Why Patrick Dorismond didn't have to die.

The New Yorkers driven to the brink of riot last week by the shooting of Patrick Dorismond claim that Mayor Rudy Giuliani's zero-tolerance policy against crime has turned their city into a police state. Giuliani's defenders respond, in effect, that you have to take the bitter with the sweet. Yes, the shootings of Dorismond and Amadou Diallo are regrettable; but they are the inevitable side effect of the aggressive policing that has sent crime rates plummeting in New York and around the nation. You can almost hear the implied threat: Stop us from shooting the occasional innocent man, and criminals will rule the streets once more.

This is nonsense. But many white liberals secretly accept it, because they don't understand the difference between two radically different theories of policing that, thanks to Giuliani, are now frequently confused: "broken windows" and "zero tolerance." Broken-windows policing gives cops broad discretion to prosecute or not prosecute low-level offenses in order to deter the social disorder that breeds more serious crime. Zero tolerance focuses not on deterring crime but on discovering it--by mandating that police stop, frisk, and arrest vast numbers of young black and Hispanic men for minor offenses, in the hope that turnstile jumpers and pot smokers may also be guilty of more serious offenses. Broken windows has helped reduce crime in cities across the nation. Zero tolerance has undermined the popular support upon which effective crime-fighting ultimately relies.

To understand how broken-windows policing, which the mayor used with some success during his first term, morphed during his second term into its ill- conceived antithesis requires understanding a debate that long predates the Giuliani administration. Indeed, the question of how to constrain police discretion while empowering officers to investigate low-level public disorder has provided much of the drama in constitutional and criminal law over the past 50 years.

For most of American history, the police enjoyed free rein to enforce or not enforce vague vagrancy laws, which protected "health, safety, and morals." The theory behind the nineteenth-century vagrancy laws anticipated the broken-windows policy: "The vagrant has been very appropriately described as the chrysalis of every species of criminal," wrote the author of the definitive nineteenth-century treatise on police power. "If vagrancy could be successfully combated ... the infractions of the law would be reduced to a surprisingly small number." Accordingly, the 1887 police manual for New York City gave officers broad power to "disperse unlawful or dangerous assemblages, and assemblages which obstruct the free passage of public streets, sidewalks, parks and places." From the late 1880s until the 1950s, more than half the arrests in America's large cities were for drunkenness, disorderly conduct, vagrancy, and suspicious conduct.

This was not zero tolerance. Vagrancy laws were enforced selectively. As Robert Ellickson of Yale Law School has noted, until the '50s, cities like New York operated under the so-called skid-row system, which tolerated far greater levels of public disorder in certain areas than in others. Skid-row neighborhoods, "along with closely related Red Light Districts," Ellickson writes, "were areas where a city relaxed its ordinary standards of street civility." The police tolerated moderate public intoxication, panhandling, and bench-squatting in skid-row neighborhoods but not on Fifth Avenue. This informal zoning system allowed the police to quarantine public disorder in particular districts, like the Bowery, which became a kind of safety valve for the rest of the city.

Public-loitering laws came under constitutional and political fire in the 1960s for the same reason Giuliani's zero tolerance policy is under attack today: discriminatory enforcement. Because of the laws' vagueness, critics charged, police used them to single out vulnerable groups--especially racial minorities. Some African Americans also complained that, under vagrancy laws, police tolerated higher levels of criminal behavior in poor neighborhoods than in rich ones.

Moved by these concerns, the Supreme Court in the '60s and '70s began striking down public-disorder laws on the grounds that they were too vague. In the landmark 1972 Papachristou case, the Supreme Court invalidated a Florida vagrancy ordinance that allowed the police to punish "rogues and vagabonds," "lewd, wanton, and lascivious persons," "common gamblers," " thieves," and "habitual loafers." By creating an "all-inclusive and generalized" list of crimes, including strolling without a lawful purpose, Justice William O. Douglas concluded, the law essentially gave the police a blank check to arrest anyone they pleased. Without standards governing police discretion, Douglas warned, such an ordinance "permits and encourages an arbitrary and discriminatory enforcement of the law."

This constitutional revolution had a disastrous effect on law enforcement. As Williams Stuntz of the University of Virginia has noted, once it became harder for foot patrols to remove the disorderly, police retreated into their patrol cars--turning to a more reactive style of policing that increased the likelihood that officers' encounters with citizens would turn violent. Ironically, this transformation occurred at the very moment empirical research began to demonstrate that reactive strategies could not effectively combat crime. In 1982, James Q. Wilson and George Kelling published their celebrated article "Broken Windows" in The Atlantic Monthly, which argued that policing lower-level public disorder--loitering, drug use, gang activity, and public drinking--best diminished the fear and social disorder that allowed more serious crime to flourish.

In the '80s and '90s, responding to the new conventional wisdom Wilson and Kelling helped to create, city councils around the country passed new laws prohibiting specific acts of public disorder. By framing the new "quality of life" laws more narrowly than their predecessors, cities avoided the untrammeled discretion that had led courts to invalidate previous legislation. Cities such as New York and San Francisco prohibited loitering for the purposes of engaging in prostitution or selling drugs, or allowed the police to disperse panhandlers lingering within 30 feet of a cash machine. In New York, courts reviewed the public-order laws carefully, balancing the expressive interests of panhandlers and loiterers against the privacy interests of innocent citizens. They upheld laws prohibiting begging in the subway--an enclosed space where passengers may feel especially threatened or invaded--but struck down laws prohibiting begging on street corners. They struck down legislation prohibiting loitering in subways but upheld laws prohibiting loitering in front of schools.

This, too, was not zero tolerance. If the new laws prohibiting minor quality-of-life offenses--from jaywalking to drinking in public--had truly been enforced without exception, most New Yorkers would be in jail. The broken-windows approach instead urged cities to use quality-of-life offenses to increase police discretion, not to eliminate it. "'Zero tolerance' is a phrase I never use, never have used," Kelling declared at a recent conference sponsored by the School of Criminal Justice at the State University of New York at Albany. "It is antithetical to what I'm talking about." On the contrary, by allowing police to choose among a wide variety of legal and nonlegal sanctions for public disorders--from informal warnings to formal citations--the broken-windows policy viewed arrest as a last rather than as a first resort. As Kelling has written, "Neighborhood rules were to be enforced for the most part through non-arrest approaches ... so that arrest would only be resorted to when other approaches failed."

In this spirit, the Koch and Dinkins administrations in the '80s began to combine aggressive order maintenance with police discretion. They tried, for instance, to rezone the homeless. In a fascinating new book called Sidewalk, Mitchell Duneier, a sociologist at the University of Wisconsin at Madison, describes the decade-long effort to control the hundreds of homeless people who took up residence in and around New York's Pennsylvania Station after the decline of single-room-occupancy hotels in the early '80s. When Amtrak prohibited sleeping in the terminal, the homeless moved downtown to Sixth Avenue in Greenwich Village, where homeless alcoholics had long worked as street vendors. The police responded by cracking down on street vending. But a 1982 amendment to New York's vending law--proposed by a liberal city council member who wanted to protect a homeless poet--said the general prohibition on street vending didn't apply to those who used tables to sell newspapers, periodicals, or books. Accordingly, between 1985 and 1993, hundreds of homeless refugees from Penn Station reinvented themselves as vendors of recycled newspapers and used magazines. In 1993, the city council closed the loophole and imposed strict zoning restrictions on newspaper vendors. By reducing the number of spots available for homeless people to sell newspapers, the new law forced the homeless to squat to preserve their spaces. In other words, New York wasn't showing zero tolerance for public disorder; it was trying merely to contain it, much like the old skid-row system did.

New York applied the broken-windows approach in other ways as well. Working with Kelling, William Bratton, David Dinkins's transit-police chief, cracked down in 1990 on low-level disorder in the subways, such as turnstile-jumping. Subway felonies dropped 75 percent, and robberies dropped 64 percent. And, in his first years in office, Giuliani picked up where Dinkins had left off. When Giuliani made Bratton police commissioner in 1994, Bratton brought his approach to the entire force. Between 1990 and 1997, misdemeanor arrests increased by more than 80 percent. And the initial reviews were positive, even in minority communities. In a New York Times poll conducted in 1997, at the end of Giuliani's first term, 44 percent of African Americans said the NYPD was doing a good or excellent job.

But, late in Giuliani's first term, the broken-windows approach morphed into zero tolerance, and a crucial opportunity to win minority support evaporated. Although the precise moment of transformation is hard to pin down, it occurred on the watch of Bratton, whose 1998 book, Turnaround, proclaims allegiance to broken windows while in fact endorsing a less flexible approach.

As Bratton describes it, his views changed when he realized that the crackdown in the subway wasn't merely deterring crime by preventing the corrosive effects of low-level disorder. It also had the unintended effect of catching serious criminals. In the course of arresting turnstile jumpers, officers discovered many who were wanted for more dangerous offenses: one in seven fare beaters had arrest warrants outstanding, Bratton notes. "We hadn't thought of it," Bratton writes, "but it stands to reason that someone coming into the system with the intention to commit a crime is not likely to pay for the privilege." Eventually, the police began seeing the arrests of fare beaters as a tool of criminal investigation rather than an end in themselves.

The public event that crystallized the shift from broken windows to zero tolerance was the arrest in June 1996 of John Royster, who was accused of brutally beating three women near Central Park. The police caught Royster when fingerprints they had taken after arresting him for jumping a subway turnstile matched those found at the scene of the crime. "If you pay attention to minor crimes, clearly you end up with people involved in the major crimes," declared Howard Safir, Giuliani's second police commissioner, who took over in April of that year. This was a subtle but critical shift. Instead of prosecuting lower-level offenses to encourage an atmosphere of social order that would prevent more serious crime, Giuliani, Bratton, and Safir began prosecuting lower-level offenses in order to catch more serious criminals. Stopping and frisking numerous ordinary citizens, they reasoned, would make the people carrying illegal guns fear that their weapons would be discovered during an arrest for a more minor offense. And this would deter them from carrying guns in the first place.

It was this approach that led to undercover operations like Operation Condor, under which officers shot Patrick Dorismond in the course of approaching him to buy marijuana he didn't possess, and to the formation of the infamous Street Crimes Unit, which shot the unarmed Amadou Diallo. Under Operation Condor, narcotics officers volunteered to work overtime to arrest people for minor crimes, such as smoking marijuana and trespassing. Between 1999 and 2000, narcotics-division arrests for misdemeanors increased by 68 percent. As The New York Times noted last week, Operation Condor has driven drug-trafficking indoors, but it has had little impact on the homicide rate, which actually increased, or on the rate of narcotic-felony arrests, which decreased by nine percent. Instead, according to the Times, 75 percent of Operation Condor's arrests have been for misdemeanors or even more trivial crimes. In other words, the zero-tolerance thesis--that turnstile jumpers would turn out, under investigation, to be carrying illegal guns--proved, after a certain point, to be wrong: many pot smokers were guilty of nothing more than smoking pot. In an effort to keep its statistics up, the narcotics unit found itself arresting and prosecuting scores of low-level offenders. These were precisely the people who, under the broken-windows approach, might have been given a warning rather a handcuff. Likewise, the Street Crimes Unit, applying zero-tolerance policies, conducted 45,000 street searches between 1997 and 1998, leading to 10,000 arrests. Once the police began thinking of low-level public disorder not as a problem to be addressed but as an opportunity to investigate more serious crimes, the incentive to arrest citizens for relatively minor offenses dramatically increased.

Zero tolerance was made possible by, and took advantage of, changes in Fourth Amendment jurisprudence, which regulates searches and seizures. In the eighteenth and nineteenth centuries, the police could arrest citizens without a warrant only for serious crimes, such as felonies, or for misdemeanors committed in an officer's presence that threatened a "breach of the peace." But, in the '80s and '90s, the Supreme Court concluded that even harmless and nonviolent traffic offenses could justify a warrantless arrest. (The Court has never explicitly resolved whether the Constitution allows arrest for infractions so trivial that they are ordinarily punished by fine rather than imprisonment.) And, once a citizen is arrested, the police may conduct a full- blown search of his person and property. Moreover, the Court held in 1996 that the police may stop citizens on a pretext--using an arrest for a minor offense as an excuse for investigating a more serious one. Instead of counseling discretion in the face of this vast expansion of police license, zero tolerance recklessly encouraged the police to test its outer limits, greatly increasing the possibility of hostile confrontations and misunderstandings that, in the Diallo and Dorismond cases, proved fatal.

And zero tolerance has not only increased the risk of accidental violence; it has proved discriminatory at its core. For, in an age of limited resources and rampant criminalization, the promise of zero tolerance is, by definition, a lie. The police cannot possibly prosecute all minor offenders with equal force: New York's jails are not large enough to put all the pot smokers behind bars. In his more candid moments, Safir has acknowledged this. Zero tolerance is a "concept," not a rule, he told a 1999 seminar conducted by Columbia Law School Professor Debra Livingston, one of the leading scholars of order-maintenance policing. In Safir's opinion, zero tolerance of quality- of-life offenses in New York means "not total tolerance." But "not total tolerance" means the police must inevitably exercise discretion about where to focus their limited resources. And discretion raises the following objection: Why should pot smokers in the Bronx be arrested under zero tolerance, while pot smokers in the wealthy neighborhoods of Manhattan light up without fear of police interference?

By promising more than it can deliver, zero tolerance inevitably prompts charges of selective enforcement and discrimination. When Giuliani proposed last November that sleeping homeless people be arrested or issued summonses if they refuse an offer of shelter and a police officer's order to move along, the reaction was hyperbolic. Hillary Clinton said Giuliani's proposal violated the spirit of Christmas, which celebrates "the birth of a homeless child"; she also declared that "criminalizing the homeless with mass arrests for those whose only offense is that they have no home is wrong." In fact, Giuliani's proposal is legally uncontroversial. Most courts have held that cities can prohibit overnight sleeping in parks not designated for camping. (The last celebrated public-sleeping case in New York occurred in 1987, when the New York State Supreme Court sustained the involuntary hospitalization of Billy Boggs, a homeless woman who lived on a sidewalk in front of a Manhattan restaurant.) Instead of trying to rezone the homeless, like David Dinkins and Ed Koch, Giuliani threatened mass arrests of homeless people and then proceeded to arrest very few.

William Stuntz has argued that order-maintenance policing inevitably discriminates against minorities, because minorities are disproportionately poor and spend more of their time on the street. They have fewer private places into which they can retreat to conceal their minor crimes than the rich, and therefore they are more likely to be the targets of the police, who need far more evidence to investigate offenses committed in private than in public. One way of alleviating this discrimination, Stuntz suggests, would be to adopt a genuinely zero tolerance approach--targeted at the rich as well as at the poor--using search warrants and high-tech investigations to surprise rich white drug users in their Manhattan apartments and suburban country homes instead of focusing exclusively on open-air drug markets in Brooklyn, Harlem, and the Bronx. Such a strategy would provoke the same demonstrations on Park Avenue that it inspired last week on Flatbush Avenue--which is why it will never happen. Alternatively, if Giuliani were serious about zero tolerance, he could arrest Wall Street brokers who smoke pot in Battery Park and use the search as an excuse to look for evidence of Securities and Exchange Commission violations in their briefcases.

From the perspective of justice and equality, Stuntz's proposal makes perfect sense. But it wouldn't do much to reduce serious crime, which is what order maintenance was supposed to be about in the first place. A better solution would be to abandon the tendentious and misleading rhetoric of zero tolerance--except in neighborhoods where it enjoys political support. In a little-noticed but highly successful experiment early last year, the Thirty- third Precinct designated West 163rd Street in Washington Heights as a "model block." Before increasing their presence, the police held a community meeting explaining their strategy for getting rid of drug dealers; they then set up barricades on both ends of the block, checked the IDs of all passersby, and investigated even the most minor infractions. Once the benefits of the effort had become obvious--crime dropped by 20 percent--residents practically begged the police not to take the roadblocks down.

The Washington Heights experiment confirms what cities that have successfully applied the broken-windows strategy have long known: In order maintenance, the politics of law enforcement are just as important as the sociology of law enforcement. "Order-maintenance policing works for crime control, but it creates a huge need for cities to improve the quality of political participation," says Dan Kahan of Yale Law School. "It's a drug whose primary effect is that it will reduce crime, and its side effect is that it may exacerbate political tensions."

In Boston, the police department reached out to gang members and ministers from black churches to build community support for aggressive strategies to reduce gun violence. In Chicago, the city council passed an anti-gang loitering law, with key support from the city's black neighborhoods. And, after the Supreme Court struck it down in 1999 on the grounds that it gave the police too much discretion to arrest known gang members loitering " without an apparent purpose," the council passed a more focused version of the same law. "The only way to make order-maintenance policing a lasting presence is to make sure that the people affected have a role in the process, so they don't feel that they're being controlled by an occupying army," says Kahan. By refusing to address the political side effects of aggressive policing and willfully exacerbating them with his inflammatory rhetoric, Giuliani has sealed his own failure.

If Giuliani's zero-tolerance fiasco teaches anything, it is that the dream of removing prosecutorial discretion, which is to say politics, from law enforcement is the surest way of subverting the public support on which successful prosecutions rely. Like the backlash against Kenneth Starr's investigation of President Clinton--another failed effort based on zero tolerance of relatively minor offenses--last week's backlash against Giuliani reminds us that effective law enforcement officials must seek the political support of the communities they serve. By contemptuously ignoring this basic lesson, Giuliani has not only imperiled his hard-earned legacy as an effective fighter of crime. He has also raised questions about whether he is temperamentally suited to hold elective office.