# Opinion: Police in the U.S. have embraced a flawed model to do their job and a new approach could save lives

That soon changed when the Newark (N.J.) police department hired the bow-tie clad Jack Maple of New York Police Department fame. Maple was second-in-command to then NYPD Commissioner William Bratton, hired by then New York City Mayor Rudy Giuliani. The three were law enforcement darlings and media sensations for their crime reduction success in New York City, which was attributed to Compstat (performance management system) and a new type of policing called the broken windows theory.  

Developed by social scientists George Kelling and James Q. Wilson, the broken windows theory (BW) postulated that violent crime was connected to neighborhood disorder and decay. BW theory is the most well-known policing theory to date. 

What is less known is that BW theory, which Kelling and Wilson published in The Atlantic in 1982, was not the result of a peer-reviewed, evidence-based study. In truth, broken windows was an untested theory published in a periodical magazine. It had not been through the rigorous process of academic review required for anything claiming to be science. 

BW advocates credit the model for the drop in crime in New York City. Its critics call it a mass arrest program that disproportionately targets poor, largely minority communities, and exacerbates tension between police and communities subjected to BW policing and its progeny Zero Tolerance Policing and Stop, Question and Frisk.  

With the advent of BW, low-level arrests were no longer laughable; they were mandated through “zero-tolerance” edicts. Officers making low-level arrests were no longer zeroes, they were heroes, rewarded with commendations, city council resolutions and coveted positions. Officers were free to patrol with little direction other than to “bring in the numbers” — police parlance for productivity-like arrests.  

What is the theoretical foundation of BW? It stems from two research papers, the Newark Foot Patrol Experiment conducted in 1978 and 1979 (Police Foundation, 1981) and Phillip Zimbardo’s vandalism study in Stanford and Palo Alto (Calif.). The Newark Foot Patrol experiment found “The…intervention did not have a significant impact on overall crime, arrests, or victimization. However, the foot patrols (the intervention) increased the citizen’s perception of safety from those they feared: panhandlers, drunks, groups of youth and strangers (or signs of disorder).  

The Phillip Zimbardo vandalism study described how an abandoned car with a broken window invited others to scavenge the car. Wilson and Kelling don’t explain that the broken window we have been taking about for decades was broken by the researcher, Zimbardo. No one touched the abandoned car, and tired of waiting, he broke the window. 

The “scavengers” were not the “others” that citizens feared in the Newark Foot Patrol experiment, but white residents from the area, well-dressed white folks at that, according to Zimbardo. How this was generalized to poor folks in black neighborhoods isn’t explained. Zimbardo has since been widely criticized for his deeply controversial studies and findings. Extrapolating from this flimsy foundation, the BW authors inferred that addressing disorder would reduce crime, and the rest is history — and a dystopian history at that for marginalized communities. 

The NYPD implemented BW policing with gusto, arresting panhandlers, drunkards, groups of youth, homeless and legions of others, as they waged a war on indicators of disorder. Crime plummeted in New York City, but interestingly, the crime rate declined across the U.S., even in cities that did not embrace BW. We know BW was not an empirical study, so what does research evidence say about BW policing, generally? Are there blowback effects?  The research evidence is mixed, at best.  

Broken windows policing is broken

This biggest criticism of the broken windows model is that it is a mass arrest program fueled by zero-tolerance policing. Kelling himself waffled on what exactly BW is meant to be, saying: “Broken windows was never intended to be a high-arrest program.” Shortly before his death, Kelling wrote, “Community policing, rightly understood, can be, and often is, aggressive and even intrusive, depending on the community’s concerns.”

Yet for all his denials, Kelling used disorderly arrests as the proxy measure in analyzing the effectives of BW. Though often defended as a community policing program, options like referrals to homeless shelters, drug rehab programs, or citizen contacts were not considered as potential proxies, he chose arrests as what would measure BW in action.   

Kelling decided to evaluate BW and its impact on crime, in response to what he called, jealous “Ivory Tower elites” detractors. According to his own evaluation of BW, he estimated that for  every 28 disorderly arrests, one violent crime was averted, preventing 60,000 violent crimes over a 10-year period in New York City.

When one multiplies 28 arrest by 60,000 violent crimes, the result is 1,680,000 arrests (of mostly Black and and other people of color). Do the means (1,680,000 arrests) justify the ends; a theoretical reduction of 60,000 crimes, with the associated burden on Black people and other racial minorities subjected to broken windows policing? Kelling chose The Manhattan Institute, a conservative think tank, to publish his analysis instead of subjecting it to peer review — again. And he was the co-author of his own evaluation.  Would one expect him to reach any other conclusion?

So, what are the identified blowback effects from broken windows policing in New York City? In the 2013 Floyd v. City of New York decision, a U.S. District Court judge ruled that the spawn of broken windows, the city’s Stop, Question and Frisk (SQF) program, was unconstitutional. The NYPD is now operating under a consent decree and federal monitor.  In response to the NYPD’s argument that SQF removed guns from the street, the judge said, “the ends didn’t justify the means.”  

Of those frisked in 2011, a weapon was found just 1.9% of the time. What about racial profiling?  In 2011, “The NYPD made more stops of young Black men than there are young Black men in the city’s population;158,406 young Black men live in New York City and the NYPD made 168,126 stops.” What of Newark’s implementation of BW?  Its police department also operates under a federal consent decree.  

The Center for Evidence-Based Policing Policy (CEBCP) at George Mason University reviewed studies that focused on the effectiveness of BW as a crime-reduction strategy. The findings are mixed, at best. No study has definitively tied disorder policing to crime reduction.  Most importantly, “There is concern that any effectiveness of broken windows policing in reducing crime may come at the expense of reduced citizen satisfaction and damage to citizen perceptions of the legitimacy of police.” 

The studies that did find a link between disorder policing and crime reduction strongly caution about the harms to police legitimacy. Yet generations of cops have been indoctrinated in BW, with low-level arrests becoming routine, instead of the rare exception. For example, Minneapolis Police Chief Medaria Arradondo testified in the trial of Derek Chauvin that possession of a $20 counterfeit bill is “typically not” an allegation that leads to someone being detained, because it’s not a violent felony.

Police practices vary by state, though. In many states, such as Texas, police officers may arrest for minor motor vehicle offenses, which some argue are ruses to conduct otherwise unauthorized vehicle searches. These are referred to as “pretext stops.” The intent is the search, not the underlying minor offense. We know that police officers in the U.S. continue to routinely arrest millions of people for low-level offenses. Why? Because broken windows has become embedded in policing’s cultural lexicon and reinforced through the reward system, with recalcitrant police chiefs unwilling to yield this mystical crime-fighting tool.  

Increasing accountability

Police chiefs groom future leaders by selecting who attends career development training and receives choice assignments, continuing the insular group-think that allows broken windows policing to continue. Policing needs science, not proselytizing. Policing can learn much from medicine and aviation, for example — fields that work tirelessly to identify harms in order to avoid catastrophic outcomes.  Ignoring the blowback effects of a policy is irresponsible, negligent and — as fomer Minneapolis police officer Derek Chauvin learned — criminal. Chauvin has been held accountable for killing George Floyd. Where is the accountability for police practices that cause mass community harm? 

I remain firm in my conviction that science, not sensationalism or dogmatic adherence to tradition will improve policing. Indoctrination is how officers learn to navigate police culture, and that culture remains frozen as police chiefs identify their protégées for grooming.  Those who adhere to the culture of policing are rewarded with key training necessary for leadership roles.  Most chief positions require attendance in three select police leadership trainings; Senior Management Institute for Police hosted by the Police Executive Research Forum; The FBI National Academy, and less often, the Southern Police Institute. Critics of the chiefs’ favored strategies aren’t likely to be selected for these highly coveted positions, thereby cementing the current police culture for generations.  

I have been a staunch critic of broken windows and other police policies that lead to disparate outcomes. This has made me no friends among police leadership. Though I often asked for permission to attend these coveted trainings as I came up through the ranks, I was rebuffed.  When I finally was scheduled to go, a change in police administration occurred and suddenly there was no money to pay the $10,000 fee. I was only the second person in the entire history of the Senior Management Institute for Police program to pay out of pocket.  I knew if I did not attend, I would never be competitive for chief’s positions. Oddly, the funding for the program materialized the following year.

I know that I am not the only officer whose careers were stymied because of their independent ideas. In response, I partnered with Maureen McGough, chief of staff at the NYU Policing Project, to launch the The 30×30 Initiative.  Our goal is to change the face of policing and how we police by removing barriers that hinder the integration and promotion of women in policing. 

The 30×30 Initiative uses peer-reviewed science as its foundation, to advocate for agencies to reach 30% female representation by 2030. The number is based is based on science which states that a marginalized group cannot affect cultural change until they reach a critical mass of 30%.  Further, we are looking to support innovative thinkers in policing by providing access to training that would otherwise be inaccessible, so that the next generation of leaders understand how to discern science from sycophancy.   

Ivonne Roman is co-founder of The 30×30 Initiative, which seeks to increase the representation of women in law enforcement. Roman has more than 25 years of experience in policing, having served in every rank, from police officer to police chief, in the Newark, N.J. police department.

More: George Floyd’s family lawyer Ben Crump: ‘It’s been 57 years since we’ve had meaningful police reform’

Also read: It’s been a year since George Floyd’s death. Here’s how you can demand racial equity — and become a lifelong ally