A Major Challenge to Policing Reform: the Absence of Good Data

A leading researcher discusses what we know about excessive use of force and its possible solutions—and what we don’t know.

Police in massive armor and by an armored vehicle on a street in Minneapolis
State patrol officers stand by a sheriff armored vehicle during a protest over the death of George Floyd on May 31, in Minneapolis. Chandan Khanna/AFP via Getty Images

The police killing of George Floyd in Minneapolis unleashed a protest movement that is unprecedented in its scale, but it’s addressing problems and injustices that many have been advocating against, and that experts have been studying, for years.

One such researcher is Barry Friedman, a legal scholar and author, and the founder of the Policing Project at the NYU School of Law. In June, New York state Attorney General Letitia James appointed Friedman, alongside Obama Administration Attorney General Loretta Lynch, to investigate the NYPD’s actions against protesters. In a two-part interview on Slate podcast The Gist, Friedman joined host Mike Pesca to discuss his research into police use of force, accountability, and reform. During part of the interview with Friedman released Thursday, the two discussed controversy over police departments’ military equipment, gaps in data about police violence, and the trouble with identifying what makes a good police department.

A portion of the discussion is transcribed below. It has been edited and condensed for length and clarity.

Mike Pesca: During this debate, you hear people argue that what we need to do is demilitarize the police—that we give them excess military weapons, essentially. And they show up on our streets in things that look like tanks—those BearCat armored vehicles look like tanks. And police officers dress like they’re ready for war. They dress like some version of Iron Man.

I’m sympathetic to that. I don’t think that that would be a bad reform. I just don’t think that would go to anything near the core of what we’re talking about [in terms of solving the problems at hand]. But what does your research show?

Barry Friedman: So actually I’m going to push back a bit.


We talk sometimes about community policing and about the relationship between police and the community. But it’s hard to relate to somebody that’s decked out like they’re going to war. You’re not going to develop much of a relationship with that person; you’re going to want to walk around them.

I remember once when I was visiting Paris with my kids and we went to the Louvre, and these officers came by with serious weaponry and all I thought was, “I want to get my kids the hell out of there.” And so I do think that if you want to develop sound relationships between the community and the police, then you’ve got to stop walking around looking like you’re headed to war. So I think that matters.

A lot of this equipment is coming from the military, but it’s not all military equipment. They get a lot of file cabinets and computers and stuff like that. And even some things that look like military equipment—like some of the armored vehicles may actually have a role to play in certain situations, including if you’ve got a shooter somewhere. But what matters is how the equipment gets deployed.

To me, that’s exactly right. That’s a deployment issue as opposed to—I’ll let you say it, but it’s fine for a department to have one of these tanklike operations, but that doesn’t mean they should be used to roll into a housing complex to try to nab some guy before he flushes his drugs down the toilet.

Yeah. We have gone insane about what they call tactical teams and tactical forces. And one of the chapters in my book—every chapter in the book starts with something that went horrifically wrong in policing. And one is this story about a mayor in Prince George’s County, Maryland, whose dogs get killed by the police in some raid where they just tear up his house that was totally based on a mistake. And we do this countless times every day quite frankly, because we started with this idea of SWAT teams for the sniper in the tower where there’s a real problem, like in Las Vegas, where you’ve just got to be able to get in there and save people’s lives. And then all of a sudden, every department wants one. And by the time every department’s got a tactical team, they are not the Navy SEALs, and they are not getting the Navy SEALs training, but they are getting the weaponry.

And that’s just a mistake. As I say in the book and as I say to people all the time, whatever happened to come out with your hands up, which most people are going to do? They get it that they’re outgunned, and they don’t really want to get wiped out. Another situation sometimes might be when you see [military equipment] used in protests. That’s a problem because it’s just too much weaponry for the job. And so I agree that there are circumstances in which departments may need to have some of this equipment, and I think I’m with you on that.

I’m a broken record: If nothing else, go look for the state laws that regulate SWAT team use. Do you want me to tell you how many there are? Take a guess.

If the answer is 48, it wouldn’t be a good rhetorical point. I’m going to say 10.

Yeah. Well, the answer would be either one or zero.


Utah has a law that Maryland once had that required keeping records on when SWAT teams were used, but the police unions went nuts about that, so Maryland’s doesn’t exist anymore. And there you got it.

Jeez. Let’s talk about statistics. Mapping Police Violence, and some other media outlets say that there are about a thousand deadly police shootings a year. Do you have any way of knowing if that’s accurate?

Well, thanks to the Guardian and the Washington Post, I think we’ve come closer to reality. We have a crisis in data around policing in this country. I have a long law review article that will come out of the Texas Law Review this coming year that just talks about this policing’s information problem. And the point is, this is an era of big data. We don’t think to regulate anything—I mean the most trivial things—without data; we are drowning in data. And yet around policing, not so much. And there are a number of reasons why that is the case. I will name three, but it’s untenable. If public safety is the thing we care about the most, and people will say that they do, then you’d think that we’d want some information and we’d want to regulate based on that information.

So one of the reasons—which is the one everybody goes to right away—is that the cops want secrecy. Now there’s some truth to that without any doubt, but in my own experience, that’s not the biggest problem. The second is that police departments run with just like—it would appall you—legacy information technology equipment. So they may have fancy surveillance equipment, but when it comes to record keeping and record management, not so good! We worked, for example, with a company Mark43, that is trying hard to change some of that, as are others. But we desperately need the mechanisms for keeping the data. And then the third problem is we’ve got 18,000 police departments in this country. So even if they all said, “We’re going to be really good about keeping the data.” They’d all keep different data because there’d be no uniformity. And you’re not going to get that uniformity unless there are mandates either at the state level or the federal level. And here, the federal government has just failed us all completely. I have some pungent reasons why if you want to hear them, but we just don’t have the data that we need to regulate policing.

And by the way, it’s not just police shootings, which is where people focus. At the Policing Project, we’ve got a transparency checklist, which I think is up on our website. It’s all kinds of information that would be useful. It’s how often do people get stopped and what kinds of surveillance technologies are used, and how productive are those surveillance technologies? For example, police departments are drunk on license plate readers, and I’ve studied them for a long time. And I am still waiting to see the data that tells me that they pay off. We just don’t know; we’re ignorant. We’re doing this blind.

My question with the killings is first to acknowledge: We don’t even know what the real number is, but also, we are in a society that is awash in guns. If you read the editor’s note on the Washington Post project, they cite a very high number of shootings where the police were actively being shot at at the time and returned fire, something around 30 percent. So there has to be some level of—I don’t know what you want to call it—“non-negligent” police shootings. Do you have any guess what that might be, or if we could get the police shootings down to the regrettable, but non-negligent number, what would the delta be between where we are now and what that number would be?

Yeah, that’s a terrific question. I want to preface my answer just by saying, people focus on the shootings. And a thousand is way too many, though it is a very large country. What we need to focus on is use of force all told. That’s the right question. That’s what we ought to be looking at. But if you take any segment of that use of force, any different types of use of force, including shootings, the wrong answer is to say, “Well, the person was armed. So the shooting was OK.” Because that forgets everything we talked about at the beginning here about, even if they were armed, could you have gotten out of the way? Could you have solved the situation in a different way?

A lot of shootings are tragic. They’re what we call “death by cop,” which is somebody that’s suffering some terrible emotional trauma, and they want to end their life, and one way to do it is to threaten cops and get shot. Well, I was at a very poignant presentation about this at the Police Executive Research Forum, and [hearing about] ways to try to avoid that, because that’s not something anybody wants to have happen. And so I know you want, like, a simple answer, like the right answer would be 237. But you’ve got to look at each case and say, “Was there a way out of that situation?” And you were right, we’re a society awash in guns, and it makes a cop’s job difficult and dangerous. But with training and tactics, there are ways to address those problems, not to speak of trying to do something about the guns.

There are often cases where a police department is put under review or the federal government offers oversight, or there is federal or state pattern or practice investigation. And sometimes they do find, maybe through the use of an expert like yourself, that the police are policing in an unconstitutional way. And those unconstitutional tools are taken away from them.

But they aren’t replaced by better tools, and I’ve heard an analysis of Baltimore that essentially argued this has happened there. The police only knew how to do their jobs (and not even that well) by really brutalizing the populace. They’re told they can’t do that, and now what? The economist Roland Fryer has a paper out that looked at pattern and practice investigations and found that when there is an investigation after a viral incident, the police really do pull back. And this is in one sense understandable given human nature and in another sense inexcusable, but he found 893 more homicides in places where there was a viral incident and then an investigation.

What do we do about finding these departments that are doing their jobs the wrong way, but also at least in the interim, giving them enough tools to do it the right way?

First, I think pattern and practice investigations are important. There was also something called collaborative reform that the Trump administration put an end to, which was: A department that wasn’t yet in crisis, but thought they had problems, could actually go to DOJ voluntarily and say, “Can you give us some help?” You know, people need help. And sometimes the departments need help, frankly, from outsiders just to be able to get their municipal government to get them the resources that they need to, for example, pay for the right kind of training. And I’ll also remind you, by the way, if the government’s gone into investigate the fact that you are not meeting constitutional standards—so, we talked about earlier about how low those constitutional standards are. So the question you want to be asking, and that I take you to be asking is, how do we have good police departments? And that’s the right question.

We could talk forever, but I’ll mention two things. One is, the U.K. has got a thing called the College of Policing, and it’s not perfect. I’d be the last to say it’s perfect, but it is a body somewhat independent from the government that involves the police, that sets national standards, and then expects people to be trained and perform to them. And we have nothing like that in this country at all. And in fact, whenever we get standards, they come out of the DOJ, which itself is conflicted, because it’s also a law enforcement body. And so, we need somebody like that. The other is that, we at the Policing Project are developing something that I probably shouldn’t even talk about on the air because it’s too nascent, but here I go. My foot’s in my mouth maybe, but: Call it a health check, which is that we’re trying to develop a set of metrics so that the department itself, the community, the municipal officials can look at their department and answer the question, is this a good department?

And I know you actually talked to Matthew Barge, who works on our team, and it came up one day, because we were sitting around the table and somebody said, “How do you know if something’s a good police department?” And we think we know that some are good, but we’re not sure. Think about that here in 2020: that we just don’t actually have an answer to that question.

Listen to this full conversation between Mike Pesca and Barry Friedman below, and subscribe to the Gist on Apple PodcastsOvercastSpotifyStitcher, or wherever you get your podcasts. Join Slate Plus, and enjoy ad-free episodes of the show.