Republican presidential nominee Donald Trump says he favors "stop-and-frisk." Here's what that means and why the policy is so controversial. (Victoria Walker/The Washington Post)

Presidential candidate Donald Trump said Wednesday that one way he’d fight crime would be to restore the police practice of “stop-and-frisk”: an officer sees someone he thinks might be about to commit a crime, so he stops them, questions them and pats them down for weapons or contraband. This practice was implemented in New York City and elsewhere in the 1990s, and in New York it was discarded after a judge ruled that it unconstitutionally targeted minorities and Mayor Bill de Blasio in 2014 ordered it stopped.

There are two things we know about “stop-and-frisk” in 2016: Crime experts completely disagree on whether it has an impact on crime, and it is disproportionately used in minority communities almost everywhere it’s used. And when communities feel targeted by police, their trust in police declines, tensions rise and relationships between citizens and officers deteriorate.

Here’s what Trump said in a taped town hall session on Fox, in response to a question about reducing violence in the black community: “One of the things I would do is I would do stop-and-frisk. I think you have to. We did it in New York, it worked incredibly well and you have to be proactive and, you know, you really help people sort of change their mind automatically…In New York City it was so incredible, the way it worked.”

Now it should be noted that presidents do not dictate policy to local police departments, and getting them to return to a practice they are largely discarding would seem unlikely. It also should be noted that statistics in New York showed about 90 percent of those stopped hadn’t done anything wrong, so that that arresting the few meant annoying the many, often repeatedly.

Criminologists differ on whether “stop-question-and-frisk” (as it’s formally known) worked, even after New York police stopped doing it, and the number of stops began to plummet in 2012. The Brennan Center at New York University published a study in April which showed that homicides continued to decline sharply in New York even after 2012, from almost 600 in 2006 to 350 in 2015, and that total crime in New York also fell even after stop-and-frisk was reduced drastically.

But others say the practice worked. Criminal justice author Heather Mac Donald pointed out that New York’s crime rate per capita was significantly lower than many cities where stop-and-frisk wasn’t used. “New York’s most vulnerable residents enjoy a freedom from assault unknown in any other big city,” Mac Donald wrote in 2013, “thanks to the N.Y.P.D.’s assertive style of policing.”

An extensive study by George Mason University criminology professor David Weisburd last year declared, “it is time for scholars to recognize that [stop-and-frisks] focused on microgeographic hot spots are likely to reduce crime.” But Weisburd added, “The question is whether this approach is the best one for crime prevention at [larger] hot spots and whether its benefits are greater than its potential negative impacts on citizen evaluations of police legitimacy.”

And that’s where stop-and-frisk ran into a problem. Statistics showed that it was disproportionately deployed against black citizens. In New York, statistics showed that police stopped black citizens 52 percent of the time Hispanics 31 percent and whites 10 percent, though blacks make up only 23 percent of the city population, Hispanics 29 percent and whites 33 percent. The policy became a central issue in the 2013 mayoral election, and after a federal judge ruled the practice unconstitutional, new Mayor DeBlasio ordered it stopped.

A similar practice was detected in Chicago, where black residents were stopped 72 percent of the time though they made up only 32 percent of the population. In 2015, the Illinois legislature passed, and the governor signed, “The Police and Community Relations Act,” requiring police to collect data on stops that resulted in a frisk or arrest and that officers issue a receipt to people whom they frisk. In August 2015, Chicago police reached a settlement with the American Civil Liberties Union to reform their stop-and-frisk practices, hire an independent evaluator and improve training.

But violence in Chicago in 2016 has erupted, with more homicides through Labor Day this year than all of last year.

Stop-and-frisk tactics were also recently criticized in Baltimore, where a Justice Department report in August said it was used both disproportionately and often with heavy-handed tactics. And similar allegations were made in Newark, where the city in March agreed to sweeping reforms under the supervision of a federal monitor.