Twenty-first-century policymakers profess a deep commitment to “public health,” and they now apply the concept to the problem of violent crime in cities. When Joe Biden took office as president in 2021, for example, he promoted the adoption of public-health interventions to address an uptick of murders; last June, U.S. Surgeon General Vivek Murthy declared gun violence a public-health crisis.
These initiatives may be well-intentioned, but they ignore a hard truth. Police departments play a critical role in controlling violent crime and should be part of the public-health approach. Advocates of such an approach to crime tend to be reluctant to engage the police, worried about racially disparate policing and incarceration statistics. Yet the wholesale avoidance of the police works against their goals.
Finally, a reason to check your email.
Sign up for our free newsletter today.
Public-health advocates regularly suggest preventing violence by, for instance, securing abandoned properties, cleaning up blighted lots, and fixing street lighting. But improving the physical environment is an old and well-established idea in policing. More than 40 years ago, James Q. Wilson and George L. Kelling called on police to help make neighborhoods safe by reversing the underlying disorderly conditions that drive crime. In their landmark 1982 Atlantic article introducing the Broken Windows thesis, Kelling and Wilson anticipated today’s trends. “Just as physicians now recognize the importance of fostering health rather than simply treating illness,” they wrote, “so the police—and the rest of us—ought to recognize the importance of maintaining, intact, communities without broken windows.” Broken Windows policing was the original public-health strategy for criminal justice, and it offers important lessons for America’s crime situation today.
Now is indeed the time to revisit those lessons. The violent-crime surge of 2020 caused increased citizen fear. Four in ten Americans recently said that they felt afraid to walk alone at night within a mile of their home—the highest share in more than three decades. Fortunately, violence is now falling again across many U.S. cities. But proactive, quality-of-life policing seems to deserve some credit for that turnaround.
Broken Windows policing has a complicated legacy, partly because police departments have sometimes misinterpreted or incorrectly applied it and partly because critics have maligned it as a racist, punitive, and ineffective approach to public safety. Those critiques are unfair. Enforcing the law against disorder, if done correctly, can control crime without posing significant unintended consequences. In fact, Broken Windows is consistent with contemporary public-health efforts to reduce violence.
In their 1982 article, Kelling and Wilson offered a compelling metaphor on disorder and crime, grounded in social psychology and practical policing experience. “If a window in a building is broken and is left unrepaired,” they wrote, “all the rest of the windows will soon be broken.” Broken Windows theory posited that rising social disorder (loitering, public drinking, vagrancy, prostitution) and physical disorder (graffiti, trash, vacant lots, abandoned buildings) generate public fear. This fear drives stable families to flee, leaving remaining residents isolated and avoiding public spaces. As neighborhood anonymity rises, law-abiding residents lose control over public spaces, inviting serious crime as offenders sense the community’s neglect. Wilson and Kelling saw crime as an epidemic, in other words, spreading through neighborhoods via environmental features.
What does the evidence say? Research generally supports strong links between disorder and subsequent fear. However, the evidence on Wilson and Kelling’s disorder/crime developmental sequence is mixed. Studies consistently show strong correlations between social turmoil and crime in varied settings, but these often fall short of causation. In practice, of course, it doesn’t matter whether disorder is a cause or correlate of more serious crimes. Police attention to troubling underlying conditions can be helpful in both scenarios. More generally, neighborhood vitality suffers when low-level harms accumulate.
A sanguine view of the existing research is that adverse conditions may cause more serious crime in specific contexts. A trash-filled vacant lot, say, may serve as a drug-sales hot spot for a violent street gang in an otherwise safe neighborhood, or a park may host repeated fights among intoxicated teenagers in an area with low capacity to exert control over public spaces.
Kelling and Wilson’s original article didn’t provide a detailed crime-reduction road map. Instead, they offered general principles, noting that communities differ in their capacity to address disorder and suggesting that “the essence of the police role in maintaining order is to reinforce the informal control mechanisms of the community itself.” They also recommended safeguards to prevent police from becoming “agents of neighborhood bigotry,” emphasizing a balance between community interests and individual rights.
Wilson and Kelling would later describe Broken Windows policing in more precise terms. A follow-up 1989 Atlantic article stressed the importance of community engagement, as police negotiate rules of appropriate behavior in public spaces. The authors also explicitly linked community policing and problem-oriented policing programs. In Fixing Broken Windows, a 1996 book, Kelling and Catherine Coles identified a range of policing actions as consistent with the original theory—from misdemeanor arrests of fare-beaters who jumped New York City subway turnstiles to community problem-solving collaboratives that cleaned public spaces in Boston and Baltimore. Broken Windows policing, then, represented police-led attempts to make neighborhoods safer by addressing underlying social and physical disorder.
Notably, Wilson and Kelling did not advocate “zero tolerance” or indiscriminate enforcement. Kelling observed that “metaphors are powerful” but lamented that “they are abstract and prone to misinterpretation.” Too many police departments seemed to misinterpret what he and Wilson had meant. Responding to critics who linked aggressive order-maintenance strategies with excessive stops and arrests of minorities, Kelling wrote: “What? We never said or intimated that.” Kelling and Wilson ultimately viewed their theory as making two key contributions: it encouraged the police to take public order seriously; and it provided a strategy to make police more effective in reducing crime.
Evidence is clear that Broken Windows policing works to prevent crime—again, when properly implemented. The data on the causal links between disorder and subsequent crime may be mixed, but evaluations of disorder-policing programs—efforts to control serious crime by addressing underlying disorder problems—show unambiguously positive results. A systematic review of 56 controlled evaluations, including 12 randomized experiments, found that such initiatives lowered crime by 26 percent in treated areas, relative to control areas, including significant reductions in violence, property crime, and drug offenses. Disorder policing, moreover, didn’t just displace offenders by moving them down the block; crime-control benefits that emerged in cleaned-up places tended to spread to surrounding areas.
How the police deal with unruly conditions matters greatly. Community and problem-oriented policing in specific crime “hot spots” worked to prevent crime, the review found. By contrast, indiscriminate and aggressive order-maintenance strategies, such as increased misdemeanor arrests and heightened pedestrian stops, applied across large police districts or neighborhood areas, were not effective. Other studies suggest that overly aggressive policing also risks generating unintended harms, from racial disparities to police misconduct.
“Arrests and summonses for disorderly behavior may be an appropriate response, but they should be made in consultation with local communities.”
These findings are consistent with Wilson and Kelling’s clarification that reducing disorder should be viewed as a tactic of community policing, not a universal panacea for crime. Community policing is a local exercise. It requires police departments and residents to define priorities together. It decentralizes decision-making authority to line-level officers. It promotes techniques to guide the design of appropriate crime-prevention strategies. Implemented thus, it aids citizen satisfaction, police legitimacy, and social perceptions—and can even guide officer decisions.
Problem-oriented policing yields more nuanced responses through the analysis of crime and disorder at specific places. As part of this process, police officers need to develop strong working relationships with local communities and know the key good—and bad—behavioral influences. Misdemeanor arrests and summonses for disorderly behavior may, of course, form part of an appropriate response to neighborhood conditions. But such decisions should be made in consultation with local communities, not imposed indiscriminately from afar.
Cast in this light, Wilson and Kelling’s Broken Windows approach matches the public-health trend in modern policymaking. Considered broadly as a field, public health deploys innovative interventions, such as food-inspection regimes, immunization programs, and water-quality improvement efforts to reduce bad outcomes. Public-health practitioners often say that they seek to identify all relevant “stakeholders” in order to undertake a comprehensive, multidisciplinary response. Public-health practitioners and researchers apply what can be called an “iterative” approach. Programs identified as effective are widely disseminated.
Cleaning up the physical environment has become one popular strategy. Greening vacant lots by removing trash and debris, mowing grass, and repurposing them for community use has been shown to reduce violence and improve community perceptions of safety. Improved lighting in public places has also been shown to enhance safety. Public health has a long history, too, of helping homeless people, providing treatment to addicts, and offering services and opportunities to at-risk youth. These activities obviously can complement police-led efforts by reducing physical and social disorder—in precisely the ways that Kelling and Wilson suggested.
Police and public-health agencies should be natural allies in reducing injuries and other harms to affected communities. But while some prominent public-health academics, such as Daniel W. Webster of Johns Hopkins, have recognized the importance of policing to public health, it is generally not considered as part of a public-health response to violence prevention. Police and public-health officials have collaborated before, however. The public-health establishment often notes its success in reducing alcohol-involved vehicle crashes, and police departments played a key role in campaigns against drunk driving. The CDC recommends “high-visibility saturation patrols” in specific places and at times that see high numbers of alcohol-related vehicle accidents and “well-publicized and high-visibility sobriety checkpoints,” where police stop vehicles to identify and arrest impaired drivers. The American Public Health Association (APHA) suggests a number of policing interventions to address alcohol-related harms, such as the interrogation of impaired drivers at the location where they had their last drink. It recommends launching not just investigations against potentially troublesome alcohol license-holders but even sting operations that send underage people into licensed establishments.
“Epidemiological analyses can facilitate a deeper understanding of crime that includes individual, situational, community, and societal risk factors. ”
Yet, on the most salient issue of all—violent crime—public-health violence-prevention efforts generally ignore or downplay the importance of police. This approach has ideological roots. The field’s reluctance, as noted, flows from concerns over racial disparities in enforcement, as well as from ongoing difficulties with police violence and abuse. But charges about racial disparities permeate American society; it is commonplace, for instance, to hear about racial inequity in hospitals and health care generally. Public-health experts do not, however, exclude hospitals from violence-prevention efforts. Nor should they decline to collaborate with police departments.
Police departments have much to gain from public-health collaborations. Epidemiological analyses can facilitate a deeper understanding of crime that includes individual, situational, community, and societal risk factors. Mobilizing a wider range of partners can enrich crime-prevention responses and enhance police legitimacy in besieged communities. Measuring the impact of specific responses to violence and disseminating findings improves understanding of effective policing practices.
Making neighborhoods safer by addressing physical and social disorder seems to be an obvious starting point for police. The public-health perspective is clearly compatible with Broken Windows. Research confirms that disorder is best addressed through community-based approaches. Dealing with disorder requires the police to identify all participants—residents, public works agencies, social-services providers, prosecutors—relevant to cleaning up the physical environment, helping troubled people suffering from mental-health and substance issues, and incapacitating violent criminals. Developing these comprehensive relationships and maintaining the capacity to intervene are key activities of community policing.
One model emerged in Lowell, Massachusetts, where a randomized experiment tested the influence of problem-oriented policing on crime and disorder hot spots. The Lowell intervention produced a 20 percent reduction in crime calls to the police by reducing social and physical disorder in treated hot spots relative to control hot spots. Changes to the physical environment, such as cleaning vacant lots, securing abandoned buildings, and removing trash, generated the largest effects. Misdemeanor arrests for disorderly behavior and minor crimes yielded much smaller ones.
Or consider “greening” public spaces. In public-health circles, Philadelphia is viewed as a leader in reducing serious violence by greening vacant lots without direct police involvement. The Pennsylvania Horticultural Society keeps some 13,000 vacant lots clean and green through gardening and landscaping. These cleanup efforts may stimulate informal social control through community use of reclaimed urban spaces; studies tend to show that greener public spaces reduce crime. Yet green vacant lots did not prevent Philadelphia from experiencing a record 562 homicides in 2021 and similarly high 512 homicides in 2022. Given that Philadelphia had been greening vacant lots well before the Covid-19 pandemic gun-violence surge, it would be dubious to claim that this initiative produced the sudden 2023 homicide decline.
The lesson: green lots, on their own, can’t stop homicides in very violent places. Efforts to stimulate informal social control in active shooting hot spots—which, of course, tend to be located in disadvantaged neighborhoods—need to be supported by enhanced formal social control: namely, an increased police presence. Public health needs law enforcement.
People want safe, orderly environments and expect police to help reduce fear and victimization in public spaces. Disorder often fuels more serious crime, especially in disadvantaged neighborhoods where residents struggle to control social and physical behavior. Broken Windows policing can provide immediate crime reduction and foster lasting safety by empowering communities to assert control over their surroundings.
As a holistic strategy, Broken Windows policing should be seen as a public-health intervention: it modifies environments, addresses harmful behaviors, and protects vulnerable people. By integrating police into public-health violence-prevention efforts, we can strengthen both public safety and community well-being.
Top Photo by Kena Betancur/Getty Images
