The heralded Compstat program to make police commanders more accountable produced “some significant but modest changes” in three police departments studied by criminologists.
Compstat succeeded in monitoring and evaluating police middle managers more rigorously, but at the price of stifling creativity, said James Willis of the University of Massachusetts-Boston and Stephen Mastrofski of George Mason University at an event sponsored by the Heritage Foundation in Washington, D.C. They were reporting on federally funded research in Minneapolis, Newark, and Lowell, Ma.
Under the Compstat model developed by then-New York Police Commissioner William Bratton in the 1990s, district police commanders are quizzed by superiors in high-pressure meetings about what actions they plan to take about crime trends in their areas. The program did help improve the “primitive” state of crime data collection, Willis and Mastrofski found.
The organizational structure in the three departments did not change under Compstat, the researchers said. Pressure on commanders for a “decisive, rapid response” to crime problems meant that officers were not trained or encouraged to devise innovations in problem-solving tactics, they concluded.
Overall, “you can’t get a sea change [in a police department] just plugging in a program like this,” Mastrofski said.
The study was not designed to determine whether Compstat helped reduce crime. The evidence is inconclusive. Crime already was headed downward in the three cities studied before Compstat was instituted in the late 1990s. The rate of reported serious crimes continued declining in Minneapolis and Newark but rose in Lowell after Compstat had been in place for several years.
Willis and Mastrofski did their research for the Police Foundation, based in Washington. It included visits to 16 police departments and intensive research in the three cities. Lowell has 260 sworn officers for its 105,000 residents as of 2001. Minneapolis has 919 officers for nearly 400,000 residents. Newark has a much
higher officer-citizen ratio, with 1,445 officers for about 276,000 people.