Peer-reviewed article · 2023
Fire Alarms for Police Patrols: Experimental Evidence on Co-Production of Public Safety
Matthew Nanes, Nico Ravanilla, and Dotan Haim · Journal of Politics 85(3): 1017-1032
In brief
A police hotline experiment in a conflict-affected Philippine province: making crime reporting easier raised reporting by 10 to 19 percentage points, showing how lowering 'search costs' improves public safety.
Abstract
Effective public goods provision requires coproduction by both citizens and the government. Search costs that complicate citizens' ability to share information constitute a critical and understudied impediment to this coproduction. We experimentally evaluate search costs in a rural, conflict-affected province of the Philippines. We randomize the rollout of a police hotline that dramatically reduces the costs of reporting and compare it against both the status quo and an alternative intervention that builds trust but does not affect search costs. The hotline increased the likelihood of reporting crimes by 10-19 percentage points. The intervention reduced perceived insurgent activity but had no perceptible impact on ordinary crime. Our findings suggest that addressing search costs substantially improves service delivery, potentially explaining why policies imported from higher-capacity countries may fail to achieve results in developing contexts.
Research annotationQuestion, variables, design, findings, and mechanism at a glance▾
- Research question
- Do the costs of sharing information, rather than trust alone, prevent citizens from co-producing public safety?
- Independent variable (X)
- Randomized rollout of a police hotline that sharply reduces the cost of reporting crime, benchmarked against both the status quo and an alternative intervention that builds trust without lowering reporting costs.
- Dependent variable (Y)
- Crime reporting, self-reported victimization by armed robbery, burglary, assault, drug-pushing, and armed rebel groups, and citizens' attitudes toward the police including satisfaction, legitimacy, responsiveness, and trust.
- Identification strategy
- Barangay-level randomization across 298 government-controlled barangays in Sorsogon, a rural, conflict-affected province of the Philippines, with 99 assigned to hotline advertisement, 99 to a community engagement arm replicating every procedure except the phone number, and 100 to status-quo policing. The engagement arm isolates search costs from trust and perceived police capacity. Estimates come from OLS on individual responses from an endline survey of ten randomly selected citizens per barangay, run after the four-month intervention, with standard errors clustered by barangay and respondent-level controls.
- Main findings
- The hotline increased the likelihood of reporting crimes by 10 to 19 percentage points, while community engagement alone had no significant effect on reporting. Among public safety outcomes, only victimization by the insurgent New People's Army declined, by roughly 0.8 percentage points, with no detectable change in armed robbery, burglary, assault, or drug-pushing. Attitudes did not improve: satisfaction with the police fell marginally, and satisfaction with police handling of insurgency-related issues fell by 0.186 points and significantly.
- Mechanism
- Search costs, not willingness to cooperate, bind the first stage of co-production, since the hotline moved reporting while the trust-building arm did not. The second stage then broke down through selective responsiveness. Follow-up interviews indicate the police prioritized deterring insurgents, a threat to officers themselves, over ordinary crime, so only insurgent activity fell and citizens grew less satisfied.
- Why it matters
- Offers an explanation for why policing policies imported from high-capacity countries fail in developing contexts, and points to a cheap, scalable fix. It also shows that co-production can fail on the state's side rather than the citizen's, because bureaucrats' priorities diverge from citizens', which means initiatives that solicit citizen effort without delivering results can leave citizen-police relations worse than before.
- Speaks to
- co-production of public goodsstate capacity and service deliveryinformation frictionsbureaucratic responsiveness and selective effortconflict and insurgency
Cite
@article{nanes2023firealarms,
title = {Fire Alarms for Police Patrols: Experimental Evidence on Co-Production of Public Safety},
author = {Matthew Nanes and Nico Ravanilla and Dotan Haim},
journal = {Journal of Politics},
year = {2023},
doi = {10.1086/723971},
}