The controversial “broken windows” theory of policing, the idea that cracking down on smaller quality-of-life crimes can help prevent larger ones from occurring, has been back in the headlines lately—especially since its most visible proponent, William Bratton, returned to New York City as police commissioner this year. As the city expands its police force and faces a backlash for some officers’ overly harsh tactics for fighting very minor offenses, even Bratton seems to be open to re-considering the approach.
Do signs of disorder and disrepair in a neighborhood really invite or encourage crime? Or are disorder and crime both symptoms of deeper and more complicated problems? Those are the questions asked by a new study in the Journal of Research in Crime and Delinquency, by Daniel O’Brien, a public policy and criminology professor at Northeastern, and Robert Sampson, a social sciences professor at Harvard University.
By analyzing hundreds of thousands of 911 calls and requests for non-emergency services in Boston over two years, authors found that “private conflict” in the community, not “public disorder,” was the biggest predictor of crime. I spoke to them about their findings; this is an edited and condensed version of those conversations.
What types of crimes and neighborhood problems were you looking at for this study?
Sampson: For public social disorder, we measured reports of drunks causing disturbances, panhandlers, and things like that. The private conflict was mainly driven by domestic violence and intimate partner disputes, but also included things like landlord-tenant disputes. These are the kinds of private conflicts that bubble up and spill into the public sphere—and our analysis shows that private conflict is strongly predictive of future crimes.
Moreover, violence in public is more important in causing public disorder than the other way around. It becomes a reciprocal model over time. So, for instance, some bars get reputations for having a lot of fights, and then people kind of assume that there are fewer repercussions for fighting in that bar. Or encounters that are becoming hostile that will actually escalate into further conflict—those are more likely in those kinds of contexts. So we’re trying to put things into a larger perspective here, and to understand what is more hidden.
O’Brien: One of the things we say in the paper is that these administrative records act as the eyes and the ears of the city. They are people reporting what they are seeing and experiencing, outside of the reach of traditional research protocols. We realized we could see into the private disorder across the city, which is not identical to public disorder—they are different constructs, and they have different implications for neighborhoods.
What do you think of the “broken windows” theory in general, and the debate surrounding it? It seems as though popular opinion of it has shifted a lot over the years.
Sampson: I do think the debate has shifted. I think a lot of things have gotten lumped into the idea of “broken windows” and policing it, and there’s a real need to unpack the different mechanisms. You can have aggressive policing, you can have signs of disorder in public spaces, you can have private conflicts, and all of these things are not the same things. I think people try to just lump them all together, and part of the problem with this aggressive policing—the idea that you can just crack down—I think is problematic.
I’m not so sure that all of those policies can be traced to the original writing; there’s been a promiscuous attribution to “broken windows” that goes way beyond the bounds of the original theory. That said, I’ve been critical in other writings of the idea that cues of disorder are themselves criminogenic. I tend to think that other things, such as poverty and resource deprivation, are more crucial.
O’Brien: I’ve done a study where, we showed people pictures of neighborhoods that they’d never seen before, and we asked them how safe they thought they were. We also had neighborhood surveys, so we knew the safety of each neighborhood. We found that people were actually relatively accurate in their assessments of the neighborhoods, based exclusively on the presence of physical disorder in the pictures. That (and other work) suggests that there is this innate responsiveness; we do have a psychological focus on disorder—at least in America, and a lot of other developed nations—we look for this kind of evidence about what a neighborhood we’ve never been to before is like.
As Kelling and Wilson said in their original article, elements of deterioration and neglect essentially communicate that this place is not being taken care of, and if that’s the case, then it’s also possible that people could commit crimes around here and not be stopped. That’s exactly the kind of thing that you want to figure out when you’re in a new place.
So it really is no surprise that the idea that those elements actually cause crime becomes a really attractive idea. That’s why I think it’s been such a mainstay of policymakers. I think, though, that our study and others have shown that there really is limited evidence that policing it works—that it’s effective in preventing more serious crime, and discouraging those sorts of escalation or expansion of crime.
But aside from a connection from “public disorder” to crime, you did find a connection between things like litter and graffiti to non-crime problems in the community, like stress, health problems, and school drop-outs, right?
O’Brien: Yes. In the whole “broken windows” literature, I’ve found that the thing that there is the most evidence for is the health effects: things like depression, self esteem problems, and hopelessness. There seems to be some reasonable evidence that the disorder itself could be playing an important role. We know that people are walking around and looking at disorder as a cue for what could occur in their environment. So observing disorder can cause a consistently elevated level of stress, and people’s bodies can actually adjust their levels of stress in response to it over time. So there is good reason to believe that disorder, while maybe not instigating further crime, has other really important effects on health and behavior.
Sampson: In retrospect, it does make sense, when you’re looking at a stressful ecology that people are living in. It’s one thing to walk down the street where people are arguing and fighting, but being exposed to that on a daily basis is another matter.
Your study really shows how some problems in communities are more visible than others, and how some things are better addressed by police than others. I wonder if sometimes law enforcement policy focuses on small quality-of-life crimes because those are the things that they can actually change. Whereas it’s not really in their job description to go into people’s homes to administer therapy, or to try to get at the roots of interpersonal conflict.
Sampson: And I think that’s related to the larger problem that we see with policing in society. We expect the police to expect to deal with all these problems, but they’re not really equipped. Many of these problems are stemming from drug abuse, alcoholism, and mental health issues, where people are on the down and out and in conflict with each other. The police really aren’t very well trained to deal with these things, and we kind of put them out there and expect them to. It’s just a mismatch with the reality of these problems....
Counseling, mental health treatment, domestic violence interventions, dispute mediations: these are things that are crucial, more from a public health perspective than as a policing aspect. The “man in the badge” is not necessarily able to handle all these things, nor should we expect, as a society, that that should be the case.
O’Brien: It’s funny, it’s another way to understand why the “broken windows” model of policing has become so alluring. After that piece came out in the Atlantic 30 years ago now, it really captured the public imagination, as well as the imaginations of policymakers and practitioners across the country—and in other countries, as well. I think there’s a simplicity to it, where the root cause, as you said, becomes accessible—it comes available to be directly accessed, and directly fixed. Since the problems that are being cited as the root causes of crime are within the general purview of law enforcement, then the challenge is just putting enough effort into eliminating them, and then you can mitigate future crime, to some extent.
I think one of the challenges of this paper is that, the question can then become, “OK, so should the focus of police interventions then be on the private domain?” That sort of assumes that, if we identify a different cause, then the approach should be equivalent to the one we’ve already developed. But because the dynamic underneath it is so different, we should be re-conceptualizing how we approach the situation.
I wouldn’t recommend trying to eradicate private conflict as a way to prevent crime. I would say that we need to be thinking about crime as an endemic social process—something that’s bubbling out from the social strife that emerges from the interactions with the people we share space with. So if we’re thinking about crime prevention measures, we should be thinking about how to provide the support to de-escalate conflict, and how to identify areas where there may be future escalation. Our loved ones, our roommates, our neighbors, our landlords and tenants: those relationships are the basis of a neighborhood.
True Crime is Lauren Kirchner's weekly column about crime and criminal justice issues.