The Global Urbanist

News and analysis of cities around the world

Has the "broken windows" theory of policing finally been fixed?

Lauded for successfully reducing crime in New York, the "broken windows" theory of policing has always been underpinned by a rather negative view of how cities worked, as Drew Reed lays out. New research now suggests that its successes might have been down to coincidences including, interestingly, the removal of lead from gasoline and thus from the population's bloodstream.

Drew Reed

Drew Reed

Cities: Los Angeles-Long Beach-Santa Ana, New York-Newark

Topics: City politics, Crime and security, Poverty and inequality, Homelessness

When we think about the cities we live in, some of us see them as nothing more than a noisy, crowded and potentially dangerous mess. Others are more optimistic, such as Lewis Mumford, who described the city as “a theatre of social action” where “man's more purposive activities are focused and work out, through conflicting and cooperating personalities, events, groups, into more significant culminations.”

While urban thinkers like Mumford have been making similar observations for centuries, the issue of how we perceive the city has more recently been taken up by social scientists seeking to apply a more methodical approach. Yet, personal opinions are still apparent in their results. One of the first well known attempts at this was the “broken windows theory”, which was as I shall explain based on a decisively negative, un-Mumfordian view of the city, seeing it as rife with disorder and in dire need of heavy-handed policing to be cleaned up.

Taking a sledge hammer to urban research

Our story begins in 1982. It’s the dawn of the Reagan Revolution in the United States, and in the March issue of The Atlantic Monthly professors James Q. Wilson and George L. Kelling put forward a new theory, in keeping with the spirit of the new political attitudes, on how to maintain social order in cities. Their article centres on a study done 13 years earlier by Stanford psychologist Philip Zimbardo, in which one damaged car was left abandoned in a "bad neighbourhood" (The Bronx) and another in a "good neighbourhood" (Palo Alto, California). Both were observed from a distance. In the Bronx, the car was robbed almost immediately, stripped of anything of value, and then generally vandalised. Its Palo Alto counterpart took a bit longer to start stirring up trouble, sitting a week untouched. Then, in a sudden bout of impromptu methodological revision, Zimbardo smashed the car with a sledge hammer. After that, the results begin to follow the same tack as in the Bronx; people robbed and vandalised the car, until it had been “turned upside down and utterly destroyed.” Zimbardo expresses disbelief at the fact that in both cases, the vandals “appeared to be primarily respectable whites.”

Building from this study, Wilson and Kelling conclude that an idea taken by disciplinarian types to be common sense — that disorder is contagious, and that a few “untended” elements in the urban environment will inevitably lead to a general breakdown in civilised society — is based on scientific evidence. Thus, the authors conclude, it is necessary for even minor infractions, such as vagrancy or drunkenness in public, to be punished to the full extent of the law. If this doesn’t happen, the consequences to urban neighbourhoods will be that, to borrow from Dostoyevsky, "everything is permitted". Simple acts of homelessness and loitering will descend into all-out warfare.

It’s worth pointing out that this theory, despite its draconian implications, takes the noteworthy step of taking into account the critical importance of the perception of urban environment in shaping our lives. Nevertheless, it denies that this perception has any implications beyond simply raising or lowering the crime rate. Ultimately, Wilson and Kelling’s view of the city is antithetical to Mumford: it’s little more than a source of constant inconvenience. And in order for that inconvenience not to devolve into a living hell, all aspects of disorder must be removed, from broken windows to (as they put it) “undesirable persons.” Their proof of this was a sociological methodology literally as direct as a sledge hammer, and packaged into an idea with a catchy name: the “broken windows theory”. Though when it was released this theory didn’t become the phenomenon it would later turn into, it made a lasting impression on a young student at the Police Staff College in Bramshill: William J. Bratton.

“Broken windows” comes to NYC

Fast forward to 1994. In New York City, an energetic Rudy Giuliani takes over as mayor, and in an attempt to pull the city out of its malaise left over from the general grunginess of the 70s and 80s, Giuliani takes action to make good on his pledge to clean up the city. He hires Bratton, who since his days as a student has risen in the ranks first of the Boston Police Department and then the New York Transit Police to become a prominent figure in policing as NYPD Police Commissioner. And Bratton wastes no time in putting the theory of Wilson and Kelling into practice. He begins by shaking things up at NYPD headquarters. Though never afraid to come down hard on rank-and-file cops stepping out of line, Bratton was more than just a hard-ass. His restructuring of the department and use of the computerised COMPSTAT crime location system won the praise of management scholars and tech-savvy Silicon Valley types alike (he was called “the Steve Jobs of police work”). Nevertheless, the core of his theory was to fix the broken windows, as it were, by cracking down with a “zero tolerance policy” on “quality-of-life crimes” such as loitering, drinking in public, and famously, window cleaning, as performed by New York’s population of squeegee-wielding car window washers.

And by all accounts, his plan worked. In the mid 90s, crime in New York dropped dramatically, making Bratton a national and international hero (his early departure was prompted by Giuliani trying to steal the credit from him), and along with him, the theorists Wilson and Kelling. Hot off his success, Kelling co-authored an entire book elaborating on the broken windows theory, and the two of them were awash in high-paying consultancy deals aimed at fighting crime in urban environments.

Nonetheless, some people were still uneasy with the heavy-handed approach of the broken windows theory. In 2004, sociologists Robert J. Sampson and Steven J. Raudenbush set out to examine the issue of urban perceptions of disorder. They launched an ambitious study, examining over 500 census-block groups in Chicago with surveys, interviews, and thousands of hours of video footage. Needless to say, their methods were a bit more sophisticated than the original Zimbardi study, and their conclusions were likewise more nuanced than those of Wilson and Kelling. They conclude that “broken windows”-style policing “may have only limited payoffs” in poor neighbourhoods, a conclusion that Chief Bratton was about to find out.

LA's Skid Row

As Bratton continued to revel in his success in New York, city officials elsewhere were eager to put the theory into practice as well. In Los Angeles, city officials at that time suffered from a decidedly negative image in the wake of the 1992 riots and the Rampart police corruption scandal. And what better way to cast off that negative image than by bringing in “America’s Top Cop”? In 2002, LA mayor James Hahn succeeded in locking down Bratton as LAPD police chief. Initially, Bratton achieved success with internal restructuring that helped to break up the secrecy culture that had produced the atrocities of the 90s. But for both his department and city hall, there was a lingering black mark on LA’s reputation that needed to be confronted: Skid Row, the neighbourhood with the highest concentration of homeless in the country. A legacy of the large number of temporary residences built to accommodate railroad workers during layovers at the end of the line, Skid Row had evolved into what was in effect the only place society’s misfits — veterans, the poor, the mentally ill, what Wilson and Kelling might call “undesirable persons” — could stay without being picked up by the police. In the 80s, the zone was the centre of a “containment” policy, in which social services were concentrated in the area, with the implicit goal of sucking these “undesirables” out of more desirable areas. Needless to say, crime was a big problem in Skid Row.

This made it a perfect target for the latest application of Bratton’s tried and true “broken windows” style of policing. And the arrival of mayor Antonio Villaraigosa in 2005 gave him the perfect chance to put it to the test once again. In conjunction with Villaraigosa, Bratton’s Safer Cities Initiative proposed to concentrate police forces on Skid Row, while also bolstering the level of social services in the area. In 2006, the plan took effect — or at least the policing side of the plan, since social services had only been given a fraction of the policing budget to work with. Levels of police concentration reached an average of one officer for every block in the entire district, which critics described as “the highest levels outside of Baghdad.” Cops quickly began enforcing “quality-of-life” crimes. Thousands of arrests were made for offences such as jaywalking and sitting on the sidewalk, with various cases of police abuse reported. Meanwhile, little thought was given to why people were sitting on the sidewalk in the first place, and whether the broken windows theory still held true in a place where external forces had concentrated poverty so densely.

Ultimately, crime did go down in Skid Row. But the programme was not seen as the resounding success that Bratton had received in New York. One of the biggest opponents of the plan was Gary Blasi, a law professor at UCLA, who released a study in 2008 in which he elaborated on the position of Sampson and Raudenbush to conclude that the decrease in crime was statistically equivalent with all other central city neighbourhoods, which didn’t have such high concentrations of police. Bratton, bolstered by a generally supportive local media, never responded to this criticism, though he did admit that some of the Skid Row homeless had simply been displaced, either to other neighbourhoods or to California’s burgeoning private prison system. However, in 2009 he abruptly cut his term as police chief short, leaving the Safer Cities Initiative in the hands of his successor Charlie Beck. The initiative is still in place today, and though proponents continue to point to the fact that crime has gone down, no one can claim that Bratton’s programme achieved the same resounding success he had back in New York.

A new era of research …

Which brings us to today. 2013 has seen some interesting new developments for the broken windows theory; in January, Kevin Drum of Mother Jones wrote an influential article openly challenging the notion that it was the broken windows theory that reduced crime in New York. The reason given instead was that the residual effects of lead in the air, created by leaded gasoline, were beginning to wear off. Drum highlights a series of studies linking lead concentration in the bloodstream to higher crime rates, an effect proven to last 20 years from contamination, which would coincide with the removal of leaded gas from the city's atmosphere. He also notes that in the 90s crime didn’t drop significantly only in New York but also in several other major American cities. While the removal of leaded gasoline may not entirely explain the drastic fall in crime either, it’s an important element that was missing from the discussion when the broken windows theory was first put into practice.

But there’s an even deeper issue that neither the original theory, nor its well received application by Bratton in New York — and not-so-well received application in Los Angeles — nor any subsequent study in the “broken windows” canon took into account. Until now. “The collaborative image of the city”, a recent study published at Plos-One by Philip Salesses, Katja Schechtner, and César A. Hidalgo of MIT Media Lab has set out to address the same fundamental issue: people’s perceptions of the urban landscape. What’s significant about this study is it goes beyond the binary input (order/disorder) and output (crime/safety) taken as a given by nearly everything related to the broken windows theory that came before, instead measuring perceptions of safety, social class, and “uniqueness” of a neighbourhood, and in subsequent phases of research, how “lively”, “boring”, “wealthy” or “depressing” it is. To do so, they made use of today’s social media saturated internet culture, presenting survey participants online with two pictures and asking them which appeared to possess more of a certain characteristic.

Though crime does figure in to the study’s final conclusions, the authors also point out that “there is a strong need to create quantitative bridges that can help us link urban perception with other social, political, economic and cultural aspects of cities.” They end by summarising the goal of their work: to “contribute to the understanding of urban environments”, to improve their “ability to include their citizens”, and finally, to “inform the construction of future cities.” In the more than 30 years the broken windows theory has been in existence, only now have scholars taken a methodological approach to urban perceptions with the goal not only of enforcement, but also improvement. An approach that sees the city not as a mere inconvenience but a Mumfordian social theatre, and analyses it as such.

… needs a new champion?

Could this be the dawn of a new age in how we think about civic perceptions? Well, maybe. It’s important to note that in order to be successful, the broken windows theory was helped by the fact that it was given a catchy name, and that it was constantly touted by Bratton, a charismatic public figure. But Skid Row proved that improving troubled neighbourhoods is more complicated than simply locking up people for sitting on the sidewalk. As Salesses, Schechtner, and Hidalgo indicate, it will require a wholesale reevaluation of the construction of the urban environment, geared toward inclusion. 

Our perceptions of the city are important not just because they may possibly provoke us to commit crimes if they are negative. They’re important because they have the power to help us see the disadvantaged not as “undesirable persons”, but as fellow human beings, if they are positive. This new research affirms this idea, and it may yet prevail where broken windows theory has not. But in all likelihood, it’s going to need a new champion who knows just how to put it into practice.


is a writer and sustainable transportation advocate. He currently lives in Buenos Aires, Argentina, where he is a graduate student in urban planning at the University of Buenos Aires.

GU