The Dead of Night

IN THE LATE 1980s, criminologist Lawrence Sherman — then a professor at the University of Maryland, now the director of the Jerry Lee Center of Criminology at the University of Pennsylvania — had a hunch. The basic facts of urban violence were not unlike what they are today. There were around 70 million handguns in circulation in the United States, many of them illegal and concentrated in the inner city. (Today, there are probably more than 100 million.) And the market for those guns was no less flexible than it is now: It seemed that no matter what efforts were made on the supply side — from permit denials to waiting periods to concealed-weapon bans — handguns, through theft or interstate straw purchase, found their way into the hands of criminals. Even if production was stopped dead, and not a single new handgun was ever manufactured or sold in a gun store again, there would be more than enough already on the streets to keep criminals well supplied for decades. But Sherman suspected that if the quantity of guns in circulation couldn't be changed, it was possible that gun carrying could. Under grants from the Bureau of Justice Assistance and the National Institute of Justice, he conducted an experiment in Kansas City, Missouri, in 1991. His hypothesis was that police officers, by aggressively enforcing laws against carrying concealed weapons at crime "hot spots" such as drug corners and nuisance bars, would discourage people from carrying guns, and that this decrease in gun-carrying would lead to fewer gun crimes. This was a great departure from the standard model of rapid response and random patrol.

Sherman's first step was to identify the "hot spots." By itself, this verified an important prior finding: Just as there are repeat offenders who commit a disproportionate amount of violent crime, so there is a relatively small group of repeat locations in cities where there's a disproportionate amount of gun violence. During a previous study in Minneapolis, Sherman found that 420 of the city's street corners had 20-plus "hard crime" 911 calls each. Once he mapped those addresses using a computer program, Sherman could group them into just 110 clusters or "hot spots" for the entire city. The hot spots were usually around bars or stores open very late. The same went for Kansas City. There were two geographically distant but demographically similar neighborhoods — called Beat 144 and Beat 242 by the police department — that had the city's highest concentrations of gun crime. Sherman, with his team of researchers, identified the hot spots in the target area, Beat 144.

The second step of the experiment was to take four Beat 144 police officers and have them look for guns. These were uniformed officers, using two squad cars, working overtime between 7 p.m. and 1 a.m. seven days a week, paying particular attention to the hot spots Sherman had identified. The officers looked for behaviors that might indicate the presence of a gun — walking down the street on a summer day with a heavy coat on, for example — but also paid attention to minor offenses, such as running red lights or public drinking, that gave them a chance to see whether a person was behaving suspiciously enough to justify a frisk. Beat 242's patrolling remained unchanged.

Sherman's study describes how right away, the hot-spot officers of Beat 144 began to find guns. A man runs a stop sign, and the police pull him over. As the driver reaches for the glove compartment, an officer sees a hard bulge under his jacket. A frisk reveals a handgun. Another driver is stopped for speeding. The officer shines a flashlight onto the backseat, and sees a shotgun. Yet another driver runs a red light. A computer check shows the man is wanted on an assault charge. The driver is arrested, and a frisk uncovers a pistol. A man is urinating on the side of a car. Officers charge him with public urination, frisk him, and find a handgun.

At the end of the six-month experiment, gun seizures in Beat 144 had increased by 65 percent. More importantly, gun crime had decreased by 49 percent. There was only one murder by handgun in all of Beat 144. Gun crime in Beat 242 — the area in which policing was unchanged — was almost identical to what it was the year before the experiment. And by tracking gun crime incidents in the areas surrounding Beat 144, the researchers determined that crimes truly had been prevented in Beat 144, and not simply displaced to these other neighborhoods. The experiment was repeated in four more cities, with similar results: Gun violence plummeted in areas with gun patrols, and remained nearly the same in areas without it.

This story may sound familiar to any New Yorker who lived through the early years of the Giuliani administration. Quality-of-life crimes such as panhandling and turnstile-jumping were enforced; high-rate offenders were found and locked away in the process; guns were taken off the street; and crime rates fell. But the New York experience is problematic for criminologists because so many things were going on at once — declining crack use, a flourishing economy, demographic shifts as immigrants and yuppies moved into once-dangerous neighborhoods, and yes, the famous change to "broken windows-style proactive policing — that it's been hard to say which had the most effect on declines in crime. That has been the significance of Sherman's work. He didn't say that the police can restore the nuclear family, eliminate poverty, bring back the industrial economy, or in any other way assuage the "root causes" many people think are behind urban violence. He made the narrow point that a police department can, by devoting some of its manpower to proactive, hot-spot gun patrols, have a demonstrable impact on the percentage of a city's murders that are committed with handguns.

The argument wasn't that so many thousands of guns are seized that criminals don't have access to them. It's that the probability of being arrested with a gun, or having a gun taken, will cause a young man not to carry it in the first place. And because so many shootings are spontaneous — a 19-year-old gets called a name in public and responds with his Bryco semi-automatic pistol — there is real prevention when a gun is left at home (or in a school locker, as is often the case). The key is to convince gun carriers that they will be stopped and their guns will be taken. That requires consistent, proactive patrolling and frisking for weapons where illegal gun-carrying takes place, and that is why many cities that have seen dramatic crime drops have done exactly that. In a 2005 comprehensive review of the research on preventing gun violence, the National Academy of Sciences called Sherman's work "compelling" and "well-designed," and said that gun patrols in hot spots have "substantial crime control effects." The title of a study from a prominent Georgetown University researcher that supports Sherman's work puts it succinctly: "Better Gun Enforcement, Less Crime."

Not long ago, I interviewed Sherman. We spoke in the Penn criminology department's library, which, until a recent move, was on the second floor of a brownstone on Walnut Street. "There was a shooting of a student at three o'clock in the morning right down there this February," Sherman says, pointing out the window. "It was a stray bullet from some drug dealers shooting it out from the corner. So we live in our laboratory here."

I ask about gun patrols.

"Looking for guns on the street is not a lock-'em-up strategy," he says. "It's not a fill-the-prisons strategy. It's a specific and focused deterrent strategy that is trying to deter one thing — and that's people carrying guns around. Because if people don't carry their guns around, and somebody bumps into them and doesn't say 'Excuse me,' or somebody looks at them with a stare that they find offensive, then they may have to go home to get their gun to do something about it. But by the time they come back, the person may not be there, and the impulse may pass."

That "stare that they find offensive" is the sort of thing over which young Philadelphians have so often been dying lately — "Stupid arguments over stupid things," as police commissioner Sylvester Johnson recently put it.

"Sometimes fights are followed several days later by an assassination, so there's no guarantee that people won't get shot when those disrespecting incidents occur," Sherman says. "But what the evidence suggests is that it's going to happen at a lower rate if they don't have their guns in their pockets."

And what about Philadelphia?

"The police in Philadelphia are not looking for guns in the street, and that has a context in this 40-year story that we have to understand," Sherman says. "But we also have to understand that as far as the evidence is concerned, the National Academy of Sciences says it is the one thing we know works to reduce the homicide rate. And it's the one thing we're not doing in Philadelphia."

---