“You Can Reduce Violence But Harm People”: A Conversation with Caterina Roman
By Greg Berman
“Libraries, parks, rec centers, pools, free internet — those are all crime prevention activities and resources,” Caterina Roman, a professor of criminal justice at Temple University, told the New York Times last year.
Over the course of her career in academia and at the nonprofit Urban Institute, Dr. Roman has taken a broad view of violence and how to prevent it. Among other topics, she has investigated reentry programs, community justice partnerships, and the social networks of at-risk youth.
I talked to Dr. Roman by phone in February, not long after she had published an op-ed in the Philadelphia Inquirer calling for deeper investments in evidence-based gun violence prevention programs. Much of our conversation was devoted to focused deterrence, an approach to reducing gun violence that emerged in Boston in the 1990s and has been broadly disseminated by the National Network for Safe Communities, under the leadership of John Jay College professor David Kennedy.
Focused deterrence begins with a recognition that most of the violence in a given community is perpetrated by a small group of individuals. It seeks to target these individuals with offers of assistance, providing them with the job training, drug treatment, and other services they might need to avoid future offending. If they do not cease their violent behavior, the focused deterrence model encourages the justice system to use all of the levers at its disposal–including arrest, prosecution, and incarceration–to halt the violence. The use of focused deterrence strategies has been associated with local reductions in homicides, according to a number of evaluations.
In addition to focused deterrence, my conversation with Dr. Roman touched on gaps in our collective knowledge base, particularly the need for more information about effective crime prevention. The following transcript has been edited for length and clarity.
Greg Berman: How concerned are you about the spike in homicides we have seen over the past year? Do you think that the increase is something that we need to be worried about?
Caterina Roman: I think we really need to be concerned about it. I think we’ve reached some kind of a tipping point. When you look at the reasons that people generally offer for any major spike in violence, all of them come into play with COVID. You have a little bit of everything. You have so many people buying guns. You have more hurt people who will hurt people. You have disinvestment that has been exacerbated. You have governments that aren’t funding parks, rec centers, summer jobs. You don’t have in-person religious services.
Police departments have the resources to do data-driven work that can be useful in communities. Where are the hot spots? How long have they been hot? What makes them different from other areas? What have we learned from both our successes, our failures in addressing hot spots from five years ago?
You don’t have outreach workers on the streets or the typical social services. And then you have compounded stress. All of these things are coming to a head together. There’s no place to go. Young people don’t have the access to pro-social jobs that will keep them busy and put them in contact with potential mentors.
Recent writing about the increase in violence seems to fall into two categories: an effort to score political points or a search for simple, silver-bullet answers to what’s going on.
I talk to my students a lot about the media. I’ve studied fear of crime, and I often ask, “Where does that fear come from?” Some of it is based in reality, but most of it is not. Having had many conversations with journalists over the years, I know that journalists often like to focus on just one thing to hang their stories on. Some of this comes from journalism itself, which requires stories to have a hook. It is also true that policymakers tend to want quicker fixes and easier answers. If you hang your hat on one cause, you have a straight line to a solution. In reality, it’s just not that simple.
You talk about the structural forces that lead to violence, including inequality and racism. The investments that we need to make to address these kinds of problems are probably generational. But we also have a need to move right now in response to an immediate crisis. How would you advise a mayor or some other political actor who wanted to make a difference? How should we balance long-term investments with short-term strategies to quell violence?
I think policymakers and politicians should be direct and transparent with regard to longer-term investments. Given that the world knows that poverty has some relationship to crime and that disadvantage has some relationship to crime, it would be great if policymakers and politicians would just be open and say, “We are optimistic that we can make longer term change, and we’re going to do it by investing in neighborhood infrastructure. We’re going to do this, we’re going to tell you where the money’s going, and we’re going to measure the incremental change over time.”
We're only looking at the outcomes related to police data and changes in violence at the aggregate level. We're not asking who benefits from the intervention and who is burdened. We're only focused on the data we have at hand.
I recognize that people want a quick fix. I’m not going to tell you that I believe that X policing solution or Y law enforcement solution is an answer. But I do think that police departments have the resources to do data-driven work that can be useful in communities. Where are the hot spots? How long have they been hot? What makes them different from other areas? What have we learned from both our successes and our failures in addressing hot spots from five years ago? This is where I would advocate for practitioner-academic partnerships because we know even the best police departments with the most data aren’t necessarily applying it in a larger, theory-based way. By creating these partnerships, we can ensure that policing will be used for smart strategies and reduce the likelihood that we’re sending police out on calls that have nothing to do with violence.
One of the problems in New York at the moment is a decline in clearance rates for homicides. Do you have some thoughts about what police should be doing to improve this?
There’s very little research out there on how to improve clearance rates. It is a huge gap in our knowledge. There’s a whole gamut of programs that are trying to achieve community-level change, whether it’s a focused deterrence/pulling levers model or Cure Violence or something else. I would advocate for researchers who have studied these models to go back and look at whether clearance rates were differentially affected in the treatment versus control neighborhoods. That’s relatively easy to do. Maybe where there’s less violence on the street and less fear of crime, clearance rates go up.
You mentioned focused deterrence policing strategy. I’m curious to hear how you are thinking about focused deterrence these days and what we can say about its effectiveness as an intervention.
I’ve evaluated focused deterrence in Washington, D.C., in its first sort of iteration right after Operation Ceasefire in Boston. And I think the big issue for me about its effectiveness is that we’re only looking at the outcomes related to police data and changes in violence at the aggregate level. We’re not asking who benefits from the intervention and who is burdened. We’re only focused on the data we have at hand. This tends to be what evaluators do. It’s easy to get arrest data, so that’s what we measure. And so we only know from the majority of evaluations of focused deterrence that it reduced violence at the aggregate level.
For every intervention we need to be asking who benefits and who is burdened ... What are the unintended consequences of using credible messengers to go into the community and be pro-social mentors and caseworkers?
We just know that we got the end result of reduction in violence. But what got us there? [John Jay professor] David Kennedy’s theory of change is that the threat of this very focused deterrence led to general deterrence. But we have no research that shows us that that is true. Yet everybody who advocates for focused deterrence is saying that it is an evidence-based program, that it’s working and this is what we should do. I don’t know if that’s true.
So, if you were going to sponsor research into focused deterrence going forward, what would that look like?
Any intervention that expects a community-level reduction in violence has to have enough research dollars behind it to fund a comprehensive survey to track individual-level behavior change. You want to be interviewing those who are targeted in the initiative, and following up with them, and then also interviewing potential high-risk individuals in the community. This won’t be cheap. You’re at a million dollars right there. But that’s where I believe we have to go. If we want answers, we need to be able to conduct the kind of studies that are going to let us measure what is really happening at the community level.
You talk about people advocating for more focused deterrence. Recently, I’ve seen a number of calls for deeper investment in Cure Violence. Do you think that the evidence merits this push?
I think you’re asking the question in the wrong way. I think for every intervention we need to be asking who benefits and who is burdened. If you frame the question that way, I would have to think long and hard about who is burdened by Cure Violence. What are the unintended consequences of using credible messengers to go into the community and be pro-social mentors and caseworkers? I don’t think there is an obvious burden to funding Cure Violence as it’s intended to be implemented, whereas there are so many things that could go wrong with focused deterrence, given its complex implementation structure. You can reduce violence but harm people.
You are advancing a kind of “do no harm” argument on behalf of Cure Violence. But do we know that it actually does what it says it does, in terms of reducing shootings?
I’m not sure. It’s supposed to be increasing legitimacy by telling the community to be collectively accountable and bringing up the moral voice of the community. But I also think you have to ask whether focused deterrence works in the long run. Is focused deterrence successful at reducing violence if once everyone knows focused deterrence is going away, they start shooting again? Please answer that for me.
Instead of Cure Violence, you could put money into victim services to make sure that every single person that's a victim of violent crime has everything they could possibly need. Victim services are basically nonexistent in most urban communities.
I don’t want us to get locked into a focused deterrence versus Cure Violence conversation, because I don’t think that dynamic is helpful.
I wasn’t arguing for one versus the other. I am using those two models as an example of investment versus policing, surveillance, and deterrence. It does not have to be focused deterrence. It does not have to be Cure Violence. I think instead of Cure Violence, you could put money into victim services to make sure that every single person that’s a victim of violent crime has everything they could possibly need. Victim services are basically nonexistent in most urban communities.
Are there other programs out there that you feel are interesting, whether or not they’ve been evaluated with any degree of rigor?
Do you know about the Chelsea Hub? This is a model where community agencies work collaboratively with the police, probation, school, and victim services. Everyone is meeting weekly. Let’s say I am a victim service agency and in walks someone who witnessed a shooting and that individual seems like they are in a crisis moment. That victim service agency would ask that person, “Would you be willing for me to present your case to a group of service providers to talk about how we could strategize about what you might need holistically?” So the Hub is a method to provide holistic services to people who are in some type of crisis. It could be a domestic violence event. It could be after a hit and run incident. It doesn’t have to be violent crime, necessarily. Philadelphia is testing the model now through Temple med school. It’s voluntary for the individual. It is a positive, full-investment, collaborative model that cuts across systems so you can get to the complexity of the issues and offer an array of useful services and gain individuals’ trust. There hasn’t been a long-term evaluation of it, but it’s a very promising model. The GRYD model in Los Angeles is also worth checking out. I think they’ve had some impact evaluation work done that is pretty strong.
Let’s talk about some of the articles you have written. A few years ago, you wrote about gang research and how to get people to leave gangs. What did you learn?
The point of that article was to look across three very large studies to identify the kinds of pushes and pulls that get somebody out of the gang. A “pull” is pro-social. It is anything from “my significant other doesn’t want me involved in that anymore,” to “my grandmother says she’s not going to let me come home if I’m still hanging out with them.” A pull is some pro-social opportunity, like a job, that is getting me out of the gang. A push tends to be more negative. A push can be: they were victimized, or they were incarcerated, or they have gotten tired of being roughed up by the police. But it could also be they just realized that the gang wasn’t what they wanted.
You have also looked at fear of crime in Washington D.C. Tell me about that research.
That piece came from my interest in social capital and collective efficacy.
How would you define collective efficacy?
Collective efficacy is the activation of social ties and informal social control among neighbors. So we ask people, “How likely are your neighbors to help out another neighbor in need?” Or, “If a group of teens is hanging out on the street corner, being rowdy, how likely are your neighbors to do something about it?” We aggregate information from questions like these to measure collective efficacy.
We don't have good evidence on prevention, because we don't research prevention.
In the survey that I did in the northeast section of Washington, D.C, we looked at how collective efficacy was related to fear of crime as measured by people reporting that they were not going to go walk outside because they were worried about crime. At first, what we found jibed with the literature—older people and women tend to be more fearful. As we added different variables to the model, we looked at the interaction of collective efficacy on Black residents versus White residents. What we found was that there was an increase of fear when collective efficacy was higher among Black respondents. We did not find any significant effect among non-Black respondents. We posited that, as collective efficacy increased, Black residents in those neighborhoods were talking more and transmitting more information about violent crime and what was actually happening in the neighborhood. And that relaying of information in the neighborhood increased fear. So higher collective efficacy meant more fear for Black respondents.
What do you think is the biggest misconception that policymakers have about crime prevention?
The truth is that we know very little about what works because we don’t test prevention. We don’t test prevention mechanisms like Pre-K. In Philadelphia, where I live, we have 4,000 more kids in Pre-K each year over the last couple of years. We don’t know if that’s going to reduce violence, because we’re not testing that. So when a policymaker goes to the evidence base, they’re looking at the interventions that were more likely to be evaluated. As we have discussed, policing programs are relatively straightforward to evaluate: You get crime data, that’s really simple. What we’re not doing is funding the kinds of survey research that would give us evidence that legitimacy is increased, that moral cynicism is reduced, that more people are integrated with their neighborhoods. We have no idea how to increase collective efficacy. That’s why we can’t solve the violence problem. So, going back to your question, I think what’s not understood is that we don’t have good evidence on prevention, because we don’t research prevention.
If you were going to make a reading recommendation to an audience that is interested in community-based violence, is there a single book or single study that you would point to?
If someone is interested in this topic, they should spend a week with an outreach worker. They should spend a week in a victim services agency. They should be inside the neighborhood. You’re not going to learn anything from a book. If you want to change a neighborhood, be inside it, and see if you can feel it.
But to answer your question about a book, I would tell you to check out the Aspen Institute’s roundtable on community change, which was turned into a book edited by Karen Fulbright-Anderson and Patricia Auspos. Dennis Rosenbaum has a great chapter in there on promoting safe and healthy neighborhoods. I’d also encourage people to read Wes Skogan, a criminologist who studied policing and community change.
Greg Berman is the Distinguished Fellow of Practice at The Harry Frank Guggenheim Foundation. He previously served as the executive director of the Center for Court Innovation for 18 years. His most recent book is Start Here: A Road Map to Reducing Mass Incarceration (The New Press).
Views expressed are the participants’ own and not necessarily those of The Harry Frank Guggenheim Foundation.