Baltimore Built a Public Safety Data Operation From Scratch
MONSE was created as a public safety agency outside of law enforcement, overseeing prevention, intervention and re-entry initiatives. But it began without its own data analytics operation.
When Muhlhausen joined the office, Mavronis said, MONSE had no in-house analytics capacity. His team began building systems that could give the agency a “single source of data truth” and allow community organizations working with the city to document their activities. The analytics team now has three people.
Baltimore also publishes a public safety accountability dashboard with information on violent crime and MONSE programs, including how many people have received services and how many community organizations are participating.
Creating those systems required the city to deal with data held by separate agencies and restrictions on sensitive information. Baltimore has used memorandums of understanding to establish reporting requirements and, in one instance, embedded a MONSE employee with the police department when officials could not share an entire data set.
READ MORE: Here is a guide to data governance for state and local agencies.
Analytics Shows Baltimore Where Programs Lose People
Baltimore officials also use the data to look inside programs rather than relying solely on citywide crime statistics.
MONSE, for example, works to connect people at high risk of involvement in gun violence with intensive services. The agency’s data showed a gap between the number of people offered those services and the number who enrolled.
“What happened there? Where are we losing people in that pipeline?” Mavronis said.
Officials are now examining the handoffs between teams and the metrics used at different stages of the program. Mavronis said unreliable numbers could undermine the strategy itself.
Baltimore has also turned to outside researchers to evaluate its programs. Mavronis said a University of Pennsylvania analysis found that implementation of the city’s focused group violence reduction strategy was associated with a statistically significant 25% reduction in homicides and shootings, without an increase in arrests or displacement of violence to other police districts.
The broader decline cannot be attributed to data alone. Baltimore’s strategy includes law enforcement, community violence intervention and services for people at high risk of gun violence. Data determines how many of those efforts are directed and evaluated.
MONSE is now developing more internal dashboards so staff members can see how the information they enter affects decisions.
“When people really start to see how it’s factoring into decision-making, that’s how we can make an impact,” Mavronis said.
The city is also drafting its next five-year violence prevention plan. Mavronis said Baltimore’s patterns of violence have already changed substantially since 2021, requiring the city to revisit the assumptions that shaped its earlier strategy.
