In 1988, Willie Horton–a furloughed Massachusetts criminal who had skipped to Maryland where he robbed, raped, and pistol-whipped his victims–starred in a campaign ad used to tag Democratic presidential nominee Michael Dukakis as soft on crime. The ad was a high (or low) point of the campaign, but almost 20 years later, it seems almost quaint. Looking back from 2007, after a midterm election in which crime was a non-issue, it is easy to forget how salient a theme it once was in national politics. And yet the Horton ad was successful because it appeared at the crest of the rising fear of crime in American cities, which spanned the late 1970s through the early ’90s. State and federal politicians prescribed long criminal sentences, most notably in the Federal Sentencing Guidelines and in California’s “three strikes” law, upheld in 2003 by a sharply divided U.S. Supreme Court.

In the years following the Horton ad, conservative thinkers warned that the nation would–because of a rising teenage population–soon
experience an unprecedented crime wave. In 1996, William Bennett, John DiIulio, and John Walters declared that “America’s beleaguered cities
are about to be victimized by a paradigm-shattering wave of ultra-violent, morally vacuous young people some call ‘the super predators.’ ” But their heated rhetoric was published just as American crime was beginning to decline. Every year between 1992 and 2004, the FBI reported a national crime drop of several percentage points. Although declines occurred nationwide, the most noteworthy was in New York City, whose “mean streets” and “needle parks” had long been etched into the national consciousness through film and TV screens.

The question remains, however: Why did crime decline so significantly? There are no uncomplicated answers, as Franklin Zimring, a law professor and criminologist at the University of California, Berkeley (who, in the interests of disclosure, has been a friend and colleague over many years) convincingly demonstrates. Zimring, the author of 18 notable books on various aspects of crime and crime policy, has once again produced an important, carefully reasoned, and readable book. Contrary to the beliefs of many armchair theorists, there is no simple and consistent explanation for the crime decline, Zimring argues. He is skeptical about the effects of the big three “usual suspects”: imprisonment, the economy, and demography. Each, he acknowledges, has some effect, but none has proven consistently meaningful. There are no unwavering correlates of crime, and thus it is difficult, if not impossible, to develop a public policy prescription for cities across the nation.

That said, Zimring concludes, in at least one city the explanation is clear: The New York City Police Department’s size, resources, use of
statistical controls, and aggressive policing have combined to account for that city’s extraordinary crime reduction. Indeed, New York’s crime
rate continues to decline, while–as I write this review in late December 2006–the FBI has reported that crime is rising in cities across the country. Thus Zimring’s thesis comes at a crucial moment–while he doesn’t provide solid answers, he knocks down simplistic theories that might serve to mislead policymakers as they once again confront the mystery of American crime rates.

It is important, in any discussion of crime rates, to ask what constitutes “crime.” Zimring, like most criminologists, examines FBI data (gathered nationally from police departments) on crimes with victims: homicide, rape, robbery, aggravated assault, burglary, auto theft, and larceny. These data have two strengths. They allow comparison of trends through time and facilitate comparison of different parts of the United States and Canada. And they permit regression analysis, a statistical method of investigating the relationship between variables, here between crime and its possible causes.

But some cautions are also worth noting about Zimring’s data. They rarely extend beyond the year 2000. This makes his analyses less timely, but not necessarily less important. Nor does he analyze white-collar crime, which might well have increased during the economic boom of the 1990s. Nor can he–or any criminologist–accurately parse the rise and fall of consensual crimes, such as prostitution, or more importantly, drug-selling. Although drug-related crimes constitute a significant portion–as much as 40 percent–of arrests, convictions, and imprisonments in the United States, purchasers of drugs (or sex) rarely, if ever, have their transactions reported to the police. Consequently, there are no numbers to report, and most drug crime does not appear in the FBI crime rate. But drug crime does have a strong influence on the imprisonment rate, which may account for the seeming paradox of higher imprisonment even as the FBI index crime declines.

Setting these qualifications aside, the key contribution of Zimring’s book is not to prove a single theory of crime causation, but rather to show how difficult it is to isolate a single factor as a sufficient explanation. Look at the three most popular reasons for fluctuations in the crime rate: imprisonment, demography, and the economy. Among national and local legislators, lengthy sentences are justified for reasons of justice, incapacitation, and deterrence. Like most criminologists, Zimring does not deny that imprisonment is justifiable; however, he questions its marginal effectiveness. We need prisons, but do we require as many as we now have? Do we need such lengthy sentences? Imprisonment is expensive (prisoners may be one of the few groups with guaranteed health care), and especially, as under California’s capacious three-strikes law, it diverts funds from other public needs, such as education. Most of our sentencing policies, in fact, have nothing to do with policy analysis–rather, they are gut-level responses, as politicians don’t want to be susceptible to a charge of being “soft on crime.”

Along with rates of imprisonment, criminologists generally consider the size of the young male population as a correlate of crime. Statistically, the most potentially dangerous person in America is an 18 or 19 year-old male (a feminist friend suggests that they suffer from testosterone poisoning), and young black males are much more likely to be homicide perpetrators–and victims–than young white males. Yet predictions based on the future size of the teen-age male population can mislead, as we saw with predictions of the “super predators” who would supposedly generate crime waves in the early twenty-first century. Indeed, crime rose during the 1980s while the size of the male youth population declined.

How about the economy? Once again, the effect is complicated and unclear. The economy might improve for the haves and the have-mores, but not for those solidly below the median, a more accurate measure than the arithmetic mean. But even if the gains were more equitably distributed, it would still not explain the drop sufficiently: A well-regarded British study suggests that property crime rises in a poor economy, and violent crime climbs in a better economy, as teenagers have more disposable income with which to purchase alcohol and drugs.

Zimring identifies a statistical tendency–regression to the mean–as another possible crime cause. Derek Jeter’s batting average will fluctuate from game to game, but his true average will emerge at the end of the season. Similarly, high crime rates will likely revert to some hypothetical mean and decline, and low crime rates will rise. But when criminologists are asked to explain the fluctuation of a few percentage points, they rarely, if ever, offer “regression to the mean” as the possible cause.

How about unwanted birth as a cause of crime? Consider the analysis by best-selling economist Steven Levitt and his co-author, John Donahue, in a 2001 Journal of Economics paper. They attribute as much as half of the 1990s crime drop to the post–Roe v. Wade abortion rise in the 1970s. There is no denying that abortions increased in the 1970s and crime decreased in the 1990s. Assuming more abortions took place after legalization (which hasn’t been proved), there would be fewer unwanted babies born. Are unwanted babies–economists might call them “babies born under suboptimal conditions”–more likely than wanted babies to be criminals? Perhaps. But so many other factors influence crime rates, Zimring concludes, that it’s difficult to specify the effect of Roe v. Wade on the crime decline or how much demography influences criminal destiny. Ultimately, there are many plausible, partial explanations. Indeed, Zimring writes that an era of crime decline is likely to encourage a “euphoric fallacy” in which we assume “that all the good news was the natural and probable result of last year’s change in police manpower or criminal punishments.”

Ironically, the great American crime decline may be reaching its end, and some experts are predicting that the current rise in juvenile crime could be a harbinger of another crime wave. In October 2006 (and in prescient anticipation of the December FBI report) the Police Executive Research Forum (PERF), an organization of urban police chiefs headed by Los Angeles Police Chief William Bratton, published a report whose title, “A Gathering Storm–Violent Crime in America,” clearly telegraphs its conclusions. But if one looks closely at the PERF report after reading the Great American Crime Decline, it turns out that both titles are a bit misleading.

The great American crime decline was real, and it was national. But it was about half a large as New York’s; indeed, as a rule of thumb, Zimring suggests, take the average drop in any crime category in the nation. If you double it, you will arrive at New York’s. Indeed, even as crime is on the up-tick in most cities across America, the gathering storm seems, at least for now, to have bypassed New York City. As of late November 2006, New York City’s FBI index crime rates have declined by 4.88 percent for the year and by 17.46 percent between 2001 and 2005. And the storm seems to have passed Los Angeles as well; in mid-December, that city reported a 29 percent violent crime decline over the previous two years.

It took some time to convince Zimring and other criminologists, including this writer, that policing could actually reduce crime. Nor would we have predicted that New York City would be the site of a singularly steep crime decline. There were good reasons to be skeptical, especially in a place that Bratton–who served as New York’s police commissioner from 1994 to 1996–once called “the crime capital of the world.” Social scientists attributed the rising crime of the late 1980s and early ’90s to underlying social and economic conditions, as well as to the crack epidemic, and we weren’t entirely wrong. But Zimring notes that New York City had 75 percent less homicides, robberies, and thefts of autos in 2006 than it had in 1990, but essentially the “same populations, schools, transportation and economy.”

Zimring credits the NYPD’s size as a major factor in its crime decline. The NYPD began growing significantly at the end of the David Dinkins mayoralty and rose even more during the Giuliani Administration. At one point, the NYPD had more than 40,000 police officers. To be sure, crime control does require a critical mass of police. But, at the time, a larger force did not seem a sufficient solution. Over roughly the same period, Washington, D. C., had a higher ratio of police to population than New York, but it nevertheless had a higher crime rate. In other words, numbers matter, but they’re not enough. What New York showed us was that those numbers need to be marshaled in the interests of crime control and community acceptance–and that demands first-rate management, which New York had.

Traditionally, the NYPD’s size, coupled with its decentralization, made it hard to manage. Before computerization, months could pass before crime statistics from its 76 precincts would be hand-delivered to One Police Plaza, the department’s central headquarters. And there was a time when New York cops routinely were found to be lazy and corrupt. The scandals investigated by the 1972 Knapp Commission began when a New York Times reporter discovered that police engaged in “cooping,” resting in out-of-the way places when on duty. Twenty years later, the Mollen Commission uncovered corruption and brutality and discovered “testalying”–a routine practice of giving false testimony, especially in drug-possession cases. All of which makes the NYPD’s world-renowned and relatively scandal-free record of crime control over the last dozen years all the more remarkable–and helps explain why better management, not just bigger numbers, was the key to New York’s crime decline.

The most important of Bratton’s managerial innovations was Compstat, a sophisticated crime-mapping instrument. The data offer top brass
timely and accurate information to hold lesser-ranking managers, especially precinct captains, responsible for reducing crime and addressing quality of life problems in their precincts. I attended some of these early meetings between department leadership and precinct commanders, which could lead to personnel changes when the latter failed to answer tough questions. By now, and especially since Raymond Kelly became commissioner in 2002, the briefings have become more institutionalized; precinct commanders have learned that if their crime numbers rise, they will be held accountable. Moreover, current precinct commanders have evolved and been promoted within the system. Consequently, they have been given more authority over the 200 to 400 officers they manage, leading to more efficient day-to-day management.

But while these steps have been successful, they have also generated problems. Bratton’s most controversial innovation was “broken windows”
policing, a tactic suggested by James Q. Wilson and George Kelling in a famous 1982 Atlantic article. Wilson and Kelling argued that if
you accepted a neighborhood’s declining appearance and street behavior, crime would follow. This led Bratton to insist upon the enforcement of “quality of life” offenses, such as spray-painting, drinking alcohol in the street, loitering, panhandling, prostitution, and loud music. Such
offenses take place mostly in high-density minority neighborhoods, which tend to have a more active street life than more affluent areas.

But since such aggressive policing–sometimes called “order maintenance”–often takes place in neighborhoods populated by by people of color, it needs to be tempered by community acceptance. When the trial of four white New York police officers accused of murdering Amadou Diallo in a hail of bullets was drawing to a close in February 2000, Bratton, by then no longer New York’s police commissioner, wrote a New York Times op-ed article criticizing the NYPD’s aggressive stop-and-frisk policing. Such policing, he wrote, makes sense when crime is rampant. But, he added, as New York City’s crime declined, its minority population “had every right to expect that one of the benefits of the safer city would be less police intrusion into their everyday lives.”

Nevertheless, the Bratton-era reforms have been unarguably successful, and the enduring success of Compstat-influenced policing strategies is borne out by the department’s successful response to the post–September 11 security challenges. It is all the more significant because the two men who led the department up to and for a few months after September 11–Mayor Giuliani and his police commissioner, Bernard Kerik–left office less than four months after the attacks, turning over the department to Mayor Michael Bloomberg and Police Commissioner Kelly.

When he took the reins of the NYPD in January 2002, Kelly faced a daunting challenge: to maintain low crime rates while addressing the threat of further terrorist attacks on the city, and settling into the job. He created an intelligence division with new deputy commissioners, initiated new anti-terrorism strategies, and stationed agents in such cities as London, Paris, Tel Aviv, and Amman. Precinct commanders, especially those in Manhattan, were given new anti-terrorism responsibilities. Thousands of officers had to be specially outfitted and trained to respond to chemical or biological or radiological attacks. But because he was building on a tried-and-true structure, Kelly succeeded in making New York the safest large city in the United States.

He also made it significantly less tense, at least racially. Kelly understood the importance of being fair and appearing fair. When dicey
incidents arose, such as the shooting and killing of 19-year-old Timothy Stansbury, Jr. on an apartment building roof in Bedford-Stuyvesant on January 24, 2004, Kelly did not reflexively support the officer, but rather made sure to convey to the public that the incident would be investigated fairly. The officer was prosecuted, but he persuaded a grand jury that the shooting was accidental, as it probably was. Nearly three years later, on December 28, 2006, Kelly announced the consequences of the NYPD’s internal investigation: The officer would be suspended for 30 days without pay, permanently stripped of his gun, reassigned to a property clerk’s office, and could be fired during the next year for any infraction.

Is the New York success story transferable to other cities? Probably not, says Zimring, who points to New York City’s extraordinarily high
1990 crime rate (on the theory that what has gone up will eventually come down), combined with its neighborhood density and low rate of
civilian handgun ownership. This is a reasonable inference, but it isn’t entirely convincing; most criminologists have argued that density
(coupled with poverty and race) increases crime–in other words, those conditions are hardly unique to New York.

Perhaps a better explanation–though not necessarily more helpful for those looking to replicate its success–is that New York from 1994 to 1996 and post–September 11 boasted two innovative police leaders, with managerial vision and skills that cannot be captured by regression analysis. Contemporary New York policing employs the Compstat principles of management developed by Bratton, which has evolved under Kelly in the post–September 11 environment. Violent crime is down, recruitment of minorities and women is up, and public opinion polls have shown that most New Yorkers appreciate the city’s safety.

There is at least circumstantial evidence that other cities can replicate New York’s experience. A natural experiment has occurred in Los Angeles, where Bratton operates under a federal court consent decree. He has imported many of his street-tested strategies from New York. And while a number of American cities that led the great crime decline–like Boston and San Francisco–are experiencing “a gathering storm” of rising violent crime, violent crime in Los Angeles has declined by 29 percent between 2004 and 2006, although its murder rate remains double that of New York.

Another natural experiment began on September 6, 2006, when Newark Mayor Cory Booker, who is black in a majority black city, appointed the NYPD’s deputy commissioner for operations, Garry McCarthy, who is white, to lead his police department. McCarthy was a highly regarded figure in the NYPD’s command staff and had run its Compstat meetings for the past seven years. Newark, whose murder rate is more than three times that of New York City, will surely offer Booker and McCarthy a challenge. So will the police department: Its numbers have declined to 1,300 from 1,475 in 2001, while its starting salary–$26,561–is among the lowest in New Jersey. While good management is necessary to reduce crime, the Newark experiment will help us understand whether it is sufficient.

We may never know precisely what caused the great American crime decline and why it is ending in most, though not all, cities across the
nation. But Zimring’s book provides powerful evidence that traditional thinking about crime rises and declines may be outmoded, while the most
conventional but least appreciated reason–good management producing good policing–may presently provide the best answer.