanism
- Implementation was uneven and incomplete rather than absent. Police complied with the assigned practices on some measures and differently across sites, and foot patrol frequency did not rise substantially anywhere. Structured qualitative data from all six teams point to three constraints the authors say may have impeded change: police leadership not prioritizing community policing, rotation of both the officers and the leaders championing the reform, and limited resources to follow up on the concerns citizens raised. The authors treat these as suggestive rather than demonstrated, since the nulls hold even in the sites with the highest compliance.
- Why it matters
- Supplies the missing systematic evidence base for a reform now adopted on six continents, through a preregistered joint analysis of six simultaneous trials that guards against publication bias. It concludes that incremental measures such as community policing may have to be preceded or complemented by structural reforms to police agencies to succeed.
- Speaks to
- police reform and crimecoordinated field experiments and Metaketastate-society relationstrust in institutions
Cite
@article{blair2021metaketa,
title = {Community Policing Does Not Build Citizen Trust in Police or Reduce Crime in the Global South},
author = {Graeme Blair and Jeremy M. Weinstein and Fotini Christia and et al. (incl. Nico Ravanilla)},
journal = {Science},
year = {2021},
doi = {10.1126/science.abd3446},
}