Search the Site

Episode Transcript

Several years back, we published a two-part series called “The Opioid Tragedy.” We interviewed physicians and economists, substance-abuse counselors and recovering addicts, and we all talked about how bad the opioid epidemic was back then.

Alicia MODESTINO: About 70,000 individuals died from a drug overdose just in 2017. That’s more Americans than were ever killed by guns, car crashes, or H.I.V./AIDS in a single year.

We also discussed a variety of solutions, some of them straight-up medical solutions.

Nicole O’DONNELL: Opiate-use disorder is treatable. It’s not a death sentence. It’s a medical condition, and it’s treatable.

And we also talked about harm reduction. That’s the idea that when it’s not practical to outright prohibit something that’s dangerous, it’s worth finding a compromise. In the case of a dangerous opioid like fentanyl, that might mean treating people with a less-addictive opioid, like buprenorphine. In a later episode, we discussed harm reduction with Rahul Gupta, Director of National Drug Policy in the Biden administration.

Rahul GUPTA: This administration has been very clear: for the first time in the history of the United States federal government, we have made harm reduction the central tenet of how we need to move forward.

Gupta talked about government-funded needle exchanges and the distribution of naloxone, a drug that can rapidly reverse an opioid overdose. There have been other developments in the fight against opioid-overdose deaths — including an intensive law-enforcement campaign to cut down on drug trafficking. Also, billions of dollars of settlement money has started to flow from the opioid manufacturers, distributors, and consultants who did such a good job of selling their products. So, with all that money, and all that law enforcement, and with harm reduction and medical treatments, you might think we had the problem surrounded. You would certainly think that opioid deaths would be falling. But they’re not.

Keith HUMPHREYS: It’s horrible. It’s absolutely horrifying. 

That is Keith Humphreys.

HUMPHREYS: I’m a professor of psychiatry and an addiction researcher at Stanford University.

Humphreys has also worked on drug policy for the Bush and Obama administrations, and for Joe Biden’s presidential campaign. He has watched in horror as annual opioid deaths continued to climb. Today on Freakonomics Radio: I wish we weren’t doing this, but we are starting another two-part series on the opioid epidemic, to try to figure out why it keeps getting worse. Some of the answers are very simple.

HUMPHREYS: Depression feels bad every day. Drug use doesn’t feel bad. Drug use feels incredible. 

But there are other, hidden factors that are driving the epidemic, and we will explore those too. We will find out which of the proposed solutions have failed, and why, and we’ll ask what might work better. We’ll try to track where those billions of settlement dollars are going. And we’ll ask some questions that may make you uncomfortable. For instance: in recent years, there has been a push to destigmatize drug use; is it time to maybe bring back the stigma?

*      *      *

Most epidemics come out of nowhere, do their damage, and fade away. Why is the opioid epidemic different? To answer that question, we need to add some context, and some history.

David CUTLER: Opioids are perhaps the most abused substance in the history of the world.

That is David Cutler.

CUTLER: I’m a professor of economics at Harvard.

Cutler is one of the most prominent healthcare economists in the world. And, like Keith Humphreys, he has done his share of government service. Cutler was an economic advisor in the Clinton administration and a healthcare advisor on Barack Obama’s first presidential campaign. He has also consulted with state and city governments.

CUTLER: Anyone who’s interested in helping make healthcare work better is someone who I’m happy to talk to. 

DUBNER: And what would you say if I said, “Well, I get that you’re an economist, and that you may know a lot about certain things, but how does that most fruitfully intersect with healthcare policymaking especially?”

CUTLER: An enormous amount of healthcare policy has to do with economics. For example, what incentives can you give physicians so that they do things that you want to happen, but not things you don’t want to happen? Or what incentives can you use for individuals to help them take the medications that they should take, and not take medications that they shouldn’t take, and not smoke and behave healthy, and so on? We’re dealing with incentives and we’re dealing with how to design a system so that it works. Those are things that economics really knows a lot about. Not everything in healthcare is just economics, but it’s also the case that if you don’t bring an economics lens to issues, you often get many things wrong. 

Cutler has recently focused his economic lens on opioids — which, as he mentioned earlier, have been around for millennia.

CUTLER: People, of course, smoked opium forever. There were wars fought over the right to import opium to China. Some of what’s happened over time is we’ve gotten better at extracting the key ingredients. So, heroin and morphine are both derivatives of the opium poppy, and those are more potent.

The ancient Sumerians used to call this poppy the “joy plant.” Today, we’re dealing with not just natural opiates, but with synthetic and semi-synthetic versions. The one you’re probably most familiar with is fentanyl, an incredibly powerful drug that was developed in the late 1950s as an anesthetic. It’s still widely used in hospitals for anesthesia and pain management. But it’s the street version of fentanyl that’s causing most of the overdose deaths today. A fentanyl overdose kills by slowing down the respiratory system so much that there’s not enough oxygen reaching the brain. And opioids are extraordinarily addictive; widespread addiction in the U.S. goes back to at least the Civil War, when wounded soldiers were given opium and morphine, a derivative developed in the early 19th century and named after Morpheus, the Greek god of dreams.

The current epidemic also has a medical history. It started in the 1990s, when the American pharmaceutical firm Purdue Pharma began promoting a new opioid called OxyContin. The big breakthrough was that OxyContin wasn’t nearly as addictive as other opioids. At least that’s what Purdue claimed. But that claim turned out to be — what’s the word I’m looking for here? — false. By the time this falsehood was widely known, Purdue was selling billions of dollars’ worth of OxyContin a year and hoped to continue. For many people, the introduction of OxyContin marks the outbreak of the modern epidemic.

CUTLER: From the mid-1990s through roughly 2010 you see increasing supplies, just massive, massive supplies, like a five-time increase in opioid prescriptions, opioid use. A massive increase in deaths from opioids.  

DUBNER: And these are mostly legal substances used by people to whom it was prescribed? 

CUTLER: Much of it was used by people to whom it was prescribed, but some of it then gets passed on to friends and relatives. There becomes a black market for it, so people who are addicted will buy it from others. There are pill mills where they’ll, in principle, examine you, but not really. They’ll give you a prescription. You pay all in cash. You get the drugs.  

We talked quite a bit in our earlier opioid series about this supply-side story. But despite all the death and damage since then, despite the anguish of millions of mourners and survivors since then, the overall problem has gotten worse. David Cutler wanted to find out why. So he started a research project, in collaboration with the economist Travis Donahoe.

CUTLER: Travis is currently a professor at University of Pittsburgh. And at the time we started out, he was a Ph.D. student at Harvard in the health policy program. 

DUBNER: You were an advisor to him?  

CUTLER: I was an advisor to him. And he actually grew up in West Virginia. If there is an epicenter of the opioid epidemic, it is West Virginia. So he was always, always interested in things having to do with opioids, and pain, and deaths due to that. 

Travis DONAHOE: I grew up in Huntington, West Virginia, which, at the time I was in high school, became widely known as a county that had the highest adult obesity and depression prevalence in the United States.

That’s Donahoe. West Virginia also has the highest rate of drug-overdose deaths in the U.S.

DONAHOE: Many people that I went to high school with, friends, have had opioid addiction over time, and there’s been a number of people that have overdosed. 

CUTLER: Travis wrote his dissertation on policies to address the opioid epidemic, particularly D.E.A. intervention against distributors and dealers and so on. There’s a lot of literature on the opioid epidemic, including a lot of very good literature on the transition from people using legal opioids to people using illegal opioids, and so on. And we were really puzzled, first by the fact that, like, “Oh my gosh, how long can this thing go on?” But then by the fact that there are reasons why people stop taking things. Like, people learned that smoking was bad for them, and they stopped. And not only that, they learned that the cigarette companies had been lying to them. And they were like, “Well, to heck with you guys. We don’t want to be using this product.” So, tobacco is a very addictive substance — tobacco use has fallen well more than 50 percent since its peak just after World War II.  

DUBNER: Now, talk for a moment about the levers that contributed to that. Because it wasn’t by accident, and it wasn’t cheap and it wasn’t easy. And there was a lot of regulatory and taxation power put to use, yes? 

CUTLER: There were a lot of public and private policies. There were public policies around taxation, around regulation of where you can smoke. There are private policies, like employers saying you can’t smoke in the workplace or you have to go outside to smoke. There was social pressure, peer pressure, “No, you can’t smoke in my house.” There’s what people were taught, just the whole attitude, “You really want to smoke?” So, it was a combination of public and private actions that led some people never to start smoking, some people to quit smoking, and others to help people stay off cigarettes. The net effect is that combustible cigarette use is very, very down. 

So why haven’t opioids followed the same trajectory as cigarettes? That question brings us back to this new research by Donahoe and Cutler.

DONAHOE: For a 30-year period, opioid-overdose deaths have been increasing continuously. What we want to ask is, why has that occurred? We know about things that have sparked it. We know that about things that have exacerbated it. But what is it that would produce this kind of a continuous trend?  

They recently wrote up their findings in a paper with a title that only an economist could love: “Thick Market Externalities and the Persistence of the Opioid Epidemic.” And what is a thick market externality? In this case, it describes the fact that opioid users end up creating more opioid users.

CUTLER: What’s going on here is the idea, “Well, maybe I’ve hurt my knee, maybe I hurt my back or something. If there’s a lot of opioids around, I’ll bump into someone who has some. Maybe I’ll experiment with them.” Whereas if not, or if I have to go into the illegal market, or I have to deal with someone I don’t know in some dangerous setting, maybe I’m not going to do that. So, just the availability of the substance can encourage others to use it. And that’s what we look for, and what we find evidence for. 

Their evidence comes from several sources, like government figures on opioid deaths and the supply of drugs — but also: from Facebook.

CUTLER: So, it looks like it’s spreading through social networks. If one county has more deaths, nearby counties have more deaths. If one county has more deaths, areas of the country that have more Facebook friends, then those areas also have more deaths. Sometimes the literal physical product will spread through networks, and sometimes just the idea, “Oh, when I was in pain, I got this opioid, and maybe you should try this opioid.” We have a quotation from someone who said, “Oh yeah, I got a call from a friend saying, ‘I just tried this thing, it’s the greatest thing ever. You just go tell your doctor you have back pain and ask for a prescription for it, and you’ll love it.’’

DONAHOE: When you look at data on initiation of opioids and other drugs, the typical age that a person initiates these kinds of drugs is pretty young. I would venture to guess that most people were not sitting in a void and then independently became curious about how to use heroin and then went out and figured out how to do that. It was probably that someone in their network was using heroin as well, and then that ultimately influenced them to learn how to do it.  

DUBNER: Do you know, what is the median age of first use of opioids?  

DONAHOE: I don’t know that I know precisely, but I have seen the number 12 to 14 float around. So, yeah, it’s quite young.  

According to some fairly reputable government surveys, roughly half of the people who abuse opioids got them for free from a friend or relative. And where do all these pills come from? The Centers for Disease Control estimates that 57 million people — that’s nearly 20 percent of the U.S. population — had at least one opioid prescription filled in a single year, 2017. For our earlier series on this topic, we spoke with Stephen Loyd, a Tennessee physician and a recovering opioid addict. I asked him how he sourced his drugs.

Stephen LOYD: At first it was out of people’s medicine cabinets. For the longest time, I had a pretty much endless supply. If I came in your house, and you had an old prescription left over, I walked out the door with it. And then the other way was doctor-shopping. You know, all my friends were doctors. And so I would just hit them up at different times for prescriptions. 

Today, Stephen Loyd is heavily involved in trying to fight the opioid epidemic. We will hear from him later in this series. As for the new research by David Cutler and Travis Donahoe: they estimate that spillover effects — social contagion, basically — can explain roughly 90 percent of opioid deaths. Spillovers, they write, “are the main reason deaths have increased for so long.”

CUTLER: When people in one area use it, they tell their friends or they give substance to their friends. And so both within the same area and then within other neighboring areas, and within other areas where they have friends, even if they’re not neighboring, then you see usage and deaths go up there. And then you get all those feedback effects, like, all those echoes, and it can get even bigger and bigger. And so that either significantly minimizes the extent to which it dies out, or it actually creates this sort of unstable spiral, where use one year translates into even more use the next year, and even more the next year. You know, 30 years into an opioid epidemic would have been more than enough time under usual circumstances for opioids to die out. But because of these echo effects, it just keeps growing and growing. 

DUBNER: Let’s talk about the previous and conventional explanations for the opioid epidemic, and the fact that it’s continued to rise in volume and intensity. Let’s talk first about what others in your profession, particularly Angus Deaton and Anne Case, have called “deaths of despair,” which I believe is a self-explanatory phrase, but also something that probably most listeners are familiar with. What components of that argument, as pertain to opioid deaths, do you think are accurate and which are perhaps inaccurate?

CUTLER: So, their work is incredibly important. It’s among the most important things that has been written about public policy in decades. The idea behind it is that people are in despair, either because of physical pain or mental pain, or really stemming back to society that hasn’t worked the way they would like, and that that leads people to use illegal substances partly as an out for it. So, some of the specific things, like, for example, there are more people in pain than there used to be, but the increase in pain is nowhere near as big as the increase in use of opioids. 

DUBNER: Now, David, you’ve written a paper on this very idea, I believe with your colleague Ed Glaeser, correct? 

CUTLER: That’s correct. What’s happened over time is not that there are just more people in pain, although there absolutely are, but the people in pain are doing different things. Once opioids were sort of safe and effective, it became, “Oh, you have a toothache, here’s opioids.” “Oh, you’ve got a hurt back because of work? Oh, use opioids.” “Oh, you’re down and out, and that’s presenting with psychological pain but also physical pain manifestations? Use opioids.” And so it is true that there is this large share of people in physical and mental pain. The way that the medical system and people have responded is different than it used to be, and it involved trying to medicalize it and treat it. That’s problematic here, when the treatment is not effective, and in fact is addictive. For God’s sakes, if the treatment is worse than the disease, it’s really a terrible thing.  

DUBNER: One thought I had while reading your new paper, on the social spillovers of opioid abuse, is that the “deaths of despair” umbrella explains many opioid overdose deaths as the result of loneliness and isolation. But what your paper is arguing seems to be kind of the opposite of that, which is that it’s not isolation or loneliness — or maybe it is loneliness, but it’s not isolation. It’s actually connection. That without connection afforded by the internet particularly, that this epidemic would not have continued to grow. Is that a fair read?  

CUTLER: Yes, the connection is absolutely fundamental here. Of course, these connections are different than the kinds of connections that Anne and Angus are writing about, in that they’re talking about meaningful connections in your life. And the internet here is about, where can I buy something, or where can I obtain the product cheaply? — which is a different kind of connection. 

So that’s one economist, David Cutler, talking about how his research — which shows that a certain kind of connectivity has helped prolong the opioid epidemic — can square up with some other economists’ argument about “deaths of despair.” But what does a non-economist think of David Cutler’s argument?

HUMPHREYS: I would praise two things about the Cutler paper. 

That, again is Keith Humphreys, the Stanford addiction researcher and drug-policy advisor.

HUMPHREYS: So one is, it drives non-economists crazy when economists show up in a new area and act as if no one has studied it before. So I actually first look at the references. Have they actually read anything addiction? And they have. So that’s good. Second thing is they are absolutely correct that social processes spread addiction, much the same way as they might spread something like Covid. People who are using invite other people to use with them. Sometimes they do that because they’re like, “Well, I need to sell to keep my own habits, so this is someone I could sell to.” But oftentimes it’s a, you know, friendship. It’s fun. Let’s do this together. And particularly if you’re in the early stages of drug use, it can look very compelling. You know, if you’re deeply addicted and you’re homeless, it’s pretty hard to persuade someone, hey, you want to live like me? But if it’s like it’s a party, you know, so that part is true. 

But Humphreys, as a drug expert, also has some critiques of the economists’ paper.

HUMPHREYS: There’s two things in analysis that I think are pretty seriously questionable. First is, they’re trying to explain: why do we keep having this epidemic after 25 years, when the average person would say, “I don’t want to do that,” and that would help make it die off, and second, the police would shut it down? But, you know, I’ll just say, there’s another epidemic 25 years that has also gotten more severe, and that’s alcohol. Alcohol deaths are up and they don’t look at that. Why don’t we shut the alcohol industry down? Because it’s a legal industry. And that is a big reason why the opioid epidemic has gone on so long. They had gold-plated protection as a legal industry. One of the most remarkable things that happened, the Drug Enforcement Administration caught distributors delivering a million pills to towns with 300 people. I’m not exaggerating. And you think that’s going to stop them. The distributors went to their friends in Congress, and got a new law passed that basically stopped the D.E.A. from doing that. 

We went back to Cutler and Donahoe to ask what they thought of these critiques. They wrote back to say that they agree that insufficient regulation “is an important reason that the epidemic has lasted so long,” but that it doesn’t conflict with their argument about social spillovers. In towns that had pill mills like Keith Humphreys was talking about — a million pills shipped to a town of 300 — opioid deaths continued to rise after those pill mills were shut down. And Cutler and Donahoe say that alcohol may actually have social-spillover dynamics similar to what they have observed with opioids. The “persistence in demand for alcohol,” they wrote, “would be related to its near-ubiquity in social settings.” Think about that for a minute the next time you have a drink with friends or colleagues. Keith Humphreys had another problem with the economists’ paper. This one has to do with how they measured opioid overdoses.

HUMPHREYS: They’re treating overdoses as an index of demand, and they aren’t an index of demand. Overdose deaths are a function of how often somebody uses, and how risky each episode of use is. 

The risk he’s describing is the likelihood that a given drug user will overdose in a given year.

HUMPHREYS: In the ‘90s, people were using Vicodin, maybe it was 1 in 200. And then Oxy raised it to 1 in 100, and then heroin raises it to 1 in 50. And with fentanyl, maybe it’s as much as 1 in 20 for a year of use. So when you say, “Oh, deaths are going up, it must mean demand is going up.” No, actually, demand could be dropping. The problem is the risks of use are going up. The number of people who say “I’m initiating opiates” is dropping. And the number of people who say they’re using fentanyl is nowhere near what it was, the number people were saying they were using the pills back when that was the heyday. It’s just that it’s so deadly. 

And that’s really the big issue, the potency of the drugs. We will get into that later in the episode. Cutler and Donahoe told us that their measure of demand was not just about overdose deaths — their analysis looked at, quote, “the total potency-adjusted quantity of opioids that are sought” — that is, the number of people seeking opioids multiplied by the amount they want to use. They wrote: “There is a lack of good data on how many people are using various substances and how much they are using. Thus, it is very hard to know what is happening to the number of people in the market for a drug like illicitly-made fentanyl … Everyone in the field wishes we knew more.”

*      *      *

Fentanyl is a synthetic opioid that’s more than 50 times as powerful as morphine. In the U.S., it’s responsible for about two-thirds of fatal drug overdoses. As I mentioned earlier, fentanyl started out as a hospital drug, for anesthesia and pain relief. The World Health Organization still lists it as an essential medicine for the management of cancer pain. Here’s Keith Humphreys, from Stanford:

HUMPHREYS: It’s terrific. I worked in hospice for a number of years. For people with late-stage bone cancer, a fentanyl patch can give them relief when nothing else can. And it had been recognized for a long time that this could really be a blockbuster on the illicit market. And some criminal gangs tried to introduce it. There was an outbreak around 2000. There was another about five years after. But they were domestic groups. They were small. And law enforcement was very good at jumping on them right away, and shutting them down.

But that changed, as the booming black market for fentanyl in the U.S. attracted foreign suppliers.

HUMPHREYS: We live in a globalized economy. And China, which has a massive chemical industry, a massive pharmaceutical industry, and an export-led economy, has a lot of people who wouldn’t mind earning a little extra money on the weekend, and they started to do it. And that combines with the technology of the internet, that they began selling what are called precursors, the stuff I use to make the fentanyl, over into North America, for example into Mexican suppliers. Also, sometimes just sending fentanyl directly in the mail. Fentanyl is so potent that you could — let’s say you had a dealer in Dayton, Ohio, you could put it in an envelope that was something that might have a Christmas card in it. Plus, this is not like in the old days where maybe, you know, Guinea-Bissau is growing a plant that’s turned into drugs, and a superpower can pressure Guinea— you can’t really pressure a nuclear-armed superpower like China to do anything.

DUBNER: Has the supply to the U.S. been significantly constrained or not? 

HUMPHREYS: The most important thing that’s happening in drugs is the departure of drugs from their agricultural base that they had for thousands of years. Let’s say you’re running a heroin business, and I’m running a fentanyl business. What do you have to do? Well, you got to get some arable land. Probably in a place with weak government, dodgy local politics.

DUBNER: Afghanistan looks good.

HUMPHREYS: Right. But now you have to pay off a warlord, and you got to find peasant labor, and you got to protect the farm, and you got to make sure no other warlord burns the crop. And then you got to get it across the border. Well, that means either smuggling costs or bribery costs, then you got to put it on a boat, and it’s big and it’s bulky and thousands of miles. And then the Coast Guard grabs it, and then you got to wait for the next growing season. And then there’s blight or a drought. Whereas me, it’s like, hey, I need some fentanyl. Bill, can you whip it up in the sink? Yep. Have it for you in a couple hours. So I’m going to put you out of business. My production costs are about one percent of your production costs. And that is what is happening. Right now in California, heroin is very hard to find. I know people who are addicted say, like, “I can’t get it anymore.” So a lot of the traditional things that countries used to do to suppress the trade are irrelevant. What do you pull up plants? There are no plants.

When a drug is that potent, and that cheap, it’s hard for government and law enforcement to do much about it.

HUMPHREYS: I testified to the Senate about this just recently. Understandably, they want fentanyl kept out of this country. But it’s so compact that our entire consumption — at least the RAND people who are very smart at this kind of stuff, think it’s only like 5 or 10 metric tons a year. That’s a truck. A truck. That’s pretty easy to hide stuff that small. So it’s very hard to keep it out of the country.

CUTLER: Most of the fentanyl that comes into the country comes through legal border crossings.

That, again, is David Cutler, the Harvard healthcare economist.

CUTLER: This is not people bringing quantities of fentanyl with them through illegal immigration. It’s basically coming into the country through a legal method, and then from there going through the supply chain to wind up being distributed to users. It gets either incorporated into heroin or pressed into pills that say it’s OxyContin when it’s not, or combined with methamphetamines, or any number of other ways, and then people die from that.

DUBNER: And I assume the concentration is far out of whack, yes?

CUTLER: The concentration is far out of whack, and the potency varies a lot from batch to batch, because you have to dilute it. And so that’s part of a high level of deaths is, I may be used to a certain dose, but if you don’t mix it correctly, I may get more dosage than that. And that can be a big problem.

DUBNER: Now, I know very little about drug dealing, but I don’t understand why it would be that drug sellers, drug dealers, would want to include a fatal substance because they’re killing their customers. Can you explain that? 

CUTLER: There are different versions of the answer to this. One version is exactly yours, which is: they don’t mean to, but they sometimes do by accident because the mixing isn’t so great. A second version is, what you really want if you’re in the market is, you want to know that the person has a potent batch. And so it doesn’t hurt if the batch has killed someone else. That may attract more customers to you. “Gosh, if someone died using a given dealer’s fentanyl, that must mean it was pretty good.”  

The current supply of street drugs in the U.S., like cocaine and methamphetamine, contains a lot of fentanyl. In a recent study of drug users in New York City, more than four out of five people tested positive for fentanyl, but only one in five said that their fentanyl consumption was intentional. Given this widespread contamination, and the massive overdose risk, and all the other suffering that opioids have caused over the past couple decades, it may be tempting to take a hardline stance against drugs, full stop. Keith Humphreys is more measured than that.

HUMPHREYS: I think people rhetorically, there are people saying all drug use is bad. But from a science viewpoint, when you realize what is a drug and how broad that category of thing that is, you realize very few of us go through our life without taking drugs. 

DUBNER: Do I have to put down my caffeine right now?

HUMPHREYS: There you go. I mean, I’ve often brought that up. You know, I love caffeine. I don’t have a problem with it. If I had to choose between caffeine and my children, I could make that decision. But I would miss them. I would really miss them. I have been in meetings where people condemn the evil of drugs, and then they all go out and have a drink, as if alcohol weren’t a drug. We go to a doctor and we get drugs that save our lives, right? When you realize that drugs are not just the thing that’s in the paper connected to a crime story about a deal gone wrong and somebody got shot, you realize it’s almost a universal human thing to take drugs. So really what we’re arguing about is, when do we take them, and how, and for what reasons, and how it’s monitored. I have been a proponent of a medication called naloxone, which is an opioid antagonist. So what that means is, when you take an opiate, it binds to particular receptors in your brain, and naloxone essentially goes to that same receptor and knocks that opioid out. Now, is there any possibility that somebody might take more opioids knowing their friend was there with naloxone? That is definitely possible. We know risk compensation happens in lots of behaviors. People drive more quickly when they have their seatbelt on. However, the equation is risk compensation minus how effective the safety device is, right? So if I think, how much more risky drug use would naloxone cause? I say it’s a pretty small amount. How different is it to overdose with naloxone than without it? Oh boy. That’s a huge effect. You have somebody who literally is dying. A few moments later, they’re breathing again, which is pretty incredible.

DUBNER: When I say “harm reduction,” is that a phrase that you generally embrace, or do you feel it’s come to cover too much ground, perhaps?

HUMPHREYS: Well, if you want five definitions of harm reduction, talk to three people. So I try, as I do with lots of words people argue about a lot, is I often will start to say, well, tell me what you mean by that. I hope none of us wake up in the morning trying to do harm, right? So we’d like to have less harm. And it shouldn’t be overdrawn, as some people do, the distinction, say, between harm reduction and treatment. Because the truth is, many people who go to treatment end up continuing to use, but less and in a less dangerous way, and a number of people who get in contact with harm reduction that ultimately ends with them deciding that they don’t want to use drugs at all. Sometimes the difference gets overstated, the clans start to fight with each other. I just look at this as a public health, public safety thing, like, what is the impact? And if it’s what I consider a good impact, then I’m in favor of it.

This gets us into another tricky area of drug policy. Some people who argue for harm reduction think the best way to get there is to decriminalize drug possession. That would remove some of the stigma of drug use; it would allow for more safety regulations; it would free up law enforcement resources for other problems. One U.S. state, Oregon, recently tried decriminalization.

HUMPHREYS: Yeah, at the end of 2020, the people of Oregon voted in the general election to remove all criminal penalties for drug use. And practically speaking, also reduce penalties for drug dealing. So that changed the character of the state. And there was unfortunately, a big increase in overdoses and — I notice, I go to Oregon a lot — a lot more public drug dealing, which has bad effects on neighborhoods. Now, it was also a pandemic, right? So things could have easily gotten worse anyway. But the faith that the advocates had, that if you removed all pressure and you removed all shame, from sitting on a park bench using fentanyl, then people would seek out care proved to be completely incorrect.

DUBNER: Did that surprise you, and others in the field?

HUMPHREYS: Me? No. I just think about the neuroscience, I think about the biology of reward, is that that would have been true if the condition were, say, depression. Depression feels bad every day. People are so happy to get rid of depression if you can get rid of it. Drug use doesn’t feel bad. Drug use feels incredible. In those short moments, you get this great reward. I mean, that’s why, you say, “Why did this person give up their family, their health, their home for this?” It’s because in the short term, it feels great. And so, most people who seek help are under some pressures. There’s some countervailing force. The spouse says, “You keep doing this, I’m taking the kids in and I’m moving out.” You know, boss says, “You show up high to work one more time, you’re fired.”

DUBNER: And you’re saying making things more easily available does not increase that countervailing force?

HUMPHREYS: No, just the opposite. And there was also sort of rhetoric, “It’s wrong to think anything negative about it. We need to fully destigmatize the behavior.” But when people come in for care, it’s very often because they do feel ashamed. They feel like, you know, I’m letting my family down. I’m letting myself down.

DUBNER: So I don’t want to say, “Are you a fan of stigma?” but do you see a useful role of stigma in the way society thinks about drugs?

HUMPHREYS: Absolutely. So when I was growing up — I grew up in West Virginia in the ‘60s and ‘70s — two things considered funny are not considered funny anymore. One was drunk driving, and the other one was hitting your wife. People joked about these things. Now they’re deeply stigmatized. And I’m glad, because that is a signal to people those are wrong things to do. So there has to be some pressure. And I say that at the same time saying, like, I’ve always been against, forever, throwing people into a cell for the mere act of using a substance. But there’s plenty of smarter things we can do than that.

*      *      *

In 2020, the Stanford drug-policy expert Keith Humphreys led a commission between his university and The Lancet, a medical journal in Britain.

HUMPHREYS: It’s one of the most influential journals in terms of global health. People read it everywhere. And they create commissions to look at global health challenges. And so there’s a Lancet commission on, you know, diabetes, Lancet commission on the health of children, Lancet commission on malaria. And unusually, they wanted to do one just on a couple countries, the U.S. and Canada. And that was because of our opioid crisis. It was so bad they said, “This isn’t global yet, but let’s do something about it now.”

The commission was asked to come up with solutions that would cut opioid deaths, from pharmaceutical marketing to medical education to policy. Humphreys has even more suggestions.

HUMPHREYS: You want to make drugs as hard to get as possible. You cannot eliminate them, that is impossible. But you can certainly raise search costs. There’s a thing called drug market intervention, that the police, community health do together to close down open-air drug markets. Why do those markets exist? They are not to serve the people in the neighborhood. If you were in the neighborhood, you would know the dealer, the dealer would know you. They’re there so that people from outside the neighborhood can rapidly find a dealer, and a dealer can rapidly find customers. And that just has all these destructive effects. You can close those down. And that’s something that’s definitely helpful.

That’s one suggestion. What else does Humphreys have?

HUMPHREYS: With the seven, eight million people at any given time who are on probation and parole, we should be drug-testing and alcohol-testing all of them, regardless of what they’re arrested for. And giving rewards and penalties based on their use — immediate, swift, and certain rewards. I’m not talking about violating them back to their original sentence, but you could have things like, “If you go through a week and you don’t use cocaine, we’ll knock a week off the end of your parole sentence.” Or the other way: “I’m afraid we’re going to add another week on your parole because you didn’t do that.”

Humphreys himself has done some work on a program like this, for alcohol abuse; it’s called 24/7 Sobriety.

HUMPHREYS: It was invented by a county prosecutor named Larry Long. Remarkable guy. He was seeing people he grew up with, in a small town in South Dakota, cycling through the court over and over with alcohol problems. And he felt bad for them because he knew — you know, we threaten you, we take away your car, we throw you in jail. Nothing works. And he said, the problem isn’t driving; the problem is drinking.

24/7 Sobriety is a court-mandated program for people who’ve been arrested multiple times for drunk driving. It involves constant and frequent testing.

HUMPHREYS: Every morning, you have to come in, and you blow a breathalyzer. If it shows negative, you get immediate reward: “Have a great day, Keith.” You know, another day of freedom. If, on the other hand, it’s positive, there’s an immediate consequence: you’re arrested on the spot. Not maybe — certainly. And you are held in a cell for just one night. But it starts that night, immediate. Now you’d think in a way, a lot of these folks have been in prison, why would they care about one night in the jail system? It’s because it’s a swift and certain consequence. And all those other consequences in criminal justice are very probabilistic and distant. So when I heard about this program, I was in the Obama administration, I thought, “Oh, come on, half these people are going to show up drunk and the other half are going to be rampaging around the countryside.” And I went there the first morning, I remember this, in Sioux Falls, and watched 200 people go straight through — all 200 showed up, all 200 blew negative. In South Dakota now, they’ve done over 10 million tests, and the success rate — meaning, the proportion of times people show up and are not drinking — is 99.1 percent.

The 24/7 Sobriety program also produced a significant reduction in repeat arrests for drunk driving.

HUMPHREYS: Violence against women also went down dramatically. When you take alcohol out of somebody’s life, other good things happen.

So that’s one set of ideas that Keith Humphreys thinks could help cut opioid deaths. What else?

HUMPHREYS: We could definitely have a decent addiction treatment system, which we do in some places but not in most of the country. Treating addiction like a serious chronic illness, meaning it would be core to our healthcare function. I think we’ve come to that with smoking. Your doctor now is very comfortable talking to you about your tobacco use and also, you know, do you want some Nicorette gum or do you want to try some Zyban or whatever. But you need to do that for all the other substances as well. We need to save as many lives as we can in the meantime. Naloxone is a life-saver. Syringe service programs: that reduces the spread of infectious disease, that is certainly a very worthwhile thing to do as well. And then the last thing is, in particular to opioids, you have to remember that the American opioid crisis didn’t start on the streets. It started in hospitals and doctors’ clinics. And we need to do a much better job regulating legal producers of drugs. It’s interesting how common it is to assert that if only these things were legal, we wouldn’t have these problems. Remember, the biggest problems we have are all with legal drugs. Eight million human beings a year die from smoking on this planet. That’s 10 times more than all the illicit drugs put together. So you’ve got to regulate adequately.

DUBNER: What about the demand side?

HUMPHREYS: The consumption is very skewed. So, you have very high-consumption people. Those are great targets for treatment. Often, in fact, the first-line treatment is an opioid, which strikes some people as — you know, how would you treat an opioid with an opioid? But they’re dramatically safer, opioids like buprenorphine and methadone, relative to fentanyl. They give people stability and they can do things like hold a job, be with their families, and that kind of thing. Every time you take one of those folks out, it’s a huge hit to demand. And then there’s the investments that need to be made on the prevention side. This is the thing that’s the hardest to get everybody interested in, because we are consumed, understandably, by the suffering that is right before us. But if you look at other epidemics that people my age have lived through — you know, Covid, H.I.V./AIDS — those were not solved by saying, let’s wait till people get really sick and then spend a lot of resources to try to help them. They were ended by getting people into recovery who were sick but even more importantly, stopping new cases. 

DUBNER: Right. So what’s the version of vaccine in the opioid problem? 

HUMPHREYS: You do as much as you can to avoid exposure, particularly when people are young. We still have a problem of, like, one in eight people go into an emergency room with a twisted ankle comes out with an opioid prescription. We still have doctors prescribing opioids to teenagers for headaches. So you want to reduce that exposure, but you also want to reduce the demand. And I emphasize that time in life, by the way, because the neuroscience shows our brains are the most plastic and changeable when you are young — as anyone who’s trying to learn a language in their 50s knows, right? So just like you can pick up your French or your Spanish really fast when you’re a teenager or even younger, that is where most addictions start. If you make it to 25 without an addiction, you will probably never develop one. That’s where prevention and health promotion comes in. When I say that, a lot of people’s eyes glaze over, they think about D.A.R.E, which was very broadly distributed, and didn’t really —

DUBNER: And not successful.

HUMPHREYS: And not successful. And, you know, you can couple that with the picture of the egg: “This is your brain on drugs,” right?

DUBNER: I always thought that egg would make stoners hungry. Sorry.

HUMPHREYS: Yeah, way to flip the discussion. But thinking about that as that’s what prevention is, it’s kind of like thinking the Tandy RadioShack 80 is what computers are about. That was a very long time ago. And since that time, there’s been really well-developed prevention.

DUBNER: Can you just describe generally what sort of program that is, what you’re talking about?

HUMPHREYS: You’re going into this period of life when kids are about 10, 11, 12 and you’re making investments in their core capacity. You don’t walk in and just say, “Don’t use drugs.” But you’re helping them with things like, how do you learn how to recognize and manage your emotions? What are some ways to cope with challenges? How do you connect with prosocial kids? How do you connect to adults? Getting them connected to community structures that they can do instead of drugs. That might be cultural, civic, religious, athletic organizations, whatever, where there’s joy and there’s fun and there’s connection, but it’s not centered on substances.

DUBNER: Are drugs even part of that conversation?

HUMPHREYS: Absolutely. This is unhealthy, as is alcohol, as is tobacco. But it’s a lot broader than that. And the reason that matters is all the things we worry about with kids, the risk factors for them overlap like 75, 80 percent. So there’s programs like, we want to stop eating disorders, we want to stop depression, we want to stop self-harm, we want to stop smoking. But the thing is, like, why do kids do all those things? And it comes back mostly to these core things of inability to deal with emotion, inability to connect with others. And so you focus on those and then, for one kid the benefit is he was going to become a drug user as a teen, he won’t. But if the other kid, that kid was never going become a drug user, but she was going to be really depressed or have bad body image and self-harm, and she’s not going to do that. So you see all these benefits as people go through it — and 10-year studies now and more likely to graduate from college, less likely to be carrying a gun as a teenager. So those are the kind of investments we need to make in kids. It’ll help our drug problem. It’ll help a lot of other problems too.

Keith Humphreys plainly has a lot of ideas about how to best fight the opioid epidemic. Many of them involve government oversight and support — which, given his policymaking background, makes perfect sense. Next time on the show, we’ll get a different perspective:

LOYD: I think the opioid epidemic is this generation’s H.I.V. and AIDS. If you paralleled that to what we’re seeing with the opioid epidemic, there are so many similarities. What’s the biggest thing that prevents people from getting treatment right now? Stigma.

Also: with billions of dollars of opioid settlement money now flowing to states and cities, how is it going to be spent?

Christine MINHEE: To use these monies to replace the status quo would be the largest travesty I could imagine from a financial perspective, and also from a victim’s rights perspective.

That’s next time, in the second part of our series about the continuing opioid epidemic. Until then, take care of yourself — and, if you can, someone else too.

*      *      *

Freakonomics Radio is produced by Stitcher and Renbud Radio. This episode was produced by Alina Kulman, with help from Ryan Kelley. Our staff also includes Augusta Chapman, Dalvin Aboagye, Eleanor Osborne, Elsa Hernandez, Gabriel Roth, Greg Rippin, Jasmin Klinger, Jeremy Johnston, Julie Kanfer, Lyric Bowditch, Morgan Levey, Neal Carruth, Rebecca Lee Douglas, Sarah Lilley, Theo Jacobs, and Zack Lapinski. Our theme song is “Mr. Fortune,” by the Hitchhikers; and our composer is Luis Guerra. As always, thank you for listening.

HUMPHREYS: If the criminal justice system were a parent, and it wanted its child to clean up its room, it would say, “Johnny, if you don’t clean up your room, there’s a 40 percent chance that 6 months from now, I will ground you for a decade.”

Read full Transcript


  • David Cutler, professor of economics at Harvard University.
  • Travis Donahoe, professor of health policy and management at the University of Pittsburgh.
  • Keith Humphreys, professor of psychiatry and behavioral sciences at Stanford University.
  • Stephen Loyd, chief medical officer of Cedar Recovery and chair of the Tennessee Opioid Abatement Council.



Episode Video