All podcasts / People I (Mostly) Admire / Summary

28. Professor Carl Hart Argues All Drugs Should Be Legal — Can He Convince Steve?

2026-06-27 - source - Read full transcript

Steve Levitt (host)Carl Hart

Key insights

Hart's legalization plan pairs regulation with access controls, not a free-for-all.

He proposes legally regulating drugs from cannabis to heroin with purity and unit-dose controls to prevent lethal contamination, plus a purchase requirement similar to a driver's license, alongside honest public education about drug combinations and safer routes of administration (snorting versus smoking).

drug-policy-legalization

The large majority of regular drug users, across drug types, do not meet clinical criteria for addiction.

Citing DSM-5 criteria, Hart gives addiction rates by substance: about a third of tobacco users, 25-30 percent of heroin users, 15-20 percent of cocaine users, similar rates for alcohol, and about 10 percent for cannabis - meaning most of the roughly 35 million Americans who use drugs monthly are not addicted.

drug-policy-legalization

Portugal's decriminalization of all drugs is Hart's central real-world evidence for what legalization could do.

Levitt raises Portugal's roughly two-decade-old decriminalization as a natural experiment; Hart says overdose rates dropped significantly and more people sought treatment, supporting his prediction that U.S. legalization would reduce, not increase, overdose deaths.

drug-policy-legalization

Addiction risk is predictable from known life factors, undermining the 'nobody knows who'll get hooked' argument against legalization.

Hart lists concrete predictors - undeveloped coping skills, unrealistic obligations, untreated psychiatric illness or physical pain, and recent loss of a stable job - as the actual drivers of addiction risk, arguing this evidence base should replace fear-based claims that any teenager exposed to legal drugs could become addicted.

drug-policy-legalization

Enforcement of drug law falls disproportionately on Black Americans and specific Latino groups, not minorities broadly.

Both Hart and Levitt agree the criminal-justice burden of drug prohibition concentrates on Black people and select Latino populations such as Mexican Americans and Puerto Rican Americans, a fact Hart treats as central to why he became an advocate rather than staying a lab-only scientist.

race-and-the-drug-war

Hart's heroin disclosure drew outsized, racially inflected media backlash compared to his psychedelic and MDMA use.

He says the same book disclosed multiple types of drug use, but only the heroin headline stuck, because heroin is coded as the 'undesired' drug of the lower classes while psychedelics have become socially acceptable among educated, largely white audiences - a double standard he says mirrors how he is treated as a Black man more broadly, just more openly.

race-and-the-drug-war

Widely cited brain-imaging differences between drug users and non-users are often statistically real but functionally meaningless.

Hart describes methamphetamine-user studies where a small average difference in prefrontal cortex volume is driven by one or two outliers, not the group; on individual cognitive testing, most users perform within the normal range, yet researchers still describe the findings with language like 'brains have atrophied.'

scientific-bias-and-methodology

Cross-sectional drug studies cannot support causal claims about drugs harming the brain, but researchers routinely write as if they can.

Levitt and Hart note that comparing user and non-user brains at a single point in time reveals nothing about whether drug use caused a difference or pre-existing brain differences predisposed someone to use - yet papers use change-implying language ('atrophied,' 'decreased') without any before-and-after data.

scientific-bias-and-methodology

Funding concentration in a single agency biases an entire research field toward one conclusion.

Hart argues that because NIDA is the dominant funder of U.S. drug research, and journal editors and peer reviewers also depend on its grants, the field converges on studying and emphasizing harm even when the underlying data would support a more balanced picture - not through deliberate dishonesty but through incentives that reward finding negative effects.

scientific-bias-and-methodology

Physical and social setting can produce completely different subjective drug effects from pharmacologically near-identical compounds.

Comparing methamphetamine and MDMA (structurally almost identical) in a standard lab setting, Hart's team found mostly overlapping effects; but when he and his wife took the same drugs in a comfortable personal setting, MDMA produced markedly more intimacy and empathy, showing that context, not just chemistry, drives reported experience.

scientific-bias-and-methodology

When you have enough information but still can't decide, minimize worst-case regret rather than optimizing expected value.

Answering a listener question, Levitt says regret is the emotion that sticks with him longest (more than anger or sadness), so his rule is to picture the worst outcome of each option and choose whichever leaves him less regretful if things go badly.

decision-making-under-uncertainty

If even the regret test doesn't resolve a decision, take the option that most changes your circumstances.

Levitt cites his own research, with a partially randomized design, showing people who make larger deviations from their status quo report being happier six months later than those who make smaller changes or stay put.

decision-making-under-uncertainty

Books referenced

Drug Use for Grown-Ups - Carl Hart - Hart's recent book arguing for full legal regulation of all drugs, including heroin, cocaine, and methamphetamine; the book's disclosure of his own heroin use drove most of the press coverage.
High Price - Carl Hart - Hart's earlier memoir/first book, mentioned by Levitt as evidence Hart started his career as a conventional lab scientist, not an advocate.
Freakonomics - Steven D. Levitt and Stephen J. Dubner - Levitt references the book's economic framing that experts' incentives diverge from clients' and can lead experts to take advantage of less-informed people.

Media referenced

America's Math Curriculum Doesn't Add Up - podcast - Freakonomics Radio episode 391, guest-hosted by Levitt, referenced when answering a listener question about how Virginia is restructuring high school math tracks.
DSM-5 (Diagnostic and Statistical Manual of Mental Disorders) - other - Hart cites the American Psychiatric Association's manual for the clinical definition of addiction he uses throughout: psychosocial disruption of obligations plus distress about that disruption.

Companies

Columbia University - Hart is a psychology professor there and sits on its Institutional Review Board, which he says does not obstruct his human drug-administration studies despite public perception that such research would be hard to approve.
National Institute on Drug Abuse (NIDA) - Hart identifies NIDA as the field's dominant funder, whose mission to study almost exclusively the negative effects of drugs biases researchers, journal editors, and reviewers who all depend on its grants.

Techniques and frameworks

Human behavioral pharmacology studies - Hart's lab method: administer controlled doses of drugs like MDMA, methamphetamine, opioids, cocaine, and cannabis to research participants and measure acute and long-term effects on cognition, sleep, food intake, and social interaction to build a factual database.
DSM-5 addiction criteria - Hart's operational definition of addiction: disruption of the ability to meet work, family, or educational obligations, combined with distress about that disruption - distinct from mere regular use.
Worst-case regret minimization - Levitt's personal decision rule when he has good information but still can't choose: imagine the worst-case outcome of each option and pick whichever leaves less regret, since regret is the emotion that sticks with him longest.
Biggest-deviation-from-status-quo heuristic - Levitt's fallback when even the regret exercise doesn't resolve a decision: take the option that most changes your situation, citing his own (partially randomized) research finding that people who make bigger life changes report being happier six months later.

Summary

Steve Levitt sits down with Columbia University neuroscientist and psychologist Carl Hart, whose book "Drug Use for Grown-Ups" argues for legally regulating every illicit drug, including heroin, cocaine, and methamphetamine, rather than just softer positions like marijuana legalization. Levitt, who has taught drug policy to undergraduates for two decades and long favored legalizing marijuana while opposing legalization of harder drugs, treats the conversation explicitly as an attempt to see whether Hart's evidence can move him. Hart opens with a concrete plan: legally regulated drugs, purity and dose controls to prevent contamination deaths, a purchase requirement similar to a driver's license, and honest public education about safer use, including drug-combination risks and routes of administration.

Much of the middle of the episode is Levitt and Hart methodically separating facts from conclusions, a distinction Levitt insists on before evaluating any policy argument. They agree that drugs reliably produce pleasure and that most regular users function normally; Hart supplies DSM-5-based addiction rates across substances (roughly a third of tobacco users, 25-30 percent of heroin users, 15-20 percent of cocaine and alcohol users, about 10 percent of cannabis users) to show that addiction is a minority outcome, not the default. They also agree that predictors of addiction, like untreated psychiatric illness, job loss, or overwhelming obligations, are identifiable rather than random, and that drug-law enforcement disproportionately targets Black Americans and specific Latino communities. Portugal's roughly two-decade-old decriminalization, which Hart says reduced overdose deaths and increased treatment-seeking, serves as the episode's main empirical anchor for what U.S. legalization might do.

A recurring thread is Hart's critique of his own field's science. He argues that brain-imaging studies comparing drug users to non-users routinely overstate small, outlier-driven statistical differences as if they were large functional deficits, and that researchers use causal language ("atrophied," "decreased") despite only having single-point-in-time, non-causal data. He traces this to funding concentration: the National Institute on Drug Abuse dominates U.S. drug research funding, and because journal editors and reviewers depend on the same grants, the field converges on emphasizing harm. Hart also describes a striking personal experiment: comparing methamphetamine and MDMA, chemically near-identical, produced overlapping effects in a standard lab setting but wildly different, more intimate effects when he and his wife took the same drugs in a comfortable personal setting, illustrating how much context shapes reported drug experience independent of pharmacology.

The conversation turns more personal when Hart explains why he moved from lab scientist to public advocate: he entered the field in the late 1980s believing drug addiction caused problems in Black communities like his own, then concluded through years of research that this causal story was largely wrong, and felt obligated to say so publicly given the racialized stakes for his own family. He also describes the disproportionate media backlash to his heroin disclosure specifically, versus his acknowledged psychedelic and MDMA use, as evidence of a class- and race-coded hierarchy of which drugs and which users society is willing to tolerate.

In the middle of the episode, Levitt answers a listener question about how to help students make an important curriculum choice, offering two personal decision-making heuristics unrelated to drug policy: when good information still leaves you stuck, minimize worst-case regret; if that fails too, favor the option that most changes your status quo, citing his own research linking bigger life changes to greater happiness six months later. By the end, Levitt reports being genuinely moved: he remains uncertain whether legalizing hard drugs would meaningfully increase the number of heavy users, but says he now thinks a state like Oregon or Colorado should test full drug legalization as a real-world pilot.

Notable Quotes

"When you are standing up for the undesired, it's a lonely place to be, but it's the right place to be and history will show that." - Carl Hart

"This idea that one has to earn pleasure. Pleasure from drugs are seen as unearned pleasure, which is this ridiculous notion." - Carl Hart

"We do, by the way, have pretty good information about the likelihood of who will become addicted and who will not become addicted... That's bullsh*t. We do." - Carl Hart

"I wouldn't have said this an hour ago, but I think it's time for Oregon or Colorado to lead the charge and serve as a guinea pig for legalizing all drugs." - Steve Levitt