More than 40 years have passed since the publication of one of the most important public-policy essays ever written. Its title, “Broken Windows,” captured the essence of a simple but deeply insightful idea: public order matters. “[I]f a window in a building is broken and is left unrepaired, all the rest of the windows will soon be broken,” wrote the late authors, political scientist James Q. Wilson and longtime Manhattan Institute senior fellow George L. Kelling, in the March 1982 issue of The Atlantic. Visible signs of chaos were like warnings: you’re not safe here. If left unaddressed, the chaos made those areas more vulnerable to further disorder, including serious crime. “ ‘[U]ntended’ behavior,” the authors maintained, “leads to the breakdown of community controls” and causes residents to “think that crime, especially violent crime, is on the rise, and . . . modify their behavior accordingly.” The areas where disorder festers become more “vulnerable to criminal invasion” than “places where people are confident they can regulate public behavior by informal controls.”

The theory—expanded on by Kelling and his wife, Catherine Coles, in their 1996 book, Fixing Broken Windows—sparked a revolution in American policing. At the direction of innovative officials like NYPD commissioner and later LAPD chief William “Bill” Bratton, and with crucial support from political leaders like New York mayor Rudy Giuliani, police departments across the country would, in the 1990s and 2000s, adopt tactics and strategies that reflected these vital insights. Proactive policing not only drove street crime down but also yielded unexpected benefits—like the illegal firearms discovered during pat-downs of turnstile jumpers in the subways and the outstanding arrest warrants discovered on the street through the enforcement of open-container violations. The historic, generation-long crime decline that resulted as Broken Windows policing took hold widely solidified legendary status for Kelling and Wilson.

Yet this law-enforcement revolution sparked acrimonious pushback from antipolice academics and activists—aided, in no small part, by how often the concept of Broken Windows policing was misinterpreted and distorted, much to the frustration of its originators. These distortions became more influential as crime continued its downward trajectory nationwide during the first decade of the twenty-first century, as large urban police departments focused on developing counterterrorism capabilities in a post-9/11 world and as a new generation of urban residents came of age with little or no awareness of recent history. Progressive critics argued for rolling back proactive policing measures and for lessening criminal-justice penalties; and a series of viral police use-of-force incidents, beginning in Ferguson, Missouri, in 2014, built momentum for these efforts, while intensifying hostility toward law enforcement. The death of George Floyd in Minneapolis in May 2020 served as the movement’s apex, triggering the deadliest urban riots in the United States since the 1960s amid widespread condemnation of police.

Perhaps not coincidentally, 2020 marked the largest one-year homicide spike in at least 100 years. Four years later, with crime—particularly gun violence—still well above pre-2020 levels in many U.S. cities, calls for American police to return to their mid-1990s crime-fighting approach have gotten louder. Unfortunately, this appeal, while entirely justified, cannot be practically pursued in the current environment. Two massive obstacles block the return of Broken Windows–style policing: the police workforce crisis; and the demonization of cops, and of policing itself, as racist. The kind of policing that led to one of the safest generations on record for American cities cannot be revived until these obstacles are surmounted.

In many U.S. cities today, proactive policing is a distant consideration. Police can often barely keep up with calls for service, maintain compliance with new, more burdensome reporting requirements, and, in some places, even staff shifts fully. According to the Police Executive Research Forum (PERF), resignations were up nearly 50 percent in 2022 relative to 2019, and retirements were up almost 20 percent; departments haven’t been able to hire enough officers to replace those leaving. Even if departments could do so, they’d be replacing veteran officers with less experienced ones. And as more departments reduce or modify hiring standards to boost recruitment, you can bet that new officers will not only be less seasoned but, in many cases, less competent.

What has the workforce crisis meant in practice? In Pittsburgh, police will no longer be responding to 911 calls that don’t involve emergencies in progress between 3:00 am and 7:00 am. In Houston, police have “suspended” more than 250,000 cases because of a lack of personnel. In Baltimore, a staffing shortage of more than 400 officers meant that on March 19 of this year, the city’s Southern District (home to some 61,000 residents) had just three officers on patrol for an entire shift, according to the city’s local Fox affiliate. Baltimore is not alone on this front. A police source in Chicago sent me snapshots of assignment schedules for some city districts several months ago. On July 20 of last year, the 1530–0200 shift in the fifth district should have had cars assigned to eight beats. Only one car was on the street for that shift. The 1730–0400 shift had cars assigned to only four of the eight beats. On June 2, the 2100–0730 shift should have had cars assigned to ten beats. It could staff only five. For the 1530–0200 shift, just two beats had a car assigned to them.

In New York, the staffing issues that have taken the NYPD from a force of more than 40,000 at the turn of the century to about 33,500 today have also driven up response times for many kinds of service calls. As I reported in the New York Post, in October 2022, the response times for critical, serious, and noncritical calls had risen 38 percent, 43 percent, and 52 percent, respectively, since October 2019. It now takes nearly ten minutes before an officer responds to a critical call for service, and more than 27 minutes to respond to a noncritical call. It doesn’t take much imagination to consider the effect on citizen willingness to call in reports on less urgent matters.

In addition to capacity problems, police face a morale crisis that has contributed to recruitment and retention issues and affected how officers engage with their environments when on patrol. In another PERF survey, many police executives tied their recruitment and retention problems to officer morale. The survey’s open-response section featured remarks such as:

“Officers are depressed over the negative national narrative about the police.”

“There is considerable concern by my officers over the future of policing. Their significant others are pressuring them to leave the profession.”

“The negative climate surrounding law enforcement has definitely affected morale.”

“Nearly every person in the department has had to change the way they conduct police work. The greatest casualty of this has been proactive work on the patrol level.”

The impact of this negative environment on police practice can be seen in sharp declines in arrests, which, between 2009 and 2019, fell by 25 percent nationally. In many cities, the decline is especially pronounced for the quality-of-life offenses—fare evasion, lewd behavior, vandalism, loitering, public drinking and intoxication—that Broken Windows told us warranted police attention. A recent report from the American Enterprise Institute found that, between 2013 and 2022, arrests for such offenses fell 74 percent in New York City, 77 percent in Washington, D.C., and 81 percent in Los Angeles.

So while it’s worth debating whether theories like Broken Windows and tactics like Stop, Question, and Frisk can help get crime back under control, what must be tackled first is the recruitment, retention, and morale crises facing police departments. Proactive policing requires manpower; and right now, too many departments are thin on the ground.

Addressing that manpower crisis, in turn, must begin with an effort to understand what lies at its root: the widespread demonization of policing as an institution in the wake of incidents like the killing of Michael Brown in Ferguson and of George Floyd in Minneapolis, as well as radical shifts in policy that have significantly raised the costs of enforcing the law and lowered the costs of breaking it. Both the demonization campaign and the policy agenda that reflects and reinforces it are outgrowths of a broader narrative about criminal justice in the United States. This narrative rests on three key claims, each fundamentally mistaken. These claims are:

The U.S. has a “mass incarceration” problem.

The U.S. has a “police violence” problem.

Because black males are overrepresented among the incarcerated and those subjected to deadly physical force used by police, the criminal-justice system is “systemically racist.”

As I have long argued, these claims do not withstand critical scrutiny.

Start with incarceration. The phrase “mass incarceration” argues, by its own terms, that we overincarcerate—unnecessarily imprison—so severely that we could decarcerate, en masse, without endangering public safety. But even a cursory review of what the data say about who’s in prison, why they’re locked up, and what tends to happen when they get out makes the dangers of a mass-decarceration program starkly apparent.

Year-end data for 2021 reveal that 62.4 percent of state prison inmates were incarcerated primarily for a violent felony. Another 3.7 percent were in on a weapons charge. Only 12.6 percent were in primarily for a drug charge. The fact that an inmate may currently be serving time for a nonviolent drug or property offense does not mean that he is not violent; inmates tend to have extensive criminal histories. Indeed, the average inmate entering a state prison has more than 11 prior arrests and more than five prior convictions. The same is true of those leaving a state prison, who, on average, have 12 prior arrests and nearly six prior convictions. Though most prisoners are chronic offenders serving time for serious violent or weapons offenses, the median time served by state prisoners as of 2018 was just 16 months, and just 2.4 years for those in on violent felonies. So much for the charge that the U.S. is a punitive society.

In case it isn’t clear what sort of outcome a mass-decarceration program invites: 80 percent to 83 percent of those released from state prison in a given year will be rearrested—and not just once but, on average, five times—over the following decade. The arrests are almost certainly a partial snapshot of a much larger picture of criminal activity that these individuals have engaged in, given that most crime goes unreported and most reported crime goes unsolved. But even these statistics don’t fully capture the harm invited by those pushing initiatives to release the more violent and chronic offenders who would need to go free to bring the incarceration rate down to where activists want it. The fuller picture is far grislier than numbers on a page can convey.

For a recent, high-profile example, consider the story of Sheldon Johnson, covered by the New York Times in early March. Johnson, a self-described shot-caller with the Bloods street gang, was released from prison in 2023 after serving more than 20 years for attempted murder, among other charges. After his release, Johnson became a prominent figure in the criminal-justice reform world and was held up as the picture of rehabilitation. Johnson was hired to work with at-risk youth by the Queens Defenders—a New York City–funded criminal-defense service provider—and even appeared on Joe Rogan’s popular podcast, where he discussed both his sordid past and alleged turnaround. Not weeks after the episode aired, Johnson was arrested and charged with murder and illegal weapons possession after police discovered him inside an apartment, where a later search revealed—per the Times—that the resident of the apartment had been decapitated and dismembered. While notable in some ways, this story is not so surprising.

In their seminal 1982 article, “Broken Windows,” George L. Kelling and James Q. Wilson captured a simple but deeply insightful idea: public order matters. (tenementcity/Alamy Stock Photo)

In cities across the country, data reveal that the typical homicide suspect is someone with an extensive criminal history. Mass decarceration isn’t merely bad policy. It’s deadly policy.

What about “police violence”? Here the advocates miss, or deliberately obscure, a great deal. We’re told that police violence, both fatal and not, is so ubiquitous as to be the defining characteristic of law enforcement. The data again refute this assessment.

Take the NYPD. In 2022, the department fielded 7.1 million calls for service. Officers made more than 189,000 arrests, including more than 4,600 gun arrests. Despite these numbers, NYPD officers intentionally discharged their firearms at just 46 individuals on just 40 separate occasions. In other words, assuming that these incidents occurred within the context of separate arrests, NYPD officers purposefully fired their weapons in just 0.02 percent of arrests. That year, the department reported only 8,270 uses of force, 93 percent of them “Level 1” incidents—those involving the lowest level of force. Even if we assume that force was used only within the context of making arrests, that still comes out to only 4.3 percent of those interactions. Other assessments have found even lower general use-of-force rates—including below 1 percent. And this is all without disaggregating justifiable force cases from unjustifiable ones. Doing so would reveal that the latter category (the one that reformers should really care about) is even rarer.

One thing that cannot be denied with respect to the available data is the existence of significant disparities between racial groups when it comes to various enforcement measures. Black men, in particular, are “overrepresented”—at least relative to their share of the population—among, for example, those arrested, imprisoned, and subjected to physical force by police. For too many in the policy community, however, the analysis stops here. Among criminal-justice reform advocates, these disparities constitute prima facie evidence of discrimination, so deeply woven into America’s criminal-justice system that dismantling its institutions is the only solution.

This is wrongheaded for many reasons. Ending the inquiry at top-line disparities in outcomes ignores substantial evidence suggesting that other, race-neutral, factors (such as rates of violent criminality and victimization) might explain why certain groups feature more prominently than others in, say, prison and jail populations. And controlling for those factors often explains most or all of the disparity cited as evidence of bias. Sometimes, the alleged disparity is even reversed, as demonstrated in economist Roland Fryer’s work on police shootings, which found that blacks in Houston were “27.4 percent less likely to be shot at by police relative to non-black, non-Hispanics,” despite being “4.35 times more likely to be involved in an officer involved shooting than non-blacks relative to their proportion in the 18–34 year old male population.”

The biggest hole in the systemic-racism charge is its singular focus on enforcement outcomes—as if these are the only important outputs of the criminal-justice system. Progressive critiques of policing and criminal punishment fail to grapple with how enforcement also produces benefits that are just as unequally distributed as its costs. Consider, for instance, the effect of the decline in homicides (a crime with massive social costs) between 1990 and 2014. Along with his colleague Michael Friedson, Princeton researcher Patrick Sharkey—a progressive critic of policing and incarceration himself—found that the homicide decline added a full year to the average black male’s expected life span while adding less than 0.2 years for white males. While the degree to which policing and incarceration helped produce that decline can be debated, it is well established that they played major roles; and subsequent research has shown that the homicide-reducing effects of hiring additional police officers, for example, are twice as large in per-capita terms for black victims as for white victims. This raises a question that few policing critics ask themselves: How can a system that produces such important benefits so disproportionately enjoyed by a particular group also be biased against that same group?

In the end, breathing life back into policing will require a successful campaign to refute the misleading and empirically dubious arguments made against the institutions of law enforcement and criminal justice. Shaping policy efforts in many state and local legislatures, as well as at the federal level, and underpinning the arguments made in the academy and throughout much of the mainstream press, these dubious ideas have made the political and legal environments in which police operate deeply adverse.

Slowing the momentum of the reformers and “abolitionists” and achieving a saner policy regime that, for example, places a meaningful limit on repeat offending through sentencing enhancements for habitual criminals is no small task. Nor will it be easy to attract, train, and retain the highly qualified, deeply motivated law-enforcement officers so many cities need. What’s necessary is a sustained, two-part effort. The first step is to think clearly about the issues at play in these debates. The second is to formulate strong, data-driven cases in favor of a more aggressive approach to policing and criminal justice—and against the kind of soft-on-crime approach that has taken hold around the country over the last decade.

A real victory for public safety will require courage and bold thinking. And it will require an intimate knowledge of the ugly realities in America’s most dangerous neighborhoods, as well as real empathy for those residing within them. In other words, it’s going to require us to tap into our own inner James Q. Wilsons and George L. Kellings.

Top Photo: Many police executives tie their recruitment and retention problems to officer morale. (Chris Moore/MediaPunch /IPX/AP Photo)


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