In 2016, Rachel Winograd began to see methadone patients who relapsed or left the treatment program where she worked start overdosing and dying at unprecedented rates. The culprit was illicitly manufactured fentanyl, which is generally 50 times as strong as heroin — with some variants an astonishing 5,000 times as potent. Fentanyl had begun to overtake heroin in Missouri.
“We were just seeing people drop like crazy,” said Dr. Winograd. But to her utter shock, staff members did not distribute naloxone, which is also known as Narcan, a nasal spray or injection that can reverse opioid overdose, to try to save their lives.
While fighting to change this policy, she discovered that many counselors, police officers, emergency medical technicians and even some doctors believed that handing out naloxone would do more harm than good. It would “enable” continued addiction and deter treatment, she was told. Or, others said, reducing fatalities would increase risk-taking among people who were already using drugs — and encourage children to try heroin.
Dr. Winograd, who is now the director of addiction science at the University of Missouri-St. Louis’s Missouri Institute of Mental Health, had encountered a concept known as moral hazard, the idea that reducing exposure to the negative consequences of a risk makes people more likely to take that risk.
While this phenomenon is a demonstrable concern for regulators of financial institutions — the 2008 crash is one infamous example — there’s little evidence it holds true in matters of health and safety. Here, moral hazard is far more of a political cudgel than a proven principle. As we face the worst overdose death crisis in American history, we can’t allow moral panic over moral hazard to drive out policies that have proved to save lives.
The University of Chicago economist Sam Peltzman introduced the idea of moral hazard to health policy in 1975. His data, he claimed, showed that seatbelt laws backfire because when drivers feel safer they take more risks, canceling out any benefit. Also known as risk compensation, the concept rapidly caught on as an argument against regulation.
But later research (as well as a continued significant decline in fatalities per mile as safety improvements continued) poured cold water on his conclusions. Researchers occasionally find a small moral hazard effect that is rarely enough to outweigh benefits. However, in most studies in areas as diverse as the influence of bicycle helmets on rider speed and the human papillomavirus vaccine on teen sexual behavior, moral hazard simply isn’t observed.
Despite the evidence, this idea continues to haunt addiction debates — specifically over harm reduction policies like drug decriminalization, programs that provide clean needles to prevent infectious disease and naloxone distribution to reverse overdose.
Some economists claim to have evidence that moral hazard eliminates most positive effects of harm reduction and increases overdose deaths. They use a method called causal inference, which, when its measures are set appropriately, can show cause and effect, unlike the observational research typical in public health.
For example, a 2018 study led by the economist Jennifer Doleac reported that naloxone distribution led to a 14 percent increase in overdose deaths in the Midwest, leading the Washington Post columnist Megan McArdle to endorse its claims of moral hazard. A 2022 study by the Vanderbilt University economist Analisa Packham used similar methods to claim that clean needle programs (which also distribute naloxone) caused a 25 percent increase in opioid-related fatalities.
These findings generated enormous controversy because they run contrary to the overwhelming majority of public health data — as well as to the recommendations of experts from the World Health Organization and the Centers for Disease Control and Prevention. More recent studies should bring humility to those who rely on data to make causal claims about behavior that they do not study directly — and to those who see moral hazard in harm reduction.
Researchers who dug into Ms. Doleac’s data found that it relied on erroneous assumptions about when naloxone availability increased in states that were studied. This is a crucial error, because if naloxone availability didn’t rise when the paper claimed that it had, it could not have caused subsequent overdose deaths. The study also had other flaws that rendered its conclusions unreliable. Ms. Packham’s research also exhibited similar measurement problems, and could not explain why small expansions of syringe exchange programs supposedly caused harm while massive expansions did not.
Moreover, a study of over 1,300 drug injectors published in 2023 found no changes in drug risk behavior after naloxone distribution and education was started. A randomized trial of co-dispensing naloxone with opioid prescriptions in Colorado pharmacies also found no moral hazard effect.
Claims that harm reduction programs encourage teenage drug use by making it less dangerous do not hold up, either. As naloxone access has boomed, misuse of heroin and prescription opioids by high school seniors plummeted. In 2007, 0.9 percent of 12th graders reported taking heroin and 9.2 percent reported misusing prescription opioids; those figures were 0.1 percent and 1 percent in 2023.
To further understand why moral hazard is especially unlikely to affect overdose, it’s critical to know how people with addiction actually behave.
For example, one methadone patient described his periods of active addiction to Dr. Winograd like this: “Look, all the money I have that day I’m going to spend on dope. All the dope I have I’m going to use.” Hansel Tookes, who founded Florida’s first legal syringe services program, shared the same sentiments. “My patients tell me they spend every dime that they have made that day. And then they wake up and they do it again,” Dr. Tookes said.
Stories like these typify the experience of addiction. And this means that even if naloxone did make addicted people more likely to take riskier amounts, it wouldn’t matter because it doesn’t provide the money needed to obtain them. (One would think economists would consider the role of economics.)
But there’s another compelling reason that naloxone doesn’t cause moral hazard, which is evident to anyone who understands the extremely unpleasant nature of having an overdose reversed. Naloxone immediately causes a distressing withdrawal syndrome — the experience people with addiction overwhelmingly seek to avoid.
Dr. Winograd has found that the more educated people are about addiction and the more closely they work with such patients, the less likely they are to endorse moral hazard arguments. The police are more likely to have these concerns than emergency medical technicians, who in turn are more likely to worry about moral hazard than addiction medicine doctors.
While it is possible that under some circumstances, reducing harm might increase risk-taking, this concern should not stand in the way of access to medications proven to save lives. It makes sense to apply the idea of moral hazard to banking — where the data shows that bailing out investors can make financiers more likely to gamble, especially with other people’s money. It does not follow that we should use the theoretical possibility that reducing harm might increase risk-taking in some settings to gamble with people’s lives.
Maia Szalavitz (@maiasz) is a contributing Opinion writer and the author, most recently, of “Undoing Drugs: How Harm Reduction Is Changing the Future of Drugs and Addiction.”
The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: letters@nytimes.com.
Follow the New York Times Opinion section on Facebook, Instagram, TikTok, X and Threads.