Search the Site

Episode Transcript

As economists go, Jeff Weaver knows that sometimes he isn’t terribly original.

WEAVER: A lot of people have been interested in this question trying to understand how does past incarceration affect, health and mortality status.

There’s a good reason it’s a popular research topic. The United States has the highest incarceration rate in the world, according to the nonprofit Prison Policy Initiative. A report they released earlier this year indicated that there are almost 2 million people in jails and prisons in this country. Most research suggests that, on the whole, incarceration is bad for prisoners’ health, and their mortality. It increases their risk for infectious diseases, mental health conditions, cancer, and violence-related injuries.

WEAVER: But it’s not clear — is that due to incarceration? Is that due to other risk factors that might affect mortality?

Jeff and his colleagues started to dig into this question themselves using data on prisoners from the state of Ohio.

WEAVER: We can see that even prior to incarceration, the people who will eventually be incarcerated exhibit much more risky behaviors.

That wasn’t the only thing they saw.

WEAVER: I can just give you one number. What we find is that people who are incarcerated for a year have about a 15 percent lower likelihood of having died after five years. And so, this is a pretty substantial effect

Their findings indicate that long-run survival, over five years, is actually higher among people who’ve been incarcerated, compared to similar individuals who never were. This was surprising.

Could incarceration be helping some people live longer?

From the Freakonomics Radio Network, this is Freakonomics, M.D. I’m Bapu Jena. Today on the show: how does incarceration impact a person’s health—and how does a person’s health impact their risk for incarceration? First, Jeff Weaver will explain what might be driving this longevity advantage he saw in people who had been incarcerated.

WEAVER: It’s really much more about what’s happening outside of prison than what’s happening inside of it.

And could better health care prevent some people from winding up in prison in the first place? Later, economist Elisa Jácome will discuss what can happen when an individual’s access to Medicaid suddenly changes.

JÁCOME: To the extent that we can keep individuals out of the criminal justice system by extending eligibility, then perhaps we could also reduce the likelihood these individuals ever come into contact with the criminal justice system

WEAVER: My name is Jeff Weaver. I’m a professor at University of Southern California. And I do work on a range of issues, involving the criminal justice system. I’m trained as an economist, but also very interested in what physicians have to do as well, since my wife is a physician.

JENA: Okay. I don’t want to hear what your wife is telling you about economists. Just keep that out of here.

The intersection of Jeff’s professional and personal lives recently led him to study the intersection of incarceration and mortality. The literature on this issue has been mixed. Some prior research has suggested that incarceration can actually lower people’s mortality risk — specifically Black men. This sounds paradoxical, but it makes sense if you think about it: outside of prison, black men have the highest mortality rates in the country. Inside prison, people are provided regular meals, housing, and health care that they may not have access to otherwise. But additional work has shown that this mortality benefit, if it exists, disappears when incarceration ends. There has been one unifying theme in past studies, though.

WEAVER: Almost all of them find where people, after they’ve been released from prison, have higher mortality risk than people who haven’t been incarcerated at all.

Jeff and his colleagues wanted to understand the specific mechanisms that were driving this higher mortality among the previously incarcerated—and whether it was incarceration itself that was the cause. They designed their study using my favorite approach: a natural experiment.

JENA: What did you do then to look at this question?

WEAVER: So, comparing people who are incarcerated to people who don’t have any criminal justice system contact, that’s probably not going to give us something that’s very informative. And so, let’s instead compare people who’ve been incarcerated to people who look similar in terms of, their age, their demographics, and who have also been charged with the same crime in these courts, but were not incarcerated for that crime. And so, what, we’re going to do is compare these groups to see what is the pre-release effect, as well as looking after release to try to understand what is the post-release effect of incarceration. The idea being if these mortality risks are similar, when we see differences between these groups, we can attribute this to incarceration rather than other things that are different. We’re going to take advantage of the feature of the criminal justice system in Ohio, which is that cases get randomly assigned. Judges are seeing cases that look pretty similar to each other, but judges have very different propensities to incarcerate people. And so then, if we see later on that there’s differences in mortality risk between these groups, we can say, “Okay, this is really due to the incarceration differences across these judges.”

JENA: You can’t conduct a randomized trial where you take a group of people and imprison some and do not imprison others and follow their mortality both in the period when one group was imprisoned, as well as the subsequent period. I mean, that would not be ethical, obviously. But this is sort of a natural experiment where a group of people who are otherwise similar, one group by chance is exposed to being incarcerated and another group isn’t just because the judge that they happened to get randomized to, had a lower likelihood of, convicting someone or, sentencing them for a longer period of time.

WEAVER: Yep, exactly. It’s a natural experiment and you say it would be unethical to conduct this as a real experiment. You might also think it’s a little bit unethical to do this., you know, in reality where someone’s life is tremendously changed as a result of the judge to whom they’re assigned. I remember when I first learned about this, I was pretty surprised.

JENA: Yeah, you would hope, that something as serious as this is not, in part due to luck, which is sort of concerning. So, what do you find?

WEAVER: When we look at the people who were not incarcerated, but charged with the same crime, we see that their mortality risk is about two and a half times higher. And so, it’s a lot higher for the people who are not incarcerated than the people who are during this pre-release period. What we then see is right around the time of release, the mortality risk for the people who are incarcerated shoots up and then looks exactly like the people who had not been incarcerated to start off with for the next 10 or 15 years. We think this is pretty convincing evidence that the reason we see this lower mortality rate while incarcerated is the result of incarceration and not due to something else that’s different between these groups.

JENA: What do you think explains this? You take two groups of people who are otherwise similar. One group gets incarcerated and, fast forward a year, if we look at the mortality, rate for that first group of people it is actually lower than the mortality rate for the other group that was not incarcerated in that same period of time. And then, when that first group exits prison, when they are released from prison, there’s a spike in mortality, for that first group, but then in the next 10 to 15 years, it’s no different.

JENA: So, overall, what do you see then? Like, if you were to say the first group versus that second group, what’s the overall effect on mortality?

WEAVER: I can just give you one number. What we find is the people who are incarcerated for a year have about a 15 percent lower likelihood of having died after five years. And so, this is a pretty substantial effect. But the message of this paper, I don’t think, is that incarceration is good. It’s that when we look at the individual causes of death, what you realize is this is really a paper, not so much about incarceration. It’s really much more about what’s happening outside of prison than what’s happening inside of it.

JENA: What do you mean it’s a paper about what’s happening in the communities outside of prison?

WEAVER: So, we see these differences in mortality, and then we look at what are the causes of death that are driving these differences. And there’s going to be three main ones that we’re going to look at. And so, the first of these is deaths due to natural causes. So, things that, you know, maybe potentially could have been prevented with medical interventions, so, cardiac episodes, cancer, et cetera. What we see is that during the time at which people are incarcerated, the likelihood of dying due to these causes drops by about a third. But then after the release, the risks look about the same. And so, what we think is going on here is that while people are incarcerated, they have access to healthcare, they may not otherwise have outside of prison. And so, even if prison healthcare is pretty well known to not be one of the highest quality sources of healthcare, it’s potentially a lot better than not having any healthcare at all.

WEAVER: And so, if we could have better extension of healthcare services, particularly to this population that has a very high risk of mortality outside of prison, this potentially could save a lot of lives.

JENA: And then what about other areas? I mean, I would think that we might be worried about opioid overdoses in this population. We might be worried about mortality due to violence, either in prison or in, the streets. How do those two play into this?

WEAVER: You have it exactly right. Those are the number two and three. When we think about prison, people may think of television shows like Oz and think that there’s a lot of murder happening there. While it’s certainly true that there’s plenty of violence in prison, when we compare people when they’re incarcerated versus outside of prison, the risk of being murdered in prison is about 50 times lower, five-zero. This is a pretty substantial drop, but when you look at the data and see, why is this the case? It tells you a little bit of a different story. Among the population we’re studying of people who appear in the criminal justice system, 85 percent of the deaths that we observe are due to firearms. And so, really what’s happening is that outside of prison there’s a lot of firearms. There’s a pretty high risk of being murdered due to a firearm. And so, as a result of there being less access to firearms in prison, the rate of gun violence, and homicide is lower.

WEAVER: I think the takeaway here is that there’s really high levels of violence, in the communities that, the offenders are in. And so, this is causing them to have really high risk outside of prison.

JENA: And talk to me about opioids.

WEAVER: We’re going to be looking at data from the state of Ohio, which is one of the places that’s been hit hardest by the opioid epidemic. And we see that your risk of death due to overdose while you’re incarcerated is about half what it is if you had not been incarcerated It doesn’t drop to zero, but it drops by half. And so, there’s a few things that could be driving this. It could be that potentially there’s less access to, drugs on which you might overdose in prison. It could be there’s better access to life-saving interventions like Naloxone. Something that we do observe is then saying, “Okay, let’s look and see what happens after people are released from incarceration. People are getting, treatment and addiction services while they’re incarcerated, and this is what’s causing this drop in likelihood of death due to overdose.” But we see that the two weeks after release, is a very high-risk period where people are even more likely to die of an overdose than the people who are not incarcerated to begin with. But then what we see for the 10 years afterwards is that their rate of death due to overdose looks basically the same as people who had not been incarcerated to start off with. And so, our takeaway from that is it seems like prison treatment services are not particularly effective, at helping this population, at least in the long term. And so, we think what’s probably driving this effect is the better access to potentially life-saving intervention and potentially less access to opioids.One other thing that we look at is, we have death-certificate data for all of Ohio, and we look to see what fraction of people who die of an overdose have some previous criminal justice system contact in terms of being arrested and appearing in court. And we see that 58 percent of the people who eventually are going to tragically die of overdose, have some previous court appearance prior to that. And so, when we’re thinking about this as a population, this is one that is really heavily exposed to the opioid epidemic. A lot of people who are eventually going to die of overdose are having this interaction with the government in terms of appearing in court. And so, if we’re able to figure out how to target the right people at this point in time, we could potentially prevent a lot of unnecessary opioid-related deaths.

JENA: You know, we think of things like hospitalizations as maybe sort of a, touch point where you can intervene on a patient. Maybe you could even think of prison as being such a place. But you’re talking about people who are not in prison, but who have some interaction with the justice system, prior to them dying of an opioid overdose, that could be another place to get involved in that care. I never thought about that before. Your study comments on the role that the environment that people are in has on their health outcomes. And we’ve talked about the role of place and geography in driving mortality. And then the next big question is, well, what do we do to improve those environments? In your case, if the mortality rate from firearms, from opioid overdoses, from lack of adequate preventive and chronic disease healthcare — if all those issues were better in the communities where these individuals are coming from, you wouldn’t find any effect, in your studies

WEAVER: I think what we’d also say is it seems like for populations that are at pretty high risk of not having much healthcare access, like the ones that we’re studying, it seems like even a little increase, even something that is documented to be not the best healthcare services, this can have a really big effect in reducing mortality. And so, there’s a lot of benefits potentially, to focusing more on healthcare services.

JENA: So, Jeff, your research, obviously, is focusing on what’s happening in the communities where these individuals are from. And I’m curious if you have thoughts on, what would be the impact of an approach that focuses on these vulnerable populations? Would the benefits only accrue to them? Or can you imagine a world in which the benefits accrue more generally? A world in which everyone sort of benefits?

WEAVER: I think this is one of those cases where there’s huge spillover benefits to, working with these populations. And so, there’s been some really great recent research that’s looked at, for example, the effect of extending healthcare access to people who potentially are at risk of later on committing criminal offenses, and showing that this actually can have some pretty big benefits in terms of making them less likely, to show up later having, committed a criminal offense. The sort of interventions that potentially could reduce those risks could have really big benefits for all society.

Extending health care access to people at risk of committing criminal offenses— to reduce the odds that they’ll do so—is one idea. Here’s another. What if people who already had access to health care never lost it?

JÁCOME: I especially find a more pronounced increase in incarceration for those individuals that at some point had been filing claims for mental health medications before losing access to Medicaid eligibility.

I’m Bapu Jena, and this is Freakonomics, MD.

*          *          *

JÁCOME: I got into economics because I am very interested in public policy and, my main research interest includes studying the interactions between low-income communities in the U.S. and the criminal justice system.

Elisa Jácome is an Assistant Professor at Northwestern University and she’s also an economist. When she started to think of research topics, she was drawn to questions surrounding mental health care.

JÁCOME: I realized that despite the large qualitative evidence on the importance of mental healthcare for the criminal justice population, there wasn’t as much quantitative evidence studying this topic in particular, thinking about how changes in access to healthcare might affect the likelihood that individuals come into contact with the criminal justice system in the first place.

In a 2016 survey from the Bureau of Justice Statistics, around 56 percent of inmates in state prisons reported a prior history of mental health problems. But only about one-fourth of prisoners said they’d ever received treatment for these issues. These individuals had been diagnosed with conditions like depression, anxiety, and post-traumatic stress disorder. Other research has suggested that having a major mental illness—like bipolar disorder, or schizophrenia — could increase the odds someone will go to jail by as much as 50 percent.

The link between psychological health and incarceration risk is well known. Figuring out what to do about it is still a big question mark. Elisa has wondered if access to health care, specifically mental health services, could reduce the chances someone would engage with the criminal justice system. Historically, this data has been difficult to obtain, particularly in the U.S. Luckily, the state of South Carolina links information across a few government agencies, and Elisa was able to peek inside.

JÁCOME: Importantly for this project, I was able to obtain access to Medicaid insurance claims. So, the Medicaid program provides a lot of behavioral health services, but of course it also provides overall healthcare as well. I was expecting to see that if individuals either gain or lose access to mental healthcare, then they might be more likely to come into contact with the criminal justice system.

JENA: Tell me a little bit about the study setting, who the population of people was that you studied, and how did you evaluate them?

JÁCOME: I’m looking in this paper at a sample of low-income adolescents in South Carolina. The context here is that in South Carolina low-income individuals in particular have access to Medicaid services, between the time that they’re infants and through 18 years of age. But when individuals turn 19, they age out of this eligibility and lose access to services. And so, what I do in this paper is focus exactly on that loss of eligibility when individuals turn 19. And I study how the criminal behavior of low-income men who were enrolled in this program changes after they lose coverage.

JENA: Someone who’s listening would say, “All right, okay, you’re focusing on the loss of Medicaid eligibility at age 19, and then you’re going to infer something about healthcare access.” And I think the follow up question someone will ask is, “Well, what if other things happen at age 19 that also increase the risk of incarceration?”

JÁCOME: Right, right. So, I think that would be sort of a central concern that the change in incarceration that I’m seeing in this data is not really coming from the loss of the insurance, but perhaps to other factors that are changing in this transition. I do a number of things in the paper in particular looking at individuals that are, residing in close geographic proximity to each other. So, who are of relatively comparable income levels so that I can make sure that it’s not, for example, that we’re just comparing a group of higher income to a group of lower income, but rather that we’re looking at adolescents that are growing up in very similar instances, who are born in the same year, who are sort of going to the same kinds of schools. And if I really see this difference arise between the two groups, just for the group that is losing access to coverage, then that might mitigate some of the concern that I’m picking up, perhaps other differences in income or other unobserved characteristics between the two groups.

JENA: Now, tell me, what did you find?

JÁCOME: I find that the two groups actually look quite similar to each other in terms of their likelihood of being incarcerated before age 19. But then after their 19th birthdays, the group of men who lost access to Medicaid — they’re 15 percent more likely to be incarcerated in any given quarter after their 19th birthdays. And I can use the law enforcement data to look at whether this effect is driven by individuals recidivating. So, they had already been in prison and now they’re just going back to prison. And I find that this increase is coming entirely from new individuals being incarcerated. And then I can split the group of individuals losing access to Medicaid, into those with and without mental health histories. When I split the sample in this way, I find that the increase in incarceration is entirely driven by those with mental health histories. To give you a sense of magnitudes, for this group of low-income men, when they lose access to services, they’re 22 percent more likely to have been incarcerated by their 21st birthdays, relative to the low-income men who are going to the same schools, who also had a history of mental illness, but who were less affected by this loss in eligibility.

JENA: What is it exactly about the mental health services that they lost that you think was keeping them out of prison? Is it medications, is it access to therapy? Is it things that are not related to mental health in particular, like just general healthcare?

JÁCOME: That’s a really tough question. What I can tell you is that of the low-income adolescents that I look at in South Carolina, many of them are using mental health medications throughout adolescence. And so, I especially find a more pronounced increase in incarceration for those individuals that at some point had been, filing claims for mental health medications in the year and a half before losing access to Medicaid eligibility. However, many of these individuals also use other types of services, whether that is, other, more conventional mental health services like therapy, but also case management services as well as other services that Medicaid covers in the behavioral, services realm.

JENA: Now, I can brainstorm a little bit. On the medication side, just two ideas. One is, suppose you found that among people on medications, there’s a set of medications for which loss of Medicaid eligibility for those people seems to have particularly large impacts, and then you could see whether or not those medications have been shown in clinical trials to be more effective. That might be something that suggests a medication loss, is at play. And another idea is, people with these conditions will sometimes take multiple different medications before they settle on something that seems to work for them. You could imagine that the duration of time which someone is on a medication, the last medication that you see them on, if it’s longer, that might indicate that they found something that worked for them. And for those people, taking them off of it would be expected to have a particularly large impact on the likelihood of incarceration.

JÁCOME: Yeah, I think that’s quite right. And one sort of related question that comes up is a lot of these individuals in this sample have an A.D.H.D. diagnosis. So, a question I often get is, is this coming, for example, from losing access to A.D.H.D. medications? That was one thing I could directly look at, and I do find that this group of individuals that has been using medications beyond A.D.H.D. medications seem to have a more pronounced effect in terms of, their likelihood of incarceration.

JENA: And do you think that it’s only about mental healthcare loss? Is there any way to tease out whether or not just the loss of general healthcare also has an impact?

JÁCOME: Yeah, so, especially as economists, we might think that losing access to Medicaid, which is a program that gives you access to free health coverage, might have an effect in the sense that the individuals who lose access to this coverage might in turn face higher medical costs, and that might make it just more likely for them to commit, for example, financially motivated crimes. One way to explore that is looking at that group of individuals without mental health histories. And so, these are individuals that might be using healthcare for alternative reasons, more physical healthcare. And for that group of individuals, I really don’t see any increase in their likelihood of, being incarcerated. And so, that suggests that, perhaps, it’s losing mental health services that is particularly key in this context.

JENA: When I read this paper, I was obviously super interested in the finding that the elimination of health insurance coverage could have these effects. But I was also thinking, you know, the primary reason that we want to offer insurance to people, is to make sure that they can live healthy lives and not be affected by the financial costs of care. But I would view this as sort of an important byproduct of that investment that we might not think about or factor into our calculations.

JÁCOME: I would definitely agree. And I think more and more, at least in the economics literature, it’s my sense that we’re starting to think about both the short-term and long-term benefits of safety net programs.

JENA: So, maybe a very straightforward extension of your findings would be that if you have people who have known mental health issues who are receiving healthcare services and who we think might be at some elevated risk of criminal justice involvement for that group, you’d want to be very cautious to eliminate their healthcare services. Right?

JÁCOME: And especially, if we think about the age crime profile, when are individuals sort of seeing this increase in their criminal propensity and the likelihood that they come into contact with the criminal justice system, age 19 is a moment in time in which criminal propensity is quite high. And so, taking healthcare away in that moment might be, something that policy makers might want to consider, especially because this is also a moment, in which individuals might be more likely to invest in their educational opportunities or otherwise improve their career prospects.

JÁCOME: To the extent that we can perhaps keep individuals out of the criminal justice system by extending eligibility, into, for example, the mid-twenties, then perhaps we could also just reduce the likelihood these individuals ever come into contact with the criminal justice system since we might put them on a different trajectory.

Giving people health care, even if they’re receiving it in prison, can change the course of their lives. So, can taking it away. As Jeff and Elisa’s work has shown, it’s a powerful tool that we might not be using effectively when it comes to the criminal justice system—before, during, and after incarceration. But we can’t, and we shouldn’t, rely on health care alone to prevent or fix these problems. We should be giving people more and better medical care because it’s the right thing to do, and if it keeps them out of prison as a result, we’ll take that too.

That’s it for today’s show. I want to thank my guests Jeff Weaver and Elisa Jácome, and thanks to you, of course, for listening.

Before I tell you about next week’s show, here’s an idea to think about based on my conversation with Jeff and Elisa, that’s sort of at the intersection of health care and the criminal justice system. A few years ago, some drugs were developed that revolutionized the way we treat Hepatitis C. The drugs were costly though and because Hepatitis C is so common in the prison population, it makes me wonder whether those costs, which were huge for state prison budgets, might’ve affected the way the criminal justice system operates. For example, is it possible that prisoners with Hepatitis C received shorter sentences, or were more likely to be granted parole, so that prisons wouldn’t have to foot the bill for these treatments? Something to think about.

And as for next week’s show:

Antibiotic resistance is one of the greatest public health threats facing humanity. Doctors know the problem is driven by the overuse of antibiotics. And yet:

LINDER:  At least 30 percent, if not up to about 50 percent of antibiotics prescribed in the United States in outpatient practice are either for not a good reason or for no discernible reason whatsoever.

We’ll talk about why physicians overprescribe these drugs, and how we might reign it in.

LINDER: It turns out that we’re powerfully influenced by what we think others think of us.

That’s all coming up next week on Freakonomics, M.D. Thanks again for listening.

*          *          *

Freakonomics, M.D. is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and People I (Mostly) Admire. All our shows are produced by Stitcher and Renbud Radio. You can find us on Twitter at @drbapupod. This episode was produced by Julie Kanfer and mixed by Eleanor Osborne, with help from Jasmin Klinger. We also had help this week from Lyric Bowditch. Our staff also includes Neal Carruth, Gabriel Roth, Greg Rippin, Rebecca Lee Douglas, Morgan Levey, Zack Lapinski, Ryan Kelley, Katherine Moncure, Jeremy Johnston, Daria Klenert, Emma Tyrrell, Alina Kulman, and Stephen Dubner. Original music composed by Luis Guerra. If you like this show, or any other show in the Freakonomics Radio Network, please recommend it to your family and friends. That’s the best way to support the podcasts you love. As always, thanks for listening.

JENA: it’s terrific to meet you, virtually. I’m sorry we can’t do this in person

WEAVER: I have to say it’s very nice to meet you as well, because I’ve listened to a ton of your podcasts and it’s a little disorienting to hear you speaking at normal speed rather than 1.7 speed, which is what I normally listen to it at.

Read full Transcript


  • Elisa Jácome, assistant professor of economics at Northwestern University.
  • Jeff Weaver, assistant professor of economics at the University of Southern California.