This piece is cross posted from the Data-Smart City Solutions blog hosted by the Ash Center for Democratic Governance and Innovation at Harvard Kennedy School.
In April, the National League of Cities’ Big Ideas for Cities conference brought together some of the country’s most innovative mayors to address the problems that keep them up at night, and violence topped many of their lists. Mayor Michael Nutter of Philadelphia identified the city’s murder rate, and the outsized effect it has on its African-American male population, as a public health crisis.
The question is: What data and tech tools might help the city address this crisis?
In his speech, Nutter laid out the basic data available: In 2013, 77 percent of Philadelphia’s homicide victims were black. Of these 191 victims, 179 were male.
“Under any other set of circumstances — under any other measurement — this would be considered a public health crisis,” said Nutter. “The No. 1 cause of death for black men and boys between 10 and 24 is homicide.” Nutter recalled the advancement of safety technology in cars — from seatbelts to airbags to automatic braking for a distracted driver — and asked why there hasn’t been a similar comprehensive technological response to the gun violence epidemic.
There is not yet a far-reaching, data-driven tech solution that addresses all facets of the crisis identified by Nutter, but there are promising public safety initiatives that are making many city streets safer for all.
For two decades, police departments have led the way in the use of data to drive performance. Beginning with New York City’s CompStat, police departments across the country have been using statistical analysis to identify and respond to crime hot spots and criminal patterns.
More recently, technological advances involving sensors give police better real-time situational awareness. For example, ShotSpotter provides instant alerts to law enforcement when gunfire is detected.
But perhaps the true tech breakthroughs will be those driven by a combination of software and hardware. Predictive analytics generally offer better approaches to complex, systemic problems.
In recent years, the Santa Cruz, Calif., Police Department has deployed a predictive policing program that uses a mathematical formula to assign risks of future crimes in areas as small as 150 square meters. The program initially focused on property crimes, and it has been hugely successful: In its first six months, it reduced burglaries by 14 percent. As the department gradually expands the program to analyze gang activity and street crime, it may prove to be a model for deterring violent crime in the streets of any city.
When combined with community policing, predictive analytics will greatly augment public safety. This can be seen in Chicago, where network analysis has helped the police move beyond hot spots to identify “hot people,” or individuals who are likeliest to be involved in future violence. Officers can then reach out to the most dangerous and vulnerable individuals to deter violence.
Philadelphia’s own predictive probation program may show a way forward. A forecasting model based on machine learning assigns each of the city’s probationers a risk of committing a violent crime. Based on this forecasting, the agency can better assign staff and other resources.
While the predictive models in Santa Cruz and Chicago provide the departments with the resources that help strategists identify places and people of interest, the next wave of enhancements will push those decision support tools to the field.
In March, Commissioner Bill Bratton announced a pilot program equipping New York City Police Department officers with tablets connected to its Domain Awareness System. This system keeps officers connected in real time to arrest records, transcripts of 911 calls, gun permits and other data relevant to their location and the task at hand. This development turns officers in the field into sensors and headquarters into a nimbler central nervous system.
Our country’s epidemic of violence will be difficult to solve, but applying predictive analytics to both public safety and social service initiatives provides a start.
Stephen Goldsmith is the Daniel Paul Professor of the Practice of Government and the Director of the Innovations in Government Program at the Harvard Kennedy School. His post originally appeared on Gov Tech.