Ferguson, Broken Windows, and de Blasio's Dilemma

by Michael Dorf

In my new Verdict column I endorse the proposal for police to be equipped with wearable cameras to record police-citizen interactions. I note the legitimate concerns raised by these proposals but conclude that they can be accommodated through careful implementation. On the whole, I agree with the view that recording is win-win: It will protect citizens against abusive policing and protect honest police against bogus allegations of abuse. Nonetheless, I explain why recording police is no panacea. There will still be disputes over what the recordings show (as in the aftermath of the Rodney King beating) and the dangers inherent in all police-citizen conflicts will mean that, even when police know they are being recorded, they will sometimes use deadly force with tragic consequences. Thus, I argue that policy makers should take steps to reduce the frequency of police-citizen interactions with the potential for violence. I suggest that the increased risk of events like the shooting of Michael Brown is one of the costs of over-criminalization.

Here I want to ask whether there are implications of my analysis for "broken windows" policing. As most readers probably know, the term "broken windows" was coined in the early 1980s by James Q. Wilson and George Kelling to refer to signs of low-level social disorder: broken windows; graffiti; litter; etc. In their theory, neighborhoods with such evidence of minor disorder embolden criminals to commit more serious crimes. Conversely, aggressive policing to attack minor crime creates liveable environments and a virtuous cycle of law-abiding behavior.

The most well-known attempt to implement broken-windows policing occurred during the NYC Mayoral administration of Rudy Giuliani. He cracked down on "squeegee men"--who "cleaned" windshields of motorists stopped at traffic lights, sometimes with an implicit threat of damage to the car or worse if drivers did not agree to pay for this ostensible service; he went after graffiti artists; he targeted subway fare-beating. And--according to the proponents of the broken windows theory--it worked. The nation as a whole experienced a substantial drop in violent crime from the peaks of the late 1980s and early 1990s, but the drop in crime in New York City was substantially larger.

How much of the drop in crime was really due to broken windows policing? That is a hotly debated topic among criminologists. The causes of the decline in crime nationally and in particular locales remain unclear. Various factors to which different scholars point in varying degrees include: more police on the street; targeted policing using big data (as in "Compstat"); more arrests; longer prison sentences; the economic boom of the 1990s through early 2000s; the "big brother" phenomenon in which young people saw the devastation wrought by crack cocaine on the generation ahead of them and were scared straight; legalized abortion; the remission of violence as a "contagion"; and more.

According to one account, broken windows policing was important but not for the reasons originally identified by Wilson and Kelling. In this alternative view, the key was New York's mandatory minimum sentence of imprisonment for carrying an unlicensed firearm (in combination with very restrictive gun licensing policies). Prior to broken windows policing, a young man living in a dangerous neighborhood in NYC might typically go out armed, even if he was not a serious criminal. But knowing that the odds of an arrest for some minor offense (like fare-beating) went up under the new policing policy, he would leave his gun at home, for fear that a minor ticket would turn into a substantial prison sentence following the stop-and-frisk. Thus, with fewer guns on the street, there was less violence.

With real disagreement and puzzlement persisting among professional criminologists, I am not going to venture a guess as to how much weight should be given to each of the factors mentioned above (or others) in reducing crime. Here is a 2002 paper arguing that broken windows policing in NYC accounted for about half of the decline in robberies and motor vehicle thefts. Here is a 2006 paper arguing that there is no good evidence for the efficacy of broken windows policing.

Suppose that you are a policy maker who does not have the luxury of waiting another 20 years (or longer) for a consensus to emerge among criminologists. Suppose further that, after consulting the best experts you can find, you think that broken windows policing does play an important role in suppressing crime. (Again, I don't take a position on whether this is true; I'm just asking readers to imagine that they think it's true.) Does that mean that your administration should implement a broken windows policing strategy?

I think that, even assuming some efficacy for broken windows policing, the answer is unclear. That's because broken windows policing may reduce or suppress crime, while at the same time causing or exacerbating other problems, like friction between the community and the police, and creating more opportunities for violent police-citizen conflicts.

What I have just described is far from hypothetical. It appears to be the predicament in which NYC Mayor Bill deBlasio now finds himself. His successful mayoral campaign rested in no small part on his opposition to the stop-and-frisk policies of the prior Bloomberg and Giuliani administrations, which were very unpopular among the city's minority population. Mayor de Blasio has taken steps to revise those policies, including in the litigation still ongoing in federal court. But at the same time, de Blasio handed over the job of NYC Police Chief to Bill Bratton, who is a believer in and practitioner of broken windows policing.

As a political matter, de Blasio's choice of Bratton was shrewd. Bratton is the closest thing one can find to a rock star among major metropolitan police chiefs. He has credibility with police based on his successful prior stint as chief in NYC and his successes elsewhere. He also has credibility with the progressive and minority constituencies who supported de Blasio's election because Bratton believes in a diverse police force and at least some version of community policing. Bratton also has credibility with these constituencies because of his falling out with Giuliani, although their feuding may have had more to do with who should get credit for reducing crime than policy differences.

But at the end of the day, these considerations will only go so far. So long as Bratton's NYPD pursue broken windows policing, arrest rates will be high, and police-citizen conflicts will occur with some frequency. That may be an acceptable price to pay for keeping the violent crime rate low, but it will inevitably trade off one set of goods for another, and for that tradeoff de Blasio is likely to pay a political price. Indeed, it appears that he already is paying that price.