Book chapter · 2024
Community Policing in the Philippines: Communication, Trust, and Service Provision
Dotan Haim, Matthew Nanes, and Nico Ravanilla · Crime, Insecurity, and Community Policing: Experiments on Building Trust (Cambridge University Press)
In brief
The Philippine site of the Metaketa IV community policing study: two phases of community engagement and problem-oriented policing across 298 barangays in Sorsogon changed what citizens noticed about the police but not their trust or their safety, except on the specific local problem each village team chose to tackle.
Research annotationQuestion, variables, design, findings, and mechanism at a glance▾
- Research question
- Do community engagement and problem-oriented policing improve citizen trust and public safety where the police face limited capacity, political subordination, an active insurgency, and a national reputation damaged by a drug war?
- Independent variable (X)
- Assignment to a community policing package delivered in two sequenced phases. A community engagement program ran for two months in early 2017, with officers contacting citizens outside emergencies, explaining how to reach the police, and soliciting feedback on local safety problems. Problem-oriented policing ran from December 2017 through May 2018 through monthly barangay meetings of the barangay captain, the chief tanod, three regular tanods, and the peace-and-order councillor, with a random subset also including two police officers; each team selected one focal safety issue and spent a budget of up to five thousand pesos on a strategy against it.
- Dependent variable (Y)
- The thirteen harmonized Metaketa IV outcome indexes, including compliance, crime victimization, perceived insecurity, perceptions of police, police abuse, crime reporting and tips, perceived police intentions and capacity, cooperation norms, state legitimacy, and community trust, plus three exploratory outcomes defined on each barangay's own focal issue.
- Identification strategy
- Randomized controlled trial in Sorsogon Province with the barangay as both the unit of randomization and the unit of analysis. Of Sorsogon's 541 barangays, 298 were designated safe after excluding those the military flagged as posing a risk of insurgent ambush, and from those 198 were randomly assigned to treatment and 100 to control. Estimation is OLS of each index on the treatment indicator plus controls for cross-randomized treatments, with standard errors clustered by barangay. Indexes are standardized against untreated barangays at midline. Community engagement effects are identified off a midline survey fielded after that phase but before problem-oriented policing, with ten respondents per barangay, and combined effects off an endline survey with fifteen per barangay.
- Main findings
- The treatment was delivered and citizens noticed: compliance is the one harmonized outcome that moves, with personal interaction with an officer running 21.3 percent in treatment against 6.8 percent in control at midline. Everything else is null. None of the other twelve harmonized indexes is distinguishable from zero, and community engagement on its own produces no perceptible difference on any attitudinal outcome. Both treatment and control barangays improved over time on most measures, which the authors read as an outside factor rather than a treatment effect. The exploratory focal-issue analysis is the one place with positive signal: treated respondents were significantly more likely to say police were focusing on the selected issue, and where teams focused on feuds or public intoxication, respondents were significantly less likely to have been personally victimized by that issue.
- Mechanism
- None established. The authors report failing to find evidence of the mechanisms through which they theorize community policing works, namely improved trust, perceived legitimacy and competence, and willingness to provide information. They rule out failed implementation, since delivery was verified and uninvolved residents knew about the program, and rule out spillovers three ways, including a test on crimes that occur inside the perpetrator's own barangay and so cannot be displaced.
- Why it matters
- Supplies the Philippine site of the Metaketa IV coordinated trial and shows the null holds in a setting stacked against it, where police are politically subordinate, cannot patrol most of the province, and compete with both village auxiliaries and an insurgency to provide order. Its constructive point is that community policing may be better assigned modest, locally targeted goals, since the only place effects appear is on the specific problem each team chose to work on.
- Speaks to
- community policing and police reformcoordinated field experiments and Metaketastate legitimacy and public safety provisioninsurgency and competing order providers
Cite
@misc{haim2024policing_ch,
title = {Community Policing in the Philippines: Communication, Trust, and Service Provision},
author = {Dotan Haim and Matthew Nanes and Nico Ravanilla},
howpublished = {Crime, Insecurity, and Community Policing: Experiments on Building Trust (Cambridge University Press)},
year = {2024},
}