Rafael A. Mangual is a fellow and deputy director of legal policy at the Manhattan Institute for Policy Research and a contributing editor to City Journal.

Now that Mike Bloomberg has announced he’s running for president, he may want to rethink his campaign messaging. Because if he doubles down on his recent comments regarding the New York Police Department’s stop-and-frisk practices, he may end up undercutting what will likely be the signature issue of his campaign: gun control.

As a founder and financial backer of one of the largest gun-control advocacy groups in the country, Bloomberg has been an ardent supporter of heavier restrictions on firearm purchases, ownership and carriage. The former New York mayor has touted both restrictions on private gun transfers and ownership of assault weapons. His group, Everytown for Gun Safety, has spent millions opposing an expansion of “concealed” carry across the nation.

But America’s gun violence is heavily concentrated in urban neighborhoods, where it is driven largely by perpetrators who, particularly because of criminal pasts, are already prohibited from possessing firearms by existing statutes. Consider the case of Baltimore, where, according to the Baltimore Sun, 85 percent of homicide suspects in 2017 had a criminal record and an average of nine prior arrests. Nearly half of identified suspects had a prior gun offense.

Simply putting more anti-gun laws on the books would do little to solve this problem. What has been shown to make a difference is a proactive, data-driven approach to policing and placing a premium on enforcing gun laws already in place. And giving gun laws meaningful effect requires the political will to detect, prosecute and incapacitate violators.

This is what makes Bloomberg’s stop-and-frisk reversal so worrisome. That tactic resulted in the removal from New York streets of thousands of illegal guns — guns that were unlawfully possessed because they were banned, because the carrier was unlicensed or because of the possessor’s status as a felon or juvenile.

By stopping, questioning and, in some cases, frisking crime suspects, police were able to reduce gun crime in two ways. First, when stops led to an arrest and prosecution, the removal of gun-wielding criminals from the streets, even if temporarily, incapacitated them for a period of time. Second, the increased probability of being confronted by officers deterred many criminals from carrying guns and committing other offenses.

Many critics of the NYPD under Bloomberg argue that the absence of a spike in New York’s gun crime following the sharp decline in stops under Bloomberg’s successor, Mayor Bill de Blasio, demonstrates that stop-and-frisk is ineffective. Others argue against stop-and-frisk because of racial disparities in the department’s stop numbers, which, they say, proves racial animus on the part of police. But both criticisms miss the mark.

By the time NYPD reports began showing sharp declines in stops, New York was a different city than it was when Bloomberg took office. Thanks to several years of aggressive policing, New York had fewer dangerous neighborhoods than it did in the early 2000s. Dangerous neighborhoods are precisely where easing up on policing would have yielded an increase in crime; without that fertile ground, there were fewer places for new criminal activity to take root. In other words, by the time New York pulled back from stop-and-frisk, the sustained use of that technique — and other robust policing tactics — had worked well enough to provide the city with some lasting protection against a resurgence in crime.

That nuance was captured in a microgeographic analysis published in Criminology & Public Policy in 2015. This study found that the NYPD’s stops-and-frisks did, in fact, significantly deter crime in the hot spots within the city’s more dangerous neighborhoods — which is where the highest number of police stops were concentrated. These neighborhoods had — and, in some cases, still have — large minority populations, which goes a long way toward explaining the racial disparities in enforcement trends. What the study’s findings also show is that, because so much of New York’s gun crime was concentrated in minority neighborhoods, those communities also benefited the most from the Big Apple’s historic crime reductions.

If New York’s example illustrates the benefits of sustained proactive policing, Chicago’s 2016 crime spike illustrates the downside risk of an abrupt rollback of policing for a city with a higher concentration of crime hot spots. A recent analysis published in the University of Illinois Law Review found that the sharp drop in stops by Chicago police — which, in 2016, went from about 40,000 per month down to about 10,000 — was responsible for 245 of 274 additional homicides that year. These murders were concentrated almost exclusively in high-crime, mostly black and Latino neighborhoods.

If the available data tells us anything, it’s that you can’t prioritize reducing gun violence without also prioritizing criminal enforcement. It may be that Bloomberg’s contrition over stop-and-frisk is simply a rhetorical maneuver to preempt jabs from his Democratic rivals. But if he continues backing away from his record in New York, he’ll severely undermine one of his main arguments for sending him to Washington.

Read more: