California scholar Franklin Zimring explains why New York City's sustained crime decline offers a useful model for other cities.
Crime rates have been declining across the U.S. But no American city has matched New York's achievement, with overall crime dropping at over twice the national rate —and sustained over a 20-year period. The so-called New York “miracle” has triggered a cottage industry of criminological research into its roots and reasons.
The most authoritative study so far has been made by Franklin E. Zimring, William G.
Simon Professor of Law and Wolfen Distinguished Scholar at UC Berkeley School of Law. His latest book, published this month, The City that Became Safe: New York's Lessons for Urban Crime and its Control (Oxford University Press) has attracted nationwide attention and comment. In an author's Q&A with Joe Domanick, West Coast Bureau Chief of The Crime Report, Prof. Zimring offers his analysis of the city's remarkable accomplishment, explains why cops matter, and punctures some once-popular theories about how to deal with criminal behavior.
The Crime Report: Why did you decide to focus on New York City's crime decline?
Franklin Zimring: My last book The Great American Crime Decline, was a study of the national crime decline in violent and property crimes that lasted from 1991 to 2000— the biggest crime decline the country had since at least the Second World War. But the decline was really difficult to explain. You could shoot down a lot of theories, but you ended up without a really clear sense of either causes or effects.
The book pointed out that the nation's biggest city also had the nation's biggest crime drop. Crime had gone down everywhere, but it had decreased with special intensity in New York City. So I decided to focus on NYC.
TCR: And what was that special intensity?
FZ: There was a nationwide phenomenon, a 40% drop in crime everywhere in that nine year period after 1991. And part of New York's crime drop was in the tailwind of that national decline. But two things separated New York [from the rest of the nation].. The city's decline was twice as big as the national average, more like 80% than 40%. And the second was that ithas lasted almost twice as long. The general American crime decline ended in the year 2000. New York's went all the way through 2009. Now in 2009 and 2010 there seems to be some new national declines. But New York had the crime-decline business virtually to itself during the period of 2000-2007.
TCR: You describe NYC's crime reduction as a “Guinness Book of World Records Crime Drop.” What do you mean by that?
FZ: New York's crime decline over the period 1990-2009 is so dramatic we need a new way of keeping score. Instead of talking about homicides dropping by 82%, another measurement is to ask how much of New York's crime rate was left by 2009? New York in 2009 had a homicide rate that was 18% of what it was in 1990, a robbery rate that was 16%, a burglary rate that was 14%, and an auto theft rate of 7 % of the 1990 rate. A 93 % decline. And NYC has remained essentially the same city—the population didn't change a great deal. But crime changed more than any big city under circumstances of social continuity where statistics were reliable than had ever before been tracked.
TCR: And what makes you so certain about the validity of your statistics?
FZ: Four or five of the crimes I mentioned can be independently measured to test whether the police statistics are accurately representing time trends and accurately representing the magnitude of the decline. For homicide there are independent [statistics] from the health departments of the five different counties of New York City. Their [statistics] are very, very similar to the police. There's a .99 correlation. For auto theft, the independent confirmation was losses reported to insurance companies because of car theft. It wasn't quite 93% in 19 years, it was 88%, but we didn't control for an increase in population.
For robbery, the essential street crime, I tracked the number of killings attributed to robberies. Since the early 1990s killings during robberies dropped a little faster than the total robbery rate. Since you can't hide a robbery killing, it's a pretty accurate index of the volume of robberies. And finally we had victim survey statistics up until 2003. And the survey answers tracked the police statistics pretty clearly for burglary and robbery. What [all this] means is that this dramatic decline over time isn't an artifact of the statistical reporting system. It's real and it's on the streets.
TCR: You give a lot of credence to hot-spot policing? Can you talk about that?
FZ: There are two situations where we have solid proof that strategies worked beautifully. One of is 'hot-spot intervention'—sending police where crime is repetitively present at high rates, sustaining the police presence there, and putting out that fire. We know it works by a combination of New York and non-New York evidence. There've been very rigorous assessments of hot-spot policing in other cities like Atlantic City. And we know it's been well replicated in New York City. That combination of external evidence and New York practice makes me confident that the hot-spots approach is a proven success.
The second [strategy] that I'm sure worked is the destruction of public drug markets. Over the first 9 years after 1991, the number of cops in the narcotics bureau went up 137 percent—the largest growth in any policing effort in the city. The central strategy was not to just increase drug arrests, but to destroy the public drug markets that make the streets very dangerous and create a tremendous amount of fighting over turf. We know this works because there are specific indicators over time that we can measure. They indicate 90 percent or more of NYC's 1990 drug killings no longer existed in the city by the middle of this decade.
TCR: So does that mean that mean that heroin and cocaine use in NYC is down by 90 percent?
FZ: No. We have a lot of different measures of drug use: drug overdose death rates, emergency room drug mentions, the number of people treated for drug abuse. Back when we were debating a war on drugs, the hard-line legalists said the only way to reduce drug violence and drug-related crime was by substantial drops in the number of people taking heroin and cocaine. On the other side, public health strategists were saying that priorities ought to be going after the most costly things —drug violence, HIV transmission—and trying to find ways of targeting those problems rather than an all-out assault on drug use itself. New York used a harm reduction approach to reduce drug violence much more than it impacted drug use.
TCR: What additional New York Police Department (NYPD) reforms or new strategies do you think also contributed to the city's crime decline?
FZ: There are several [other] approaches I think probably worked well in NYC. One was increasing the number of street police. But this isn't a proven success, because New York's [crime decline] advantage over the other 9 biggest cities in the US was 21 percent [between 1991 and 2000]. After 2000, when NYC reduced its population of police by 4000, NYC's comparative advantage was still 21 %. That doesn't mean that the extra manpower wasn't a factor; there was a different scale measuring police per capita after 2000, because the crime rate was dropping faster than the police force's [numbers.]
Secondly, even with the manpower decline, many more cops were on the street, and police productivity probably increased. So it's probably been a success. COMPSTAT brought mapping and centralized decision making to the Big Apple, and this is also a very probable success. Not a proven success, because they didn't rigorously test it in Atlantic City, and it happened in NYC where lots of other things happened.
TCR: The NYPD became very aggressive in its policing, most controversially with its “stop and frisk” policies. Has this been a factor in the crime decline?
FZ: There were 700,000 stops and frisks in 2010 [in New York City]. Each year, the number of misdemeanor arrests has more than doubled. The NYPD has been using these arrests as a way of taking potential robbers and burglars off the street: you make a misdemeanor arrest, you get fingerprints, find out whether there's a warrant. The misdemeanor is used as a pretext for going after people believed to be serious street criminals. Now, everything that NYC cops did after 1993, they did very aggressively.
Hot spots policing worked, and was used with extreme aggression. We know that hot spot policing worked, but what we don't know is whether the aggressive way in which it was pursued created added value. The police think so, but again that hasn't been proven.
TCR: With “stop and frisk,” you've used an analogy of the late NYPD strategist Jack Maple, who said: “go after the sharks, not the marlins.” Hasn't the NYPD been going after the marlins as much as the sharks?
FZ: Certainly not in an order-maintenance situation. The misdemeanor arrests are of people the cops think are streets criminals. Of the number of people arrested for marijuana possession, 89 % are black and Hispanic and 93 % male. But almost half of the people carrying marijuana are female, and white rates of use are as high as black rates, and higher than Hispanic rates. So we know that the population that is getting busted are in high-risk areas, and secondly they look like high-risk folks. That might add value, but we don't know. But it would be good to find out, because it's costly in human and financial terms.
TCR: What strategies haven't worked in NYC?
FZ: Broken windows, or quality of life policing. That never went on in NYC. If you read the [James Q.] Wilson and George Kelling's [classic] article on broken windows, they maintain that police effort should be concentrated in marginal neighborhoods on the verge of falling into deep trouble, but that you shouldn't concentrate resources in places with the very worst problems, because they're beyond hope. Oh, really? That's the exact opposite of what the NYPD did with hot spot policing. If you only try [to save] marginal neighborhoods, nothing that you can do is going to create an 82 percent drop in homicides, because marginal neighborhoods aren't where the homicides are.
Point two: the NYPD was never much concerned with publically visible vice crimes. Street prostitution is the broken windows crime that is the perfect test of whether quality of life [enforcement] was ever tried, because it's not a good way of arresting robbers and burglars. But it is a good way of taking people who are obviously violating the law off the streets. After 1991 prostitution arrests never went up. The police force expanded by 40 %, but the vice squad was cut in half. So it isn't so much that broken windows policing didn't work in New York; it just never existed there.
TCR: What can other cities learn from NYC experience?
FZ: A number of things. The first is that cops matter a lot in the prevention of street crime. Twenty years ago that was in doubt. [The reasoning was] if you send three cops to 125th Street and 4th Avenue, the robbers will just go to 140th Street. Or if you send cops out on Tuesday, the robbers will wait until Thursday. This notion of dedicated and persistent criminality was why we thought cops couldn't make big permanent dents in crime.
But the fact is if we prevent a robbery at 125th Street on Tuesday that doesn't create an extra robbery on 140th Street, and it doesn't create an extra robbery on Thursday. The impulses that govern crime rates at the margin are much more situational and contingent than we thought. So if you prevent a robbery on Tuesday at 125th Street, there's one less robbery in NYC that year. And that's something that's also true in San Francisco or Toledo. Look at the crime control legislation of 1994. All the conservatives were laughing at midnight basketball. [The attitude] was if you have robbers play basketball on Friday at midnight, they'll just rob people on Saturdays. But it turns out if you're eliminating a robbery on Friday, it's probably going to eliminate a robbery for 2011.
So cops matter because criminal offenders are much more situationally responsive and contingent in what they're going do, and how much they're going do, than we thought. That's very good news for the police, probation officers and anybody invested in the individual changeability of criminal offenders.
TCR: What about the impact of mass incarceration on the crime rates?
FZ: The data shows that the criminal activity of people coming back to NYC from the prisons dropped as the crime decline proceeded. In 1990 the odds that a prison released from prison coming to NYC would get reconvicted of a felony over the next three years was 28 %. But over the next 17 years, the odds of being reconvicted of a felony dropped to 10 percent.
TCR: And that is attributed to…?
FZ: Call it an environmental effect. If all your friends are doing less crime and you're hanging out with them, so are you. Here's the point: imprisoning people for long periods of time because you're positive that they're at risk of committing a lot of crimes over long periods of time [is problematic.] Why? Because in New York City the cost-effectiveness of locking up people like that has gone down by 64%. In other words, the amount of crime they would have committed if they weren't locked up turns out to be variable and rather low in New York after 2006.
TCR: Why?
FZ: The street situation changed and so had the things that their friends were doing. People were now smoking marijuana and drinking wine. Cocaine use was down. Street robbery has gone down 84 %. Burglaries 86 %. And that meant that the people that the released offender used to hang out with as a persistent offender from a high-risk neighborhood, are no longer doing those things. So he's not doing crimes with them.
What I want to suggest is this: if the crime rate is down 84 percent and the population hasn't changed – which it hasn't— then despite the fact that these statistics on individual crime risks are dramatic, they were also inevitable. I should've expected these fabulous changes in individual crime rates, because as I said before, individual crime propensities are situational and contingent.
TCR: So are expensive policies of mass incarceration that states like California have pursued to prevent crime cost-ineffective?
FZ: We used to think that all we could do with high-rate offenders is lock 'em up or they're going to offend on the street. But NYC has 28 % fewer people locked up in 2011 than in 1990. And it has 80 % less crime. The [individual] criminals didn't go anywhere. They're just doing less crime. So the bedrock of prediction on which incapacitate imprisonment was built, has turned out to be demonstrably false. And the proof of that is in New York City.
TCR: And what are the implications for harsh mandatory-minimum laws like California's three strikes law?
FZ: People who are really interested in three strikes' lessons should be studying New York City as much as they're studying California.
Joe Domanick is West Coast Bureau Chief of The Crime Report, and Associate Director of the Center on Media, Crime and Justice. He welcomes comments from readers.