By Franklin Zimring, New York Post
The drop in street crime in New York City after 1990 is not only the largest decline ever documented in a major city but also a major test of the conventional wisdom that has dominated crime policy in the United States for a generation.
Between 1990 and 2009, the rate of murders, robberies and burglaries fell by more than 80%, while auto theft in 2009 was down 94%. Such huge declines in crime require a new method of score-keeping — instead of measuring the size of the drop, it is more informative to ask what proportion of the old crime level remains, and the answer in most cases is “not much.”
In less than 20 years, the robbery rate is 16% of its 1990 high, the auto theft rate is 6% of its former level. The homicide rate is 18% of its 1990 high, and lower in 2011 than it was in 1961.
And these amazing statistics are not just creative accounting by the police — independent tests available to check police crime statistics confirm both the size and the timing of the official crime statistics for homicide, auto theft, robbery and burglary. Gotham is now a safer place to live in 2011 than half a century ago.
Part of New York’s good fortune was the tailwind of a national crime decline during the 1990s, but the New York decline was twice as large and almost twice as long as the national drop. Why was that? What can we learn from this experience to help other cities?
A close reading of the history in New York City points to four policy surprises. Institutions that were thought to be ineffective (like police) produced great results while the major instrument of crime control everywhere else in the United States — expanding prisons — were not needed in New York. Here are the city’s four unexpected lessons about combatting crime.
1 Not all “broken windows” are created equally
Twenty years ago, social scientists believed that police efforts couldn’t make a substantial dent in urban crime, because police couldn’t be everywhere. Criminals could simply outwait the police or relocate.
But even a careful discounting of New York City crime numbers — taking out the 40% crime drop that most places experienced in the 1990s, as well as the effects of population change in Manhattan — still leaves plenty of credit for the cops.
Almost half of the city’s record-breaking drop in robbery and burglary and almost a third of the disappearing auto thefts are the product of better policing, statistics show. But why did New York’s police make so much of a difference?
It turns out that criminals aren’t quite as persistent as we used to assume. If police prevent a robbery Tuesday night on 125th Street, that doesn’t mean an extra robbery on 140th Street or even next Thursday. Instead, the impulses that produce muggings are situational and contingent and that means the short-term interventions can have long-term consequences for crime rates.
The epic success of police in New York came at a time when the number of cops increased, their strategies of enforcement changed and street police became much more aggressive. We still don’t know with precision which of these changes had the most dramatic results. In particular, we don’t know whether large numbers of misdemeanor arrests and massive stop-and-frisk campaigns add value to preventive policing.
But the popular theory that New York used “order maintenance” or “broken windows” policing strategies is not correct. The broken-windows approach concentrates efforts on marginal neighborhoods rather than the highest crime areas where the situation is hopeless. But New York’s strategy concentrated cops in exactly those “hot spots” of the city where violent crime kept happening.
New York cops used arrests for minor crimes and disorderly conduct not to maintain order but to get the fingerprints of people they suspected were threats to commit more-serious crimes.
And it turns out that not all lower level offenses are created equal in this regard.
Marijuana arrests jumped in the mid-1990s to a peak more than 10 times the 1990 volume and stayed very high. Whereas gambling arrests increased between 1997 and 2001, then dropped off suddenly and continued to decline through 2009.
The most plausible interpretation is that police used gambling arrests as part of an aggressive patrol for some time but didn’t find the persons brought in to have records, warrants or other indices of danger that were targets of the policy.
Prostitution, meanwhile, would appear to be a classic affront to public order. Yet the rate of prostitution arrests never got higher than it is in 1991, even as the police force expanded. Why was prostitution never a priority while marijuana misdemeanor arrests skyrocketed? The only plausible answer is that order maintenance and quality of life were labels but not real motives for the pattern of arrests.
As Jack Maple, one of the designers of the NYPD’s policing strategy in the 1990s, once said, it’s about looking for “sharks” and “not the dolphins.”
2 New York won its war on drug violence without winning the war on drugs
One cause of the epidemic of violence that engulfed New York City in the late 1980s was the disorder and conflict that came with the rise of crack cocaine in the mid-1980s.
There were two contradictory theories about how best to combat many problems generated by urban crack. The hardline view of drug warriors like national drug czar William Bennett was that the only way to combat drug violence was to substantially reduce all forms of illegal drug use — more focused priorities were considered halfway measures doomed to failure.
But public-health analysts argued instead that the government should isolate the most threatening problems associated with drugs — things like drug killings, the takeover of public streets by open air markets, AIDS transmission — and produce concentrated efforts to reduce these harms. While this “harm reduction” approach was not associated with police in the 1980s, it was the NYPD that proved harm reduction worked.
During the 1990s, police increased narcotics unit manpower by 137% and dedicated their efforts to destroying open-air drug markets.
This gave the streets back to the public and also reduced drug-related killings by 90% as the turf wars associated with open-air markets ended.
Yet drug sales continued indoors. Most indicators of hard drug use in New York show stable trends in drug use over the past two decades — including emergency-room drug mentions, hospital drug abuse treatment and urine tests of criminal defendants.
The number of drug-overdose deaths in 2004 and 2005 is 90% of the volume in 1990 — indicating pretty flat trends in the use of at least the high-lethality illicit drugs. The volume of drug-involved homicides in 2005, meanwhile, is only 5% of the number in 1990.
The police were happy enough with this outcome to reduce the manpower in the narcotics unit in 2008 back to fewer officers that were on duty in 1990. Drug harm reduction was a proven success in Gotham.
3 Cutting crime doesn’t require increasing imprisonment
The sevenfold expansion of imprisonment in the United States since 1972 was based on the assumption that high-rate criminal offenders could only be controlled if they were locked up. But New York City has cut its crime rate by 4/5ths since 1990 while reducing the number of persons in its prisons and jails by 28%. The incarceration rate in the rest of the United States increased by 65% during the same period.
Over the 11 years after 1997, prisons and jails sent 20,000 more persons back to the streets of New York City than they admitted.
If national trends had been followed, by 2007 New York City would have been locked up 58,000 more people than it did. At a modest $25,000 per person per year, the public savings of that many fewer prisoners istwice the cost of the expansion in policing.
That the city that used incarceration much less than its neighbors yet ended up reducing its crime much more than other places turns the conventional wisdom of American crime control upside down.
But where have New York’s robbers and burglars gone if not to prison?
Here’s one major clue — 28% of felons sent back to the city in 1990 were reconvicted of a felony in the next three years, but by 2006 the re-conviction rate had dropped to 10%. This signals that the personal crime rates of these offenders have dropped by 64% in 16 years. Those who used to be active offenders commit far fewer crimes without being locked up.
The problem with all these statistics is that they are after the fact of behavioral changes. There is little direct observation of changing patterns of street life in former high-crime neighborhoods. There are anecdotes from drug-scene workers of younger persons mixing marijuana and alcohol instead of crack in the mid-1990s, but no sustained studies.
While many questions remain, however, it’s fair to say that lower levels of crime lead to lower levels of crime. What we used to call “career criminals” are much easier to alter than had been believed. The crime rates of mid-career offenders are quite sensitive to upward and downward variations in general crime rates.
The same situational and contingent forces that work for others deter persons with substantial criminal records. In other words, crime doesn’t pay — at least not as much as it used to.
4 Small changes can make big differences in urban crime
Not all the civic changes in the last decade have been positive. The 2011 report card for urban progress is uneven at best — jobs are scarce, schools are problematic, single parent families are common and economic inequality is rampant. But violence and the imprisonment of the poor have both dropped by huge margins.
Even the major changes in police force levels and tactics haven’t been of overwhelming scale — it’s a few thousand more police in a city of 8 million. The people and places and patterns of life in the city have changed only modestly. And yet the threat of murder and robbery and home burglary have dropped not by half but by more than 80%. If this is a “New York miracle,” it is much closer to the loaves and the fishes than the parting of the Red Sea.
The objective of programs like closing open air drug markets and intensive patrol of hot spots is to make small changes in the environment in which criminal decisions get made. Proponents of long-term incarceration could scoff at police closing two or three square blocks of a city to public drug traffic because they argue that persistent offenders can always buy or sell drugs elsewhere. And there is evidence that drug purchases and sale and consumption continued at a pretty steady pace in New York. But drug killings dropped, and there were fewer public settings residents consider dangerous and inaccessible.
Even though the extra police presence is transient and the extra police numbers are not huge, most criminal offenders seem responsive to modest and even temporary alterations in the environment of the city. The further removed people who may commit a crime are from hardened criminals, the better the prospect that any attempt at intervention will work.
The broad lesson that emerges from the huge drop in New York is the variability of most crime to modest changes in circumstances. That doesn’t guarantee that after-school tutoring will keep kids from failing in school or from being vandals. But it shows that much of the population at risk is malleable, their propensities prone to change with modest provocation. Anything that is prone to work at all can work well with a population not that difficult to change.
Epidemic levels of crime and violence are not hardwired into the structure of urban life and that is good news for cities all over the world.