In 1990, more people were murdered in New York City than ever before. In fact, by late November of 1990, every single murder set another new record. Nor is New York alone. The Centers for Disease Control recently announced that, nationwide, the risk of death by homicide for Afro-American males is approaching the death rate of combat soldiers in Vietnam.

Increases in crime are nothing new. As far back as 1965, crime was so bad that President Lyndon Johnson felt obliged to create a national crime commission. We did not know then just how bad it was going to get.

Then, police believed they had the answer to climbing crime rates. They promised that if city governments only gave them enough officers, cars, high-tech equipment, and a new universal “911” response system, the police would cut crime.

Policymakers and politicians in city after city believed them. More police were put into more patrol cars, to respond more rapidly to calls, and to patrol high-crime areas. Police did arrest a lot of people, some of them many times. Alas, they did not control crime. Twenty-five years later, this strategy is in shambles.

Although police have no control over many factors that contribute to the rise in crime, we should not underestimate the extent of police failure in recent decades. A good deal of research suggests that the very heart of the strategy—police riding around in cars and racing to calls for help—had negligible effects on serious crime.

At the same time crime was rising, so was citizen fear of crime. This is unsurprising, but not unimportant. For understanding the causes and effects of fear goes a long way towards explaining the police failures of recent decades.

To begin with, fear of crime is itself one of the worst consequences of crime. For every person mugged in the park there are hundreds or thousands who do not walk in the park. For every one mugged or robbed in the neighborhood, there are hundreds who triple-bolt their doors and abandon the streets, many others who sell out or do not move in. There are good teachers who, learning of the neighborhood’s reputation, will teach elsewhere instead.

Moreover, social scientists have known for some time that the fear of crime is not caused solely by “real crime”: murder, rape, assault, robbery. As important is “disorder,” the physical and social evidence that a neighborhood is not quite under control: youths hanging out on the corner; panhandlers, hustlers, and suggestively dressed prostitutes on the street; public drunkenness and rowdiness.

For some years James Q. Wilson and I, joined by a growing chorus of other social scientists, have argued that this fear of disorder is justified, for disorder leads to crime. Just as an abandoned house with one broken window-a sign that no one cares-will soon have all its windows broken, a neighborhood which permits disorderly behavior will soon find itself plagued by serious crime as well. Disorder drives citizens from the streets, leaving them unsurveilled; it emboldens punks to graduate to crime and criminals to try out new turf.

Until recently, however, those of us who believed that disorder matters, and that police ought to spend more time maintaining the orderliness of neighborhoods, and less time in traditional crime-fighting, did not have much more than common sense and anecdotal experience on our side. Now we have Northwestern University Professor Wesley G. Skogan.

Professor Skogan’s book, Disorder and Decline: Crime and the Spiral of Decay in American Neighborhoods, tests the “broken window” theory and examines the relationships between disorder, crime, fear, and neighborhood life. Skogan’s data confirm that disorder accompanies and probably breeds serious crime.

Professor Skogan’s analysis is based largely on surveys and field work conducted in 40 neighborhoods in six American cities. Citizens reported on their neighborhoods, their fears, the actions they have taken to defend themselves, and whether or not, or how often, they have been victims of crimes. Trained observers noted signs of neighborhood disorder: graffiti, poorly kept properties, vandalism, youth gangs, drug dealing, prostitution, drunkenness, and other such conditions.

Several important points emerged when Professor Skogan looked at the data. First, despite fretting by social scientists and policymakers that one person’s sense of disorder might be another’s entertainment, that different races, ethnic groups, and classes might have vastly different definitions. Dr. Skogan’s research of disorder, revealed a remarkable consensus among neighborhood residents, regardless of back ground. People know what offends them, they are scared by it, they believe and fear it will lead to serious crime, and they act on those fears. If they sense that things are out of control, and they cannot move away, they restrict their activities.

Professor Skogan notes that previous researchers have shown that in disorderly neighborhoods, neighbors fear and distrust each other and are far less likely to believe that their neighbors would call the police in the event of a crime. Other studies have shown that the fear of crime (which may be caused by disorder) reduces “people’s willingness to take positive action when they see crimes being committed-many even balk at calling the police.”

Skogan’s exhaustive study showed similar results. People were asked whether neighborhood residents tended to “help one another” or “go their own way.” In disorderly neighborhoods, residents were much more likely to say that “people go their own way.” People were asked when they had last asked a neighbor to watch their home while they were away. In disorderly neighborhoods people were less likely to engage in such cooperative informal crime prevention.” More surprisingly, even such highly individual anticrime measures as marking an identification number on one’s household goods seemed to be less common in disorderly neighborhoods.

As Skogan sums up this portion of his findings:

Disorder may undermine in several ways the capacity of communities to preserve the conditions they value. Disorder may foster suspicion and distrust, undermine popular faith and commitment to the area, and discourage public and collective activities. Disorder may also undermine individual morale and the perceived efficacy of taking any possible action. Since there is little that individuals seem able to do about many forms of disorder, they may feel disheartened and frustrated, rather than motivated to do more, even to protect themselves.

Skogan also found overwhelming evidence that people who live in disorderly neighborhoods believe they are surrounded by crime. In fact, disorder was the most important factor in persuading people that they lived in a high-crime neighborhood. Fear is also linked to disorder. Residents were asked how safe they “would feel out alone in your neighborhood at night.” Residents of disorderly neighborhoods were significantly more likely to say they would not feel safe.

Most importantly, he found that people who live amongst disorder are right to be afraid. More than any other causal factor, disorder either precedes or coexists with serious crime. His analysis of neighborhood robbery rates showed that people were much more likely to be robbed in disorderly neighborhoods (see Figure 1). Skogan’s conclusions:

The evidence suggests that poverty, instability, and the racial composition of neighborhoods are strongly linked to area crime, but a substantial portion of that linkage is through disorder.... [Direct] action against disorder could have substantial payoffs.... These data support the proposition that disorder needs to be taken seriously in research on neighborhood crime, and that both directly and through crime it plays an important role in neighborhood decline. “Broken windows” do need to be repaired quickly.

Those of us who have long believed that the key to fighting crime and fear is to fight disorder have argued that police forces should shift their resources out of high-tech, quick-response anticrime tactics and into something that has come to be called “community policing.” Community policing has gotten a lot of press play in New York recently, because Police Commissioner Lee Brown, who as Houston’s police chief ran an important community policing experimental program, is a firm believer in the idea and is trying to implement it in New York. Community policing consists essentially in using the police to help make communities more crime-resistant rather than using police primarily to catch criminals. In community policing, police engage in “problem solving.” Because community policing is aimed at fear and disorder as much as at crime, the problems to be solved are largely defined by the community: People who live in the community know what they are afraid of and what constitutes disorder. Under community policing, police spend more of their time enforcing community standards. They fight disorder, thereby shoring up what Northwestern University political scientists Dan Lewis and Greta Salem have called the neighborhood’s “moral reliability.”

Moral reliability is what makes us assume that people will behave properly while in public. Moral unreliability, on the other hand, threatens urban life. When strangers swear at people in the park, intimidate shoppers, or threaten those who won’t “spare some change,” the sense of civilized community is lost. Once a certain threshold is crossed, citizens begin to walk the other side of the street, shop in sequestered malls, abandon public transportation and parks, and move from the cities. They become afraid to stand up for their fellow citizens. Moral unreliability is contagious. It encourages others to misbehave, attracts troublemakers, intimidates citizens (especially the weak and elderly), erodes social bonds, and finally, erupts into serious crime.

If disorder both accompanies crime and causes it, this has enormous implications for the police failures of the last several decades. To police officers speeding around in their cruisers, the drunks and panhandlers must have seemed very small stuff. To their superiors, obsessed with “efficiency” and the idea that the way to fight crime is to arrest people who have committed crimes, we know that taming disorder seemed insignificant. Like American strategists in Vietnam, they assumed that if the enemy body counts got high enough the payoff would have to come eventually. Under that assumption, the distinction between the efficiency of the system and its effectiveness disappears.

For police engaged in community policing, efforts to shore up “moral reliability” or “solve problems” can be as simple as spending more time enforcing laws against public drunkenness or writing citations against liquor stores that cater to public drunkenness. It can be as complex as coordinating the efforts of community groups so as to find something better to do for the youths hanging out on the corner than frightening their elderly neighbors.

Because the community must help define problems, and because community policing aims at reassuring citizens that they may use the streets, it always involves a great deal more contact between police and citizens than the 911 model, under which, for most citizens, cops remain figures behind a speeding windshield. Cops in community-policing programs spend more time in their base areas and less time responding to calls outside the area. They spend more time walking the beat and less time in cars. They get to know people in the area by ringing doorbells and asking residents about their concerns. They usually set up highly visible storefront substations where citizens can walk in with complaints, problems, or suggestions. They attend community meetings and even assist in organizing neighborhood blockwatch and other self-defense programs.

Dr. Skogan reviews the evidence on the two most well-known and extensive community policing experiments of the 1980s, the one run by Lee Brown in Houston and another in Newark. He cautiously concludes that they worked: Community policing is well received by neighborhoods. It reduces both disorder and citizen fear of crime. It therefore may well reduce crime itself, and certainly contains its bad effects.

The Houston experiment employed different elements of community policing in different experimental areas. In one area, the police set up a very visible walk-in storefront station and used that as a base from which to work with the community. Within a few months 65 percent of the people in the community knew about the station, and many people did walk in to report problems. The police also organized community meetings outside the station, distributed newsletters, and made other efforts to get to know people. The officers were given a great deal of latitude by their superiors to launch programs and respond to community problems. It seemed to pay off:

Station officers heard repeated complaints that the park had been taken over by rowdy youths; nearby residents were reluctant to use it. The police began patrolling the park regularly, and during the summer months they organized athletic events there. Residents returned to the area, and a vending machine . . . [which] had been removed after being vandalized repeatedly, was reinstalled at the park’s swimming pool.

In a second neighborhood, police created a community self-help organization to assist in fighting crime; in another, they visited homes, giving residents a chance to discuss their problems and to establish relationships with particular officers, who always left personal business cards.

All three programs were then measured for their effect on social and physical disorder, community fear of crime, satisfaction with the area as a place to live, and satisfaction with police performance. All the experimental neighborhoods showed improvement in every category, though for some categories in some neighborhoods the improvement was statistically insignificant.

The Newark experiment, which was much more extensive, used most of the strategies tried in Houston, except that they were used in combination. Here the results were even more impressive: In every one of the same five categories, neighborhoods got better in the course of the program. In every case the results were strong and significant. It is not yet clear whether these experiments reduced serious crime. But the reduction of fear itself is an enormous accomplishment, since fear may be the worst effect of crime. And of course the reductions in disorder suggest that crime will be curbed over time.

In their book, Beyond 911: A New Era for Policing, Harvard University scholars Malcolm K. Sparrow, Mark H. Moore, and David M. Kennedy assume that maintaining order and fighting fear are essentials of good police work. Sparrow and company concentrate on how to change the culture of police departments so that they can pursue these goals successfully and adapt to concepts such as community policing.

They accept as a starting point the historical failure of police to control crime, particularly under the 911 or “reform” model. The traditional model of managing police, modeled after the hierarchic bureaucracy of the military, was to control officers and limit their discretion through layers and layers of command. The 911 model tightened the system even further by keeping police officers in their cars and under centralized control. They were in constant radio contact with headquarters but cut off from the community.

Community policing, of course, tries to reinsert officers into the community. But that attempt raises very real questions: How will officers now be controlled? How do police officials ensure that the old days of corruption, police favoritism, and “rough curbstone justice” do not recur? How are police officers to be creative and innovative in problem solving while still fitting into a paramilitary bureaucracy? To whom will the officer be accountable and how will a department measure an officer’s effectiveness once his main job ceases to be responding to logged calls for service? And if patrol officers are the key to successful community policing, how do we keep them on patrol rather than bucking for promotion off the beat?

Drawing on the examples of six innovative departments and their chiefs (again including Lee Brown and his Houston force, see box on page 63), as well as the lessons of companies in the private sector, Sparrow, Moore, and Kennedy try to answer their questions. Their many answers really come down to one answer, so bold that they cannot quite say it out loud: The paramilitary, bureaucratic model of police organization must be substantially abandoned.

Policing is a dangerous job, requiring brave men and great esprit de corps. For that reason police officers will always need to draw on the military virtues. But the bureaucratic model, by definition, crushes innovation and flexibility. Community policing requires innovation and flexibility: not good and obedient soldiers, but trusted, self-directing professionals.

How do we get there from here? Well, if Commissioner Brown has his way, over the next few years New Yorkers will have a front-row seat on the most ambitious attempt yet to do so. The Beyond 911 authors have some concrete and persuasive suggestions. First, it will be impractical for officers in the field to tailor their every action to the rulebook. Under community policing, officers must fully grasp for themselves the goals and values behind the rules. Police executives must clearly state their vision of policing and the values that should guide police practice, and make sure officers, as fellow professionals, understand and share those principles.

Since success in the field will be measured largely by the community’s reaction to the officer’s work, the department must work to establish strong, trusting relationships with community institutions, such as schools, churches, and neighborhood councils. Chiefs must rethink the role of middle managers—sergeants, lieutenants, and captains—so that rather than being guardians of the status quo, they become the advocates of innovation, reward creativity, and tolerate risk-taking and failure.

Perhaps most importantly, not only the role but the status of patrol officers must change. For years patrol officers have been at the bottom of the police hierarchy. Promotion meant either a change in rank and a move off the beat or a move to the detective squad or some other specialized unit, which also meant moving off the beat. Those at the bottom of the ranks, those closest to the street, may have the best sense of how to solve a problem, or even close a case. But the system’s rigid respect for the protocol of rank and specialty has frequently squandered the knowledge and insight of regular officers, as when detectives move onto a crime scene and brush off the contributions of the original reporting officers.

This, say the authors of Beyond 911 must change. Good patrol officers who demonstrate their ability to solve local problems and even local crimes must be rewarded, not only financially, but with official status and the respect of the organization. They suggest a number of ways to do this, from creating a new rank of “master patrol officer,” to loosening the ties between rank and pay, to allowing patrol officers to run the investigating teams for crimes occurring on their beat, with the detectives called in to help patrol at patrol’s discretion.

Policing has turned a corner. Professor Skogan’s research confirms an intuitively reasonable idea: Untended disorder leads to serious crime. New Yorkers, beleaguered by disorder and crime, will read these volumes with special interest.

Neighborhoods will only be safe if we can restore their moral reliability. Police can help. In really tough areas, they must help. Their presence alone, on foot, will reassure citizens that when something untoward happens police will be there to restore order. When obstreperous youths, vagrants and other disorderly persons behave unpredictably, police demonstrations will rein them in, thereby reassuring citizens.

Ultimately, however, safe neighborhoods are the responsibility of citizens. Being morally reliable certainly means that citizens will refrain from uncivil and threatening behavior. But it means more than that. Moral reliability obligates citizens to do good things predictably: control their own and other neighborhood children, rebuke citizens who violate basic rules of civility, come to the assistance of neighbors and other citizens who are threatened, accept responsibility for vigilance in their own neighborhoods, and strengthen the basic character-building institutions of society: family, neighborhood, school, church, and commerce. Public civility has been undermined not only by the spread of moral unreliability from obstreperous person to obstreperous person, but also by the failure of responsible persons to do the right thing. It takes courage to be a good citizen, but there is no alternative.


Brown is black; the department and the city’s powerful were white. He was an outsider in a department that valued its own. He came under the unfamiliar banner of neighborhood policing into a department that had long prided itself on taking names and kicking ass. Despite all this he reformed the Houston Police Department thoroughly and fundamentally.

Before Brown’s arrival, Houston’s police force prided itself on its own brand of tough, shadowy, no-quarter policing, emphasizing cruises through high-crime areas and lots of felony arrests. Old-timers on the force today still boast that in the good old days a burglar caught in a house was likely to die there.

Brown arrived as a firm believer in community policing, but he was convinced that community policing requires a complete transformation of policing and police departments. His job was to change how the Houston department thought, how it worked, what it believed, what it valued.

Brown knew where he wanted to go, but he did not presume that he knew how to get there. He saw his programs as opportunities for his department to team, and taught his officers to regard failure as an informative detour. The department, consequently, is in a ferment of innovation. Brown has been extraordinarily successful in blunting conventional policing’s fear of and resistance to change.

Brown thought a new deal could be struck with the citizens of Houston. Instead of a department that was indulged so that it would do the dirty work of protecting society, Brown would settle for no less than a department trusted, and supported, to do right. Instead of a department cut off from local communities, Brown planned a department tied to the City’s residents and neighborhoods. Instead of a department that did what and how it pleased, Brown imagined a department that focused on the problems defined by communities.

Brown summarized these democratic values in an official statement meant to guide the department’s daily operations. He was convinced that rules and procedures could not deal adequately with the varied circumstances officers confront. Nor can rules guarantee proper conduct, because so much of an officer’s work occurs beyond effective supervision. For Brown, teaching values was not only an instrument of reform but the only effective way of controlling officers’ conduct.

That Brown, starting from where he did, could make the credo of democratic values stick in Houston, and marry it to a program of ambitious experiments in neighborhood policing and fear reduction, is a sign of considerable promise for policing everywhere. Brown’s experience suggests that not even the most entrenched police departments need be immune to growth.

—adapted from Beyond 911

Books cited in this essay:

  • Wesley G. Skogan, Disorder and Decline: Crime and the Spiral of Decay in American Neighborhoods. New York: The Free Press, 1990.
  • Malcolm K. Sparrow, Mark H. Moore, and David M. Kennedy, Beyond 911: A New Era for Policing. New York: Basic Books, 1990.

City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next