The challenge
Residents in many US cities rarely report crimes or service problems to 911 and 311.
The case
Buffalo, New York's Clean Sweep Initiative brings city departments, community police officers and local organizations to high-need blocks each week. Teams go door to door to ask residents about quality-of-life problems and connect them to services, remove debris and graffiti, board up vacant houses, maintain lots, and carry out community-oriented code and law enforcement. Harvard researchers used city data from 2009 to 2021 to compare 77,955 treated and matched untreated households. Within three months, treated households were 32% more likely to have a drug crime reported (6.2 vs 4.7 per 1,000 households) and 9% more likely to have a blight-related 311 incident reported (43.7 vs 40.1 per 1,000). Within 5.5 months, drug-crime reporting was 42% more likely. Other 911 and 311 reporting showed no difference.
Why the results were mixed
Highly visible visits that build relationships between residents and city institutions may encourage reporting of the problems people can see on their block, such as drug activity and blight, but they did not change reporting overall. Because households were matched rather than randomly assigned, the authors describe this as a relationship with reporting, not proof of cause.
Sources
- Dickens, E., Greifer, N., Robb, K., Marcoux, A., Creighton, J., Allegar, C., & de Jong, J. (2025). The effects of a place-based intervention on resident reporting of crime and service needs: A frontier matching approach. Justice Quarterly.
- Dickens, E., et al. (2023). The effects of a place-based intervention on resident reporting of crime and service needs (working paper version). Harvard Kennedy School, RWP23-033.
Spot an error or know a better source? Tell us