Bad Science: the Bogus Case for More Cops

In this bonus episode of In Plain English, I talk with defense lawyer Alec Karakatsanis about the paper “The Injustice of Under-policing in America,” by Christopher Lewis and Adaner Usmani. We cover the undisclosed assumptions, missing statistics, and cynical worldview behind this piece of propaganda masked as a scientific paper.

You can read Alec’s response to the paper here:

If you liked this episode, you can reach out to Alec on Twitter @EqualityAlec, or submit questions or comments via the Continue the Conversation tab on our website!

To learn more about this subject, check out Our Enemies in Blue, by Kristian Williams.

Remember to follow In Plain English on Facebook , Twitter, and Instagram to keep up with the latest announcements! If you are able, please consider becoming a Patron to help support In Plain English!

In Plain English also has a Discord server! Join here to chat with other listeners, past experts, and guests about science, suggest new episode topics, and more!

And check out our Where to Listen page to subscribe to our podcast on your favorite platform!

Photo Credit: Grayscale photo of police riot team on pedestrian lane. Original public domain image from Wikimedia Commons

The Injustice of Under-Policing in America

Reply to Alec Karakatsanis (Authors’ Rebuttal)

Episode Transcript

Jamie Moffa: Hello everyone, and welcome to this special episode of In Plain English. I’m your host, Jamie Moffa, and today, instead of talking about a good scientific paper that has rigorous data and well done statistics, we are going to be taking a look at what not to do when you are writing a scientific paper, and what to be on the lookout for if a paper that you’re reading seems to be a little bit suspicious. And today, we are going to be talking about that with Alec Karakatsanis. Alec, welcome to the podcast.

Alec Karakatsanis: Hi, thanks for having me.

Jamie Moffa: Would you like to tell the audience a little bit about yourself and what you do?

AK: Sure. So I’m a civil rights lawyer. I started my career as a public defender, actually, representing people who are accused of crimes, but who are too poor to afford their own attorney. And then in 2013, I got a grant from Harvard Law School to start a nonprofit organization that would try to challenge systemic injustices of the criminal legal system in the U.S. And so since then, I’ve been going around the country and identifying injustices that are systemic and pervasive and horrific. In the U.S., we have the most people in cages of any society in recorded world history. And so there’s no shortage of these devastating and ineffective policies that jail people and separate families all over the country. And our job is to sue cities and counties and states and judges and prosecutors and police officers and parole and probation officers all over the country, private companies that are profiting from this, to try to resensitize our culture to the horrific nature of human caging and also to challenge the way people think about what keeps us safe and what doesn’t. And so this work has often gotten me interested in the ways in which elite academia through, I think, really serious intellectual and evidentiary and empirical and ethical flaws has brutally perpetuated the rise of mass incarceration in the United States.

JM: Yeah, absolutely. And that actually brings us nicely to the paper we’re going to be discussing today.

This paper is entitled “The Injustice of Underpolicing in America” by Christopher Lewis and Adaner Usmani. And these are two professors at Harvard, I believe, who published this paper in the American Journal of Law and Equality. And so Alec, can you tell me a little bit about how you found this article and why you decided to shed some light on it on social media?

AK: I think it’s really important to understand a few basic pieces of context first. So we have about five times the number of human beings in prison as the national average prior to about 1970. So prison rates exploded in the United States after about 200 years of pretty stable rates of human caging. Similarly, we have about five to 10 times the incarceration rate of other comparable countries. And so something happened in the 1970s that led to the explosion in the way that our society thinks about public safety. And it wasn’t that people in the United States became worse people or more evil or more—or committing more crimes or anything like that, right? It was actually that policy changes were made and a lot of people profited from these policy changes. There’s a tremendous amount of money, there are multi-billion dollar industries at every single stage of the criminal punishment bureaucracy, from the people that make the tasers and the guns and the police vehicles and the sirens and the military equipment and the surveillance technology, the people that make the software, the people that profit off the surveillance, the police unions, the bail bond industry, which is multi-billion dollar industry. There’s a multi-billion dollar industry that profits from prison and jail phone calls, multi-billion dollar industry that does medical care inside prisons and jails.

You get the point. There are millions of people that depend on this system for their jobs. There are tens of billions of dollars spent on every single aspect of the system. So there’s a lot of money here. And over the course of the last 20 or 30 years, the federal government, some state governments, and the private sector have spent a tremendous amount of money to influence so-called scientific research around this area. All to sort of create this idea that more investment in police prosecution and prisons keeps us safe.

Now, of course, they have a really, really tall task because we have really high levels of violence in the United States relative to other wealthy countries. And yet we also are spending so much money on police surveillance, repression, human caging, et cetera. And so many people would be very reasonable and having a very basic question. I mean, if money spent on those things made us safer, we would have the safest society in recorded world history, but we don’t. And so there’s been a concerted effort in academia, among a relatively small number of institutions that are getting most of the money, to produce research that justifies and rationalizes and supports these massive expenditures on the bureaucracies of punishment, rather than investments in systems of care, like universal health care or inequality reduction or early childhood education or a lead abatement, et cetera.

I use those examples because in—despite all of the money spent on police and incarceration research, the scientific consensus is relatively robust that each of those other things I mentioned, systems of care, so in general kind of reducing inequality is actually far more associated with reducing levels of violence and harm in society. So in a situation where you’ve got a basically a scientific consensus that investment in reducing inequality and building systems of care actually reduces violence, you’ve got to work extra hard if you are sort of the industries and the unions and the interests that profit from this punishment bureaucracy. You’ve got to work extra hard to try to convince people that all of this stuff is needed, and that’s kind of how I came to this article, because I’ve been studying the body of research in this area and picking it apart and analyzing it. And so whenever a new article like this comes out from the small group of elite academics that are trying to influence journalism and influence the media into talking and thinking a certain way about investments in carceral bureaucracies, people send them to me.

So someone sent me this article and said, can you believe what these people are saying? And I wasn’t prepared for actually how bad it was.

JM: Yeah. So can you go over on a broad overview level, what is the thesis of the authors of this paper? What are they trying to demonstrate?

AK: The thesis of the authors in this article is essentially that the United States has too few police. They make a number of claims, but I think the core of the article is basically saying that the United States has fewer police per capita of other comparable countries, a claim which I think is false. And then second, that police are really good at preventing what the authors call crime. Now, we’ll get into a little bit later how this is such a silly idea because the very concept of crime is socially constructed. And when these authors talk about crime, they’re not talking about crimes committed by the police themselves. They’re not talking about crimes committed by government officials or corporations. They’re not talking about wage theft or tax evasion. They’re not talking about a whole wide range of crimes committed by the wealthy. But anyway, their main point is that police reduce crime, a claim that I also think is false. And so because of those two things, because we have so few police compared to other places and because we have—because police are so good at preventing crime, we should have many, many more of them.

Then they make a series of laughable sort of what they view as back of the envelope calculations. They see this article as more of a think piece. So they deliberately say in the article that they’re not trying to fully justify their claims, which I think is really odd. They make a series of back of the envelope calculations and sort of conclude that, you know what, we need maybe 500,000 more cops. And to put that into perspective, the greatest expansion of policing in modern US history was the notorious 1994 crime bill, which Joe Biden got heavily criticized for in the campaign trail in the Democratic primaries when he was running for office, and which Hillary Clinton was, you know, widely reviled for in circles of study of these issues and in communities that were impacted by the incredible introduction of 100,000 new cops in 1994. And this would be five times that, right? This would be the greatest expansion of police repression in the history of the world. And these authors sort of casually offer that as like the only solution that’s feasible to reduce crime in the United States. It’s a really stunning, really stunning set of claims.

JM: Yeah. So let’s get into it by first talking about you mentioned that they claim that the United States has, you know, fewer police per capita than the majority of developed countries. And this gets into a bit of what they’re trying to display on figure one. And you had dug up the statistics that they used for that. So can you talk a little bit about where their number for the number of police in the United States is coming from?

AK: Yeah, I mean, I think there’s a lot of dispute about how to count police officers. One of the things I think is most dangerous when doing scientific writing and scientific research is making a series of assumptions, but not describing to your readers what assumptions you’re making. It makes it very, very difficult to evaluate the strength of someone’s claims when you don’t know what kinds of discretionary assumptions they’re making. So here in this example, they didn’t acknowledge to readers that this is a widely disputed and very, very difficult thing to measure because different countries have different kinds of jobs that they call police officers. So for example, in some countries, they call police officers what we would call in this country, you know, border patrol. Or they think of police—they have a lot of people labeled as police officers that we would think of as civilian employees of police departments, or they sort of like combine their police and their militaries in certain ways to make it hard to count. And there’s just bad data in other countries. So it’s a very, very difficult thing to measure. Similarly, in the United States, it’s very, very difficult to measure because our surveys for collecting this information are sort of not very comprehensive.

So I was immediately skeptical when I saw them claim that the United States has way fewer cops. I mean, for one thing, we spend a lot more money on police than any other country per capita. But for another thing, I knew that some of the data sources they were using had all of these limitations. And it was a huge red flag for me that there wasn’t even a footnote, let alone a section of the article where they describe the data limitations and explain that this idea of counting these cops is in any way uncertain or controversial.

So I actually sent them an email, the authors, and I asked them sort of which data source you’re using and why basically, right? And one of the professors responded that they chose to use the number 697,195, which is from the Uniform Crime Report, which is an FBI reporting survey. Now, this was really alarming to me because anyone who studies these issues, including these authors, knows that the Uniform Crime Report excludes many, many, many local police departments around the country who just don’t send them the data. So right off the bat, we know there’s a really large exclusion going on. So they’re deflating the number of cops in the United States to make it seem like we don’t have as many as we do. He admitted to me that the number may be much higher, like 900,000, right? So from just under 700,000 to 900,000. Even if you were to go on Wikipedia, for example, Wikipedia says there’s about 900,000 cops in the US based on a source from a police nonprofit organization. So even if you were to do basic, just like internet googling, you would see that there’s some real discrepancy. That’s 200,000 person difference in the numbers.

Then the professors admitted privately to me over email, again, not included in their actual published article, that the number of US police in the census count, which there are a lot of reasons to think that the census count is the most accurate because this is a survey that has the most sophisticated methodology and the most comprehensive methods, and so this is the number of people who self-report as their job being a police officer. And that was actually 1.2 million, a little bit more than that, 1,227,788. So that’s 76% higher than the number they chose to use in their public article, right? And 500,000 people higher than what they said. So actually, it’s kind of ironic, you know, if you, if they call for 500,000 more cops and the number of cops is actually 500,000 more than they said there were, then really they’re calling for no more cops.

JM: Yeah, wrap it up. We’re done. We did it.

AK: You know, then once it became clear that like someone was on to what they were doing, they switched gears in our email exchange in a way that I found even more troubling. So now recall like the very, you know, beginning of the article emphasizes, and I think this is a very strong and sexy rhetorical point for them, that, you know, all these other countries around the world, they have more cops than the US. And maybe that explains, you know, why they have less crime. They then shifted away from that. They said, yeah, okay, it doesn’t matter if we, if we’re wrong about that, and if, in fact, the US has more cops per capita than all these other countries. What really matters is something else. And this is where it got really confusing and really weird. They said instead what matters is the number of police per prisoner.

Well, let me, before I get into that, let me just say one more thing. I’m sorry. They also admitted to me that they excluded privately paid police officers from their calculation. This is a really, really big red flag. So when you’re talking about the extent to which the US is a heavily police society, you cannot exclude private police because, because of the recent boom in privatized police forces, you know, for example, Harvard University or the University of Chicago, or look at downtown Detroit, which is basically patrolled now by privatized police owned by a local billionaire. They didn’t even tell readers they were excluding privatized police forces, let alone explain why it would be a good choice to exclude privatized police forces, let alone explain how that affects the relative calculations of the number of police in different countries. And by estimates that I’ve been able to find through investigative journalism, there are about 1.1 million private police officers in the US, almost the same number of public police officers, right? So now we’re talking about multiplying by a factor of about four, the original figure that they gave in terms of the number of police in the United States. I mean, it’s all just so laughable. It’s laughable if it wasn’t being written by two up and coming elite Harvard professors who are publishing in a what sounds like a major journal. Now, I’ll note that this journal was actually created at Harvard and published through MIT as part of the university’s public response to the murder of George Floyd. So it’s absolutely ironic that within two years of creating this journal that was supposed to be about equality and a response to horrific police violence, you have shoddy scholarship by pro police professors calling for 500,000 more police. I mean, it’s really just a tragic tale of, of how the interests that kind of underlie a lot of the academic research in this area sort of manipulate and for years and years and years play a very important role in socializing our culture and normalizing the idea that police keep us safe when they don’t.

JM: Yeah. So then can you talk about why now are they looking at police per prisoner and what that does for their argument?

AK: Right. They basically said, look, we’re using per capita comparisons, although what really matters is, is police per prisoner and police per homicide. They said that if you actually look at the amount of police per prisoner, the US has really, really low numbers of police. I pointed out, well, yeah, because we have a lot of prisoners. So if you look at the number of police per prisoner, you’re artificially inflating the point that there’s not enough police by not acknowledging that we have the greatest incarceration rates in modern history and we have really, really long sentences. So all you’re telling people is that we have really long sentences, which everyone already knows.

And in fact, one interesting thing about this paper is these two professors, this is a tactic, a political tactic you see a lot in the pro police scholarship now, they concede that the research is very, very clear that prisons don’t make us safer. So one really interesting part of this article, which we’ll get to later, it’s one of the worst parts of the article because it makes no sense, is that in order to pay for 500,000 more police, we should reduce the number of prisoners from like around 2 million to about 300,000. So they want the greatest decarceration in US history. They claim to want that. I’ll get into a minute why it’s so nefarious. They say that because prisons don’t work and because the US has dramatically expanded incarceration and there’s no scientific evidence that it helps, we should dramatically reduce incarceration as part of dramatically increasing police.

Now, what they don’t explain, we should talk about a little bit later, is one of the main reasons we increased incarceration is lobbying from police. And these police organizations are so strong and so powerful. Even as it stands now, we stand no chance of dramatically decreasing incarceration rates when people have been trying for many years. If you dramatically increase the power of police and police unions and the companies that profit off of them, it will make it even harder to ever reduce. And there’ll be much more pressure to increase incarceration because all of these bureaucracies and industries are tied up together. So they never address—they claim that their whole proposal is based on political feasibility, but they never address the most important point from most of the literature on this, which is that it is the increase in power of police themselves that is actually one of the main factors in increasing incarceration.

So anyway, they say that police per homicide is much more important as well. And that is also just really silly. It’s not—it’s just measuring the fact that the United States has a lot of homicide. It’s not telling you anything about the appropriate number of police in the society. So there’s a lot more that they do in there that’s very, very strange. But I think that’s the basic setup of the article. And that’s the foundation that is laid for what they do later, which is, I think, way more ugly and way more significant than even the mistakes they made in that initial setup.

JM: Yeah. And just to point out, one of the first things I noticed when I was looking at these figures was that they’re not really using any statistical methodology at all. So in figure one and figure two, they’ve done essentially the same thing except for in figure one, they’re looking at police and prisoners per 100,000 people. And in figure two, they’re looking at police and prisoners per homicide. And so what they’ve done, to explain to our listeners, is that they’ve plotted police per 100,000 on the X axis, the horizontal axis, and prisoners per 100,000 on the vertical axis. And then they’ve taken a bunch of different developed countries and they’ve plotted them. How many police do they have? How many prisoners do they have? And the majority of the countries fall along this line, and they’ve drawn a line. If they had done actual statistics, which they haven’t, this would be called a regression. And what you would be able to glean from this is if you have about 250 police per 100,000, you would expect to have about 125 prisoners per 100,000.

And then they plotted the United States of America, which unsurprisingly has many, many, many more prisoners per capita. And based on their bunk numbers has about middling police per capita. And so this point looks like an outlier. And so usually you’d try to then spend time explaining why it’s an outlier. Instead, they have drawn a separate line connecting zero comma zero to the point for the United States of America, which has no scientific meaning whatsoever. And they’ve done this for both of these graphs. So I also wanted to point out that that’s completely not even using statistics in any meaningful way.

AK: It’s really remarkable. And it goes to show something that I’ve learned when I’ve studied this area is some of the most shoddy and unethical research practices. And also a big problem in this area is research that is perfectly fine as far as it goes, but claims that are made about what the research says are often distorted. If you’re doing it in a way that supports powerful institutions, and there’s no institution more powerful in the US than the police, military, and other forms of sort of repression that get a tremendous amount of state and corporate and philanthropic investment, if you do that, then your research is judged by different standards of rigor, different ethical standards. An article like this appearing in a major journal that was making the opposite claim, that the police should be abolished, for example, would be held to much higher ethical standards. People would be picking it apart. And the scholars that did it would have to be extremely careful and meticulous with every little thing they did because it would get so much more scrutiny. And I think that’s just a general point about my evaluation of hundreds of studies in this area. If you’re reporting a result that is consistent with what the powers that be want, you get less scrutiny. And if you’re reporting a result that is different from what the powers that be want, broadly speaking, I’m being a little bit crude about it now, but generally speaking, your paper gets less scrutiny.

JM: And that makes it easy to kind of get some of these not even Statistics 101 things through and still publish. So they make this claim then about there’s this what they call a first world, first world balance between police and prisoners. And they spend the rest of the article trying to justify that. So can you talk a little bit about how they try to justify that and what quote-unquote statistics they use?

AK: Yeah, keep in mind that all these sexy terms they use, like first world balance, they admitted over email, this is essentially false, right? Because we’ve now admitted that actually the US might have more cops than other first world countries. So when it’s talking about getting the US to the first world balance, it’s all just totally false. So we really at this point abandoned a lot of the marketing and the claims that began the articles. But in any event, the core of their article still stands if even if you strip away some of the, the way they’ve marketed the article as being about how the US compares to other countries, the core point of their article is really, and this is where it gets just astonishingly bad.

They purport to find that the costs of adding more police are significantly and overwhelmingly lower than the benefits of adding all of these cops. So what they admit to me over email is essentially we don’t need the international comparisons. You know, the real core of our paper is that because police are so good and because they do so much good and because they don’t have all these costs, the cost benefit analysis shows that we need way more cops. And that’s really the intellectual core of the article. The rest of it is just window dressing.

And I think the first big thing to say about that is it’s total nonsense. You can tell it’s total nonsense by the fact that, when they go through the costs of policing, they leave out virtually every major cost of policing. It’s really incredible. I mean it’s incredible to me that not only would a journal publish this, but that other academics looking at this wouldn’t be offended and wouldn’t be worried about the integrity of their work. Other sociologists at Harvard, other people at Harvard Law School, you know, one of them is a sociologist, one of them is a law professor, wouldn’t be outraged at the way in which scholarship like this really undermines all of the scholarship that people are doing at these institutions and in these journals.

So what they, what they say is the only cost of police that they identified is they predict that adding 500,000 more cops would result in 7.8 million more arrests. They call this the quote main downside end quote. So what we’re saying here, they’re saying look yeah our proposal would be, you know, it would have some costs. Almost 8 million more people every year would be arrested. But, and again, when they talk about this arrest, they don’t talk about infectious disease in jails, they don’t talk about family separation, they don’t talk about convictions, sentences to prison, they don’t talk about coerced guilty pleas, they don’t even talk about any of the attendant, they don’t talk about strip searches, they don’t talk about any of the attendant harms of an arrest.

JM: It’s violent just to be arrested like and then there’s all these other violences that are stacked on top of that.

AK: Absolutely. And one amazing thing is that you’ll notice if you read the article or you read my article about their article in my newsletter, which is called “A Warning to Journalists about Elite Academics.” At every single stage of what we’re about to talk about, there is an entire academic and scientific literature that actually does quantify these things in many respects, and they’ve just ignored it. And this is another I think really important lesson about doing social science, you have to engage in the relevant bodies of research. When you don’t do that, and you just assert things as comprehensive, it’s really fraudulent because you mislead people into thinking that that there isn’t a wealth of knowledge about some of these other issues.

So why don’t I go through just really quickly some of the costs that they leave out. I want to start at a really high level. So there’s a tremendous historical, economic, and political science literature about the cost of policing that—and again like some of the stuff you may not agree with, because some of this stuff is, you know, I have a particular view of the police that might be different from yours and I encourage you to read some of my other work on this where I’ve set forth my views more. But there is a whole literature out there about how one of the central roles that police have played for the last 150 years is in preserving inequality and then blocking investments in social welfare. You don’t even have to go as far back as when modern police forces in the U.S. really derived from old slave patrols in the south. And in the northern and western more industrial cities, they actually evolved from the interests of private industrial conglomerates in crushing union organizing. That was really the development of what we think of now as the modern police force. It was really designed to crush union labor organizing and to support local political machines by political bosses and it sort of, it changed its marketing in the mid-20th century to be about public safety. But really the origins of the United States policing system is really in preserving inequality. Whether it’s racial and gender inequality or economic inequality.

And if you look back—and that hasn’t changed even though the marketing of police has changed. There’s really great historical research showing how essentially every single social movement for progressive social change, from seeking the right to vote for women, to LGBTQ struggles throughout the second half of the 20th century, to labor organizing, to immigration, to environmental protests, now to the abortion movement, etc. You name it, the police have infiltrated, surveilled and crushed it. A huge percentage of police budgets over the course of the civil rights movement, for example, over the course of of the current sort of modern professionalization of police. A huge percentage of their surveillance and policing budgets goes towards tracking left-wing movements that are calling for social justice. It’s just always has been a major function of police. And they don’t even acknowledge, they don’t acknowledge that the police murdered Fred Hampton, they don’t acknowledge the police infiltrated and tried to destroy Martin Luther King, that they infiltrated and destroyed Malcolm X, they don’t acknowledge the role that police play—playing right now in destroying environmental protest and protests for abortion, etc. I use some examples in my article.

It’s pretty wild to me that an article calling for many more police would ignore vast swaths of literature in every relevant field from sociology to history to economics to public health, etc. that police actually increase inequality by destroying and targeting social movements.

Another really, really important point is currently right now in most American cities, police and police unions spend a tremendous amount of money lobbying local officials to block the creation of public health measures that have actually much stronger connection to reducing violence and harm, like violence interruption programs, needle exchanges, safe injection sites, access to permanent supportive housing for people that are experiencing homelessness, after-school programs, hiring more teachers, you name it, right? So actually when you look at, and this is something that I do in my day job, I’m actually involved in these fights all over the country where doctors and nurses and teachers are asking for more money for provenly effective early childhood education programs and drug addiction intervention services and the police are blocking it. And the police are actually instead using their political power to get more money for surveillance programs, etc. So just ignoring all that is a really odd thing to do.

And then I think at a slightly lower level of generality, the police are and always have been central to gentrification, redlining, evictions, immigration enforcement, civil forfeiture where they steal the property of, of people who are accused of drug offenses, another sort of low-level crime, and generally the depletion of wealth in poor communities. All over the country right now police are playing a central role in enforcing abortion restrictions and investigating and enforcing new state laws in places like Texas, investigating the parents of trans children for giving them medical care. So I guess my point here is that how can you write an article about how our society needs 500,000 more police without even mentioning that the only way that the new abortion restrictions all over the country are going to be enforced is through more investment in police. And not even mentioning that as a cost shows just how narrow the paper is thinking about the cost of policing. In every single place that I’ve worked in my entire civil rights career, the police have organized to oppose what we’re doing, whether it’s bail reform to get people out of jail because they can’t afford to pay, whether it’s fines and fees reform which, you know, we sued the Ferguson police department etc. for running a debtor’s prison for years where the police would just arrest people tell them to pay them a few hundred dollars cash, and if they didn’t they would be stuck in jail. And we’re talking about several hundred thousand people a year in this country jailed by the police just because they can’t pay cash on low-level debts. None of this stuff is mentioned at all in this article’s discussion of the various costs of police.

Another point I think that is just worth hitting at the really you know high level is that they claim that more police would not lead to more police violence, and this is a really incredible part of it. They cite nothing to support it but they say something along the lines of, well if one of the reasons that there is so much police violence is that policing is really dangerous, and maybe if we added more police and we reduced the number of the amount of violence in our society, there’d be less need for police violence and we actually might see a reduction in police murders. The casual nature of the really bold claims they’re making is astonishing, but again note when they’re talking about police murders they’re leaving out all of the other forms of police violence. Never mind the fact that 7.8 million more arrests are violent. You know, an arrest meets the legal elements of a kidnapping. The only reason that it’s lawful is if there’s probable cause that someone has committed a crime, but the act of arresting someone is essentially the same physical act as kidnapping. So what they’re what they’re really saying is 7.8 million more kidnappings are justifiable, even though you know around the country a huge percentage of arrests actually lack probable cause so we’re talking about like actual kidnappings here.

Forget about the arrests think about the many many more, there’s orders of magnitude more police stops and beatings and taserings and shootings then there are arrests, right, so for every arrest there are several more times when children are stopped and roughed up a little bit, searched on their way to school, or people are stopped by the police in really tense and humiliating and horrific experiences. None of that is even mentioned in the article as a—they don’t even mention police stops and searches as a cost, let alone civil forfeiture or police take those property millions of times a year in those stops etc. I’ll stop there because I could keep going on and on. I think you get the point that they don’t talk about almost any of the major costs to policing.

JM: And the last point that you were making about them claiming that more police would lead to fewer police killings: they try to justify that in figure five, where again they’ve made this scatter plot, and on the x axis they put police per homicide, which is already an issue because again like you said earlier, dividing by the number of homicides is just going to be a measure of how many homicides there are. If there are more homicides, you’re going to have fewer police per homicide, that just makes sense. And then on the y axis they’ve put police killings per million people. And they’ve plotted all of these countries, and showing that there’s a negative linear regression—they haven’t even actually done a linear regression, but they’re, they’re trying to show that if you have more police per homicide then you have fewer police killings per million, which I don’t even know what kind of meaningful number that works out to. And again all they’re doing is just demonstrating that if you have more homicides, all they’re doing is dividing by the number of homicides to mask the actual trend, which is—would make much more sense that if you have more police you probably have more police killings because there’s more police doing the killing.

AK: It’s really incredible. And, and I guess one more point here is that, note at every single turn here they’re making assumptions and decisions that skew their results, so by talking about homicides and not total deaths, they’re able to ignore a wide-ranging literature about how more police actually leads to more death. So for example, there’s really good literature about how police violence reduces the birth weight of black children in the U.S., right? Extremely robust scientific literature about various public health effects of policing on Black communities totally ignored in this article. And by talking about police-recorded homicides instead of say, premature deaths, they’re able to kind of elide all of that, that, that literature and research.

JM: And then, so they also, on the flip side they try to justify, not even really the benefits of policing, they try to justify or quantify the cons of other forms of violence in society, and then draw a tenuous link between more police and less crimes. So can you talk a little bit about that side of their argument?

AK: Yeah absolutely. Let me just say one more thing too, because this is something that I—this is one of the main things that I study, which is: another thing they do not mention is the role that hundreds of millions of dollars of police funding play in public relations and in changing the way our society thinks. So the example I use in my article is that the sheriff’s department in Los Angeles county has itself 42 full-time public relations employees that are cops, okay. So that’s just one department in one place. The Los Angeles police have another 25. So just those two departments in one city—and there’s, there’s dozens more police departments in Los Angeles county. And you know, Beverly Hills has his own police department etc. Just those two places have 68 cops that are focused on public relations. They shape every media story: you know when you see—turn like local news every night and you see you know five different crimes committed by four people, often people of color, and mug shots in every story? That’s because police are spending a tremendous amount of money flooding the news with certain kinds of stories about certain kinds of crimes by certain kinds of people. They are distorting all over the country, with hundreds of millions of dollars and thousands of police officers, they’re distorting on social media, they’re distorting in traditional news, print media and tv media, they’re distorting what we think about safety. They’re not stories about the much more significant link between eviction and death, or between lack of access to health care and premature death etc. They’re portraying a very particular worldview that is actually distorting our society’s perceptions of safety. And that’s one of the big reasons our society is so bad at public policy that actually promotes health, safety, and well-being, is literally because there’s so many cops around the country that are, that are shaping the news every single night. That’s another big cost they ignore.

And local governments all around the country, when I talk to officials, they tell me: we know that the evidence says that we should open up a new wing of this hospital, or we know we should do permanent supportive housing for our homeless population because that’s actually going to reduce all of these problems that people are complaining about with respect to so-called crime etc., but the police won’t let us do that, and they’ve threatened us. And if you look just in the last few months, there’s been major police raids of the homes of progressive politicians around the country who are proposing reducing police budgets. There’s, there’s a culture of retaliation that is really, it’s hard for me to describe to people that don’t deal a lot with local politicians, but all over the country local politicians are absolutely frightened about police. It’s one of the most significant threats imaginable to our democracy.

And I think this article is being written and calling for 500,000 more police at a time of rising fascism in the United States, where I don’t think a lot of people really appreciate how close we are as a country to losing even the semblance of democracy with these very powerful bureaucratic interests that are extremely violent, but also extremely organized and extremely effective. And they change the nature of governance because people are afraid of going against them. They know that if they piss off the police union and the police department and even city, and the companies that make billions of dollars from contracts with the police etc., they have a really, really strong chance of losing their next election. And so even when they’re not being physically threatened, which happens a lot, they are repeatedly and constantly and continuously not doing what they and their constituents think and know is right, and what’s supported by the actual scientific evidence, because of the political power of these forces.

Now I think I—I’ll turn now to the other side of the equation, because I think this is another area where the lack of ethics in this scholarship is astounding, but they then claim that, okay having said that the costs of police are really minimal, only 7.8 million more arrests, the benefits of police are extraordinary. This is really the core of their paper, right, because if, for example, it was true that more police didn’t really reduce what they call crime, then the whole point of their paper would be meaningless. This entire claim is based on the scholarship of like one or two other professors who are, who work at the police or very pro-police, who find that increasing police reduces certain kinds of crime. These professors don’t note that even in that other pro-police research, those professors find that it’s actually not the case in certain kinds of cities: cities that are majority black for example, and it’s actually only certain kinds of crimes. Whereas it also increases illegal arrests and illegal stop and frisks and other things that actually increase violence in society. So it’s actually way messier than these professors acknowledge.

But even if you take their point at, at sort of face value, but they claim I think that adding 500,000 cops would lead to 4,000 fewer homicides. So what we’re talking about, the main benefit that we’re going to do all this stuff, we’re going to increase the political power of police, we’re going to add the most police that any society’s ever added in the world history, we’re going to get 7.8 million more arrests of people, and all of that is being done to reduce homicides by 4,000.

Now the science that goes into this is pathetic. The papers that they’re talking about don’t even themselves argue that the relationship between adding more cops is linear. There’s no evidence that like if adding 1,000 more cops reduced, you know, three homicides, that adding 10,000 more cops would reduce 30 homicides. Like there’s no reason to believe that this, even if you accepted this relationship, there’s no reason to believe that it’s linear. Just the level of shoddiness in this is pretty remarkable.

So we have about 20,000 homicides a year in the US, so what they’re saying is we’re going to reduce the number of homicides by about 20% and essentially creating a police state as a result. What they say is, this benefit is so clear [laughter] that we have to do it. And I think that’s just even—so I guess like another theme of the article is, even if you take all of their claims at face value, they’re kind of laughable, you know? If you think about a little bit, which they’re hoping people don’t do.

JM: And all of this also not to mention that there’s no reason to believe that these associations are causal. Even again if they were, if these numbers were vaguely accurate, there’s no reason to believe that having more police causes fewer homicides. Like you need to do all sorts of other scholarship that they don’t do in order to prove that this isn’t just a link that’s being mediated by some other outside factor.

AK: You know, another point that is really important to make about that is that—the point I was making a few minutes ago about non-homicide death. So they don’t even acknowledge that there actually is literature about the causal nature of more police and certain forms of non-homicide death. And not just death but sort of, illness and, you know. For example like the studies that I mentioned before about birth weight of Black children. There’s a lot of other literature about the causal nature of more police and more police arrests, and more police stops. And the—there’s a lot of really good literature about how arrests more—or low level arrests and short periods of jailing, there’s actually some natural experiments that have been done on hundreds of thousands of cases that show that by arresting low level people and keeping them in jail for even a few days, you actually increase their likelihood in the future to commit crime. You actually reduce their future lifetime earnings by measurable amounts. So there’s all this literature out there, when we’re talking about the benefits, is just totally ignored. And so a lot of these benefits even that they’re talking about are actually outweighed by other literature that they ignore.

And it’s important to take a step back just for a second and say, recall that like these—one of the things that makes this paper so dangerous, and I think one of the reasons that it, that it got such little scrutiny by the people that published it, and why it was initially celebrated in a lot of academic circles, you know, not—not because of the academic rigor, but because of it, because it, it was thrust into a national debate at a time when Joe Biden is asking for 100,000 new cops. Democrats in congress are pushing for this. These two professors present themselves as progressives. That’s very very key to the harm that they’re doing, right? They are self-consciously, even in this article, claiming a certain politics. What they’re trying to do is provide a progressive flanking, like a—they’re trying to make it okay for self-described progressive politicians, like many members of the Democratic party that are more sort of pro-police and more pro-corporate, they’re trying to provide an intellectual framework and justification and sort of rationale and political cover, essentially, for those people to do things that are deeply regressive. By calling themselves progressive, they get less scrutiny from people on the left who might otherwise find this horrific.

And, and that, I think, is why they inserted this meaningless stuff into their paper about how, well actually prisons are horrible, and jails are horrible, and so we actually want to reduce the number of people incarcerated while increasing police surveillance and police oppression. Some of the same group of people have admitted in recent weeks, you know well after this article was passed, if you actually look at what people are saying when they’re lobbying in congress. And if you talk to some of these people privately, they’ll admit that we don’t have the political power to reduce the prison population at all. So we should do is first, first we should dramatically increase the police, and then if these scholars are right it’ll be easier to reduce the prison population, because crime will have gone down so much.

And I want you to, I want you to understand just how disingenuous and, and just, and these people know exactly what they’re doing. These professors who are calling for more police deliberately inserted into their article so-called, you know, progressive-looking points about how we should dramatically increase—decrease incarceration, precisely because they know that that is impossible. And the people that are going to use their article to argue for more police are going to get those more police, and we’re never going to get the decrease in incarceration. And that’s why they feel so comfortable in inserting that point. We had crime reductions for about 30 years, and prison populations increasing almost every year. We know that just lower amounts of crime are not going to reduce prison populations, because the prison populations are driven by other interests and other factors.

And so I think that’s something that makes this article just so, so harmful is it’s, it’s not scientific scholarship being done in a vacuum. And this is another big lesson for doing science, and social science in particular as well. Your claims and, and research that you make, exists in a political context. And when you’re a professor at Harvard you’re very, very aware of the political context because it helps determine whether you get tenure, it helps determine whether you get philanthropic donations that get you a named professorship, it helps determine whether you’re invited to be on certain commissions and be in certain rooms. This is one of the most profound threats to good science, I think, is the way in which many of the most successful people in academia are able to use politics and not the, the nature of their and quality of their research to advance their careers. And one of the easiest ways to advance your career in the United States in academia is by producing research that helps the military or the police.

JM: And then in the article as well, they also try to address the, well if there are problems with more police because there’ll be more arrests, and there are problems with prisons, well then why don’t we just do massive social welfare spending? So can you talk a little bit about the points they try to advance to say why they think that we can’t just do that?

AK: Yeah this is just one of the most laughable parts of their paper. I mean they, they claim that, yes it’s true that reducing inequality and investing in reducing concentrated poverty, they call concentrated disadvantage, and really we’re talking about things like health care and housing and inequality and et cetera right? Yes it’s true that those things actually are the main driver of violence, but it’s politically infeasible in the United States to reduce inequality, and given that it’s politically infeasible in the United States to reduce inequality, we’re going to take that as a given. So our research question that we’re choosing to answer—this is a really great lesson for young social scientists in terms of like, how you constrain yourself and how that task dependency sort of leads to, to certain outcomes—because reducing inequality in our society is impossible, we’re going to write a paper about increasing police repression, because given that we can’t reduce inequality the only other way to reduce crime is to add more police and to create a police state within a society that is a fixed level of inequality.

This is profoundly cynical. I mean, imagine if during the civil rights movement, researchers at Harvard had said, you know there’s no chance of reducing segregation or inequality in the United States, so instead what we’re going to do is, how are we going to use the police to manage segregation? How do we hire more cops to enforce segregation, because we have no hope of actually ending segregation? I mean it’s, it’s just such depoliticizing nonsense for young people to see, scholars taking as an assumption that we can’t create a more equal world.

This isn’t a scientific critique of their article, it’s a moral one, and it’s one of strategy actually, because these people claim to be strategic leftists. I mean one of these two guys is actually on the board of Jacobin, the socialist magazine’s academic journal called Catalyst. These are people that, that purport to be interested in strategic political analysis and calculus and social movements etc., and instead what they’re basically saying is: there’s no hope, and because there’s no hope we have to do the next best thing which is more cops. And you know, setting aside all of the other flaws in, in the article, that to me is one of the most sad aspects of this entire scandal, which is: it makes me really sad to think that there are students taking classes from these people. And I got DMs from their students actually, and I got messages from their colleagues at Harvard, all of them were really appalled at this. And not just at the, the ethical lapses in their, their work, but in the message they’re sending to young people, which is I think one of the saddest aspects of this.

JM: I actually think that’s a pretty good place to leave it. I wanted to ask, you know, since this is such bad scholarship, where are some places that you would recommend people who are interested in scholarship on crime and policing and public safety in America, where are some more reputable places that they could look?

AK: That’s a good question. There’s a couple of structural problems here, right, which is that crime data is kind of owned and controlled by police, and so in order to get that data you have to work with the cops. So there’s some—and it’s also expensive to do some of this research. And so the only people paying for this are the government, and police, and in some cases large companies. So there’s a lot of bad scholarship out there. There’s, there’s not a lot of hubs where you could go and see all the great scholarship, but there are a number of really good academics who are doing really good scholarship in this area. Ruth Wilson Gilmore, Alex Vitale, there’s really really impressive stuff happening at the University of Chicago, a professor named Robert Vargas, there’s a fantastic professor at Princeton named Naomi Murakawa. These are just like individual people who are doing really good research in this area. And then there’s many many hundreds of other incredible scholars that are producing one-off research, you know like research showing that actually more police doesn’t increase public safety, research showing that more police, you know, carries with it many different harms, like the researchers I mentioned earlier who studied the effect of short periods of jailing on public safety: Paul Heaton and Megan Stevenson.

So anyway there’s lots of people but, but I’m not aware yet—and this is something that we are talking about across the academic field of building—but I’m not aware yet of like a hub where people can go to see all the good research in this area. But one place to start, I think, is a book called Our Enemies in Blue, which is a, I think the most fun and easily accessible and comprehensive and well sourced history of US police. And that’s a really good place to start because I think, as a scientist or social scientist coming into this area, it’s really important to understand the political context first. And so Our Enemies in Blue I think is a really really good introduction.

JM: Thank you so much for joining me for this discussion. Where can our listeners find you on social media?

AK: You can find me @EqualityAlec on Twitter. I also write a newsletter about copagana and the roles of media and changing how we think about safety. And that’s where I wrote this essay on this particular article, and that’s Alec’s Copaganda Newsletter at Substack. And those are the main places where you can find me online, and of course in the civil rights organization that I founded, it’s called Civil Rights Corps, and you can find them on our website, or on social media @CivRightsCorps.

JM: Great well thanks so much for being on this podcast.

AK: Thank you so much for having me.

JM: A few weeks after our conversation, Alec reached back out to me to let me know that the authors of “The Injustice of Underpolicing in America,” Christopher Lewis and Adaner Usmani, were circulating a response letter to his original critique via a private ropbox link. Alec sent me this letter, which will be uploaded to the website alongside the original paper and Alec’s response to it. The response letter engages with a few, but hardly all, of the substantive critiques that have been levied against the author’s paper. After a condescending introduction that dismisses most of Alec’s critiques as unserious, Chris and Adaner take a few pages to justify their choice of estimate of the number of police in the US, concluding that no matter what number they use their argument stands. They do not acknowledge the fact that this justification ought to have been included in the original paper rather than in a privately circulated letter. I find this lack of transparency a far more serious example of academic misconduct than their use of a lowball estimate.

They go on to dismiss concerns about their dubious use of the police per homicide and prisoners per homicide statistics, brushing over the obvious issue with this methodology, namely that dividing the number of police and number of prisoners by the high homicide rate in the US makes both the number of police and prisoners look artificially small. They hardly even engage with the criticism that they have failed to account for the majority of the actual costs of policing, repeating the line that there are also costs associated with crime.

Much more concerning than any of this, their response does not even touch on the serious methodological issues present in their paper. They do not discuss why they failed to conduct a basic linear regression on their data, why they simply drew a line over the graph instead, and why they drew a second line just to the data point for the US. They do not discuss, for that matter, why they failed to conduct any statistical tests on their data at all. They do not address their elementary mistake of routinely confusing correlation for causation, and for failing to provide any scientific justification for why more police would directly lead to, for example, fewer police killings. Nor do they answer for their failure to cite the relevant academic literature in the fields of public policy, sociology, criminology, etc.

They end their letter with the following sentence: “Students today too often graduate from college and graduate schools unable to engage with people who disagree with them in good faith, or even to make the relevant counter-arguments. Instead they look to institutions to censor the views which make them uncomfortable.” This kind of non-argument is a tired refrain from those who have power and are accustomed to using it, wielded against those who are trying to hold them accountable, or at least to hold them to the standards the rest of us must reach. Rather than hiding behind this sentiment, I invite the authors to engage with the extremely relevant counter arguments and critiques that I’ve posed in this episode.

Thank you for listening to this special episode of In Plain English. My guest was Alec Karakatsanis, and we’ve been discussing the paper “The Injustice of Underpolicing in America” by Christopher Lewis and Adaner Usmani. You can find the paper, Alec’s response, and the author’s rebuttal on our website at You can also find the transcript for this episode on the website.

In Plain English now also has a Discord channel where you can chat directly with me and several past experts and guests, ask questions about science, and propose future episode topics. You can find the Discord link on our website as well. You can also follow us on Twitter, Facebook, and Instagram @plainenglishsci, that’s P-L-A-I-N-E-N-G-L-I-S-H-S-C-I. Make sure to subscribe to In Plain English on Google Podcasts, Spotify, or wherever you get your podcasts so that you never miss an episode. And you can become a supporter of this podcast on Patreon.

Thanks again for listening and tune in soon for another episode of In Plain English.

[Outtro music]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s