New York City Police Commissioner Bill Bratton, left, stands with Mayor Bill de Blasio at a news conference on August 2, 2016 in New York City. (Photo by Spencer Platt/Getty Images)

New York City Police Commissioner Bill Bratton announced on Tuesday that he will resign next month amid demonstrations by frustrated activists and acrimony between the department’s rank and file and the city’s civilian leaders, apparently bringing an end to his divisive and influential career in law enforcement.

As the top cop in America’s most populous city, Bratton implemented methods that some criminologists credit with helping make New York City much safer. Crime has declined nationwide since Bratton began his first stint as commissioner in 1994, but the decrease has been even more dramatic in New York than in other major cities.

Bratton left his post in New York in 1996, before returning to the helm again in January 2014. He will be succeeded by James O’Neill, who is currently the department’s chief.

Bratton “has left an indelible mark on American policing, and he has worked diligently to try to advance the field,” said James Bueermann, the president of the Police Foundation in Washington, D.C.

At the same time, Bratton relied on strategies that other experts say did little to reduce crime, and have potentially damaged public safety by eroding confidence and trust in the police. Demonstrations against racial bias have partly defined his tenure as the head of the department, which began when he returned to New York after serving as police chief in Los Angeles.

Computers and manpower

New York’s most dangerous year was 1990, when there were 30.7 homicides for every 100,000 people in the city, more than average for large U.S. cities at the time. Now, though, New York is substantially safer than the typical big city. There were 4.1 homicides per 100,000 people there last year, a decline of 87 percent over a quarter century.

It is likely that much of that decline was the result of factors beyond the control of the police, including Bratton. Indeed, violence began to decrease before Bratton first took office.

Yet Franklin Zimring, a criminologist at the University of California at Berkeley, has argued that Bratton and his lieutenants contributed to making New York City’s streets safer — and that what Zimring calls the “New York model” helped reduce crime in other cities, too.

Zimring compared New York to other major U.S. cities, and concluded that about half of the decline in crime was probably because of changes in policing, if not to Bratton’s leadership specifically.

For example, the number of officers in New York increased by about 41 percent between 1990 and 1999, Zimring said, which helped reduce crime.

For his part, Bratton is widely credited with requiring his officers to collect data on crimes and use it to make strategic decisions about how and where to police. Other police departments emulated the New York program, called “Compstat” for computerized or comparative statistics. One recent analysis concluded the new systems reduced homicide nationally by about 11 percent.

Compstat has its critics.

They argue that the system forces officers to act more aggressively — writing up stops without reason and issuing summonses or making arrests instead of just giving warnings — so that their superiors can provide data showing their units are working hard. Researchers and former officers say such patterns developed in New York mostly after 2002, under Raymond Kelly’s leadership as commissioner of the department.

Broken windows

Proponents acknowledge the data alone is not enough to curb crime, and that police also need to know how to use the information they gather with systems like Compstat. A strategy that encourages officers to act disrespectfully can be counterproductive.There are polite and impolite ways of interacting with citizens,” Zimring said.

Bratton’s name is associated with the “broken-windows theory” of law enforcement, which instructs police to focus on minor offenses to maintain order in public spaces and deter more serious criminals. Bratton promoted the broken-windows approach vocally during his early years as commissioner.

For Bratton’s detractors, broken-windows theory is an excuse for police officers to harass people on the street, especially African Americans and Latinos.

The police officers who accosted Eric Garner on Staten Island in 2014 were trying to control sales of loose cigarettes, which is exactly the kind of minor crime that opponents of broken-windows policing say the theory wrongly addresses in the name of public order. Garner died in that altercation.

The evidence in support of the broken-windows theory is mixed. A few studies suggest that by focusing on maintaining public order, police can reduce homicide and violent crime, but the relationship is tenuous. Other factors in a neighborhood might have more influence on the level of crime, such as whether residents trust the police and are willing to cooperate with them.

John Eck, a criminologist at the University of Cincinnati, praised Bratton as “probably one of the smartest people ever to be in policing,” but he argued that Bratton’s emphasis on broken-windows policing was a missed opportunity. Bratton should have focused on other, less controversial theories, according to Eck.

Eck argues for a different approach to preventing crime, in which police focus on the kinds of problems that cause crime — whether it is a club where fights break out regularly or an escalating feud between rival gangs — and find ways of addressing them. Several studies have found that departments that take this approach reduce crime.

“I think the New York City police could have gone a lot further,” he said.