We know that correlation does not imply causation, but lack of correlation does imply a lack of causation, and it appears that stop and frisk falls firmly into the latter category:

If you grew up in New York City in the 1970s, the number can be hard to get your head around: 291. If you were a reporter in New York City in the early 1990s, the number can almost make your head explode: 291 murders in 2017, the lowest total since the 1950s.

But the number is perhaps most striking when set not against the numbers of murders in other years, but against this figure: the roughly 10,000 police stops conducted in 2017.

The longstanding rationale for the New York Police Department’s widespread use of what came to be known as stop-and-frisk — encounters between officers and people they suspected of suspicious behavior — had been that it was an essential crime-fighting tool. Such stops got guns off the street, the theory went, and low-level enforcement helped sweep up criminals destined to commit more serious crimes.

………

Ultimately, a federal judge, Shira Scheindlin, found the NYPD’s enforcement of stop-and-frisk racially unfair and unconstitutional. A new mayor, Bill de Blasio, and the judge’s orders for reform, prompted a radical scaling back of stop-and-frisk. Critics predicted a disastrous return to, depending on one’s age and experience, the 1970s or the 1990s.

The disaster never happened. Instead, what many scholars and police officials thought nearly unthinkable — further reductions in crime after two decades of plummeting numbers — did.

Holding murders under 300 was just the headline of 2017 statistics that saw considerable reductions in almost every category of major crime.

Stop and frisk was never a tactic to reduce crime, though the naive might have believed that, it was a way to keep people of color down.  (With a bit of law-enforcement theater thrown in.)