In a windowless hangar at New York's John F. Kennedy International Airport, dozens of law enforcement officers sift through packages sent through the U.S. Postal Service, looking for fentanyl. Here, one officer finds Oxycodone pills.

How Iron-Clad Is the Iron Law of Prohibition?

In late May 2023, at the National Institute of Standards and Technology in Gaithersburg, Maryland, Edward Sisco opened a package of swabs taken from used syringes and empty bags. Sisco, a research chemist, ran the samples through a mass spectrometer, a device about the size of a washing machine, and got back a chemical profile.

Sisco receives these swabs from needle exchanges in Maryland and four other states, to track what drugs people are using. Those drugs range from illicitly manufactured fentanyl, an opioid that has largely replaced heroin; benzodiazepines, a class of drugs used to treat anxiety; and 20 different cutting agents and adulterants, such as caffeine, quinine, and mannitol. The surveillance system, which the federal agency piloted in October 2021, also frequently turn up drugs that are contaminating the illicit drug supply in the U.S. Among these are xylazine, a veterinary sedative that goes by the street name tranq, short for tranquilizer; clickbait headlines sometimes refer to xylazine as a “zombie drug” because it causes necrotizing wounds, skin ulcers and knocks people out. Last year, Sisco found another animal tranquilizer, medetomidine, in a small percentage of samples. Xylazine has never been approved for use in humans, and neither sedative is a typical opioid, which complicates the use of antidotes, such as naloxone, to reverse overdoses.

More than 100,000 people die every year due to drug overdoses in the United States. President Joe Biden and his drug policy advisor announced, in 2022, a call to “beat the opioid epidemic.” But academics and harm reduction activists have attempted to reframe these deaths not strictly as an opioid crisis but as a crisis of a toxic drug supply. Sisco’s testing aims to make more of the unknowns known. As he explained, getting quick turnaround results to people who use drugs allows them to make smarter choices.

But these observations reflect broader fluctuations of the illicit drug supply, particularly the detection of veterinary tranquilizers in street drugs, and ultimately raise more fundamental questions for public health. If lawmakers follow through on proposed legislation that would criminalize xylazine and make it illegal, then, Sisco said, suppliers could potentially replace it with yet another substance. “Are they going to try to shift to another veterinary sedative?” he said. “We’re in this continual loop of cat and mouse, trying to keep up with whatever the bad guys are deciding to do.”

The phenomenon has a name: The so-called iron law of prohibition. The more intense law enforcement cracks down, the theory goes, the more potent drugs become, in part due to increased risks in transporting and manufacturing. The theory suggests that drug criminalization spurs traffickers to minimize the size of their product, leading to an uptick in potency; a market that favors the most potent supply, arguably, makes drug use less predictable and, in turn, more deadly.

In 2017, two public health experts, Leo Beletsky and Corey Davis, argued that the lethal cycle was both predictable and practically inevitable. The pair looked at data from Prohibition, the federal ban on alcohol in the U.S. in the 1920s and 1930s, which shifted the market from beer and wine to strong distilled spirits, and revisited the idea in the context of opioids. As it turns out, opioids seem to follow a similar trajectory in response to War on Drugs policies, shifting from prescribed painkillers to heroin, and then from heroin to more concentrated and compact substitutes such as fentanyl.

The iron law of prohibition is nothing more than an “opinion with no empirical foundation.”

The iron law of prohibition has been cited in medical journals, legislative hearings, and scholarly treatises, and it has real-world policy implications. If the theory holds true, then it suggests that supply-side efforts to curtail drugs will never accomplish their intended goals and, instead, will exacerbate their lethal unintended consequences. By invoking the iron law, Beletsky and Davis questioned the punitive drug regime: Why should anyone expect criminalization, penalties, and funding law enforcement to combat the current crisis and save lives?

While researchers in the field of addiction medicine cite Beletsky and Davis’s canonical 2017 paper, evidence supporting the iron law of prohibition remains mixed, and data supporting the overarching claim is ambiguous. As such, many experts balk at the immutability implied by the law’s name. Keith Humphreys, a professor of psychiatry and behavioral sciences at Stanford University, points out exceptions, most notably the increasing potency of legal cannabis and alcohol content of wine, as two examples when the repeal of a drug prohibition apparently had similar effects as the enactment of prohibitions.

“The Iron Law of Prohibition is looking pretty rusted out these days,” Humphreys wrote on X, the platform formally known as Twitter, in response to a posted graph depicting the rising concentration of THC. In an email to Undark, he elaborated: The law is nothing more than an “opinion with no empirical foundation.” Jonathan Caulkins, a professor of operations research and public policy at Carnegie Mellon University’s Heinz College, described it in simpler terms: “Bogus.”


The phrase “iron law” descends from Das Göttliche, a hymn by the late 18th century poet and naturalist Johann Wolfgang von Goethe, and it’s come to mean firm, impregnable, unchangeable, something observable that just happens. Roughly 200 years later, economist Richard Cowan introduced the concept of the iron law of prohibition in a 1986 essay on drug policy and the demand for crack. (Cowan later became executive director of NORML, the National Organization for the Reform of Marijuana Laws.)

Like the national Prohibition on alcohol, Cowan argued that outlawing crack had created incentives to minimize volume to evade law enforcement, and maximize profits. Regulation, he wrote, affected the market in perverse ways (italics in original): “The iron law of drug prohibition is that the more intense the law enforcement, the more potent the drugs will become.”

In 1991, another economist, Mark Thornton, published a book and an analytic paper with the Cato Institute, the libertarian think tank. Neither commentary appeared in a scientific publication, lending credence to a common criticism: The iron law is rooted in economics, a soft science, and the iron law of prohibition originated, first and foremost, as a polemical invective. As Humphreys put it, the concept was “formulated to give a scientific veneer to antiprohibition sentiments.”

In 1923 during the Prohibition era, revenue agents in Washington, D.C. raid a speakeasy. Data suggest the potency of alcohol increased 150 percent pre- and post-Prohibition, a phenomenon which experts say lends credence to the iron law of prohibition. Visual: Bettmann via Getty Images

But while early critics published in openly partisan forums, Beletsky and Davis’s paper placed the concept in a more citable form, resurfacing the iron law in a peer-reviewed scientific journal, the International Journal of Drug Policy. Beletsky, a professor of law and health sciences at Northeastern University, and Davis, director of the Harm Reduction Legal Project at the Network for Public Health Law, laid out the evidence in a short research article. As Davis remembered it, the paper came together rather quickly, and part of its enduring appeal, he said, is that the analyses were not particularly nuanced.

When nothing was done to address the demand of illicit drugs, Beletsky and Davis wrote, efforts to curtail the supply backfired — increasing the harm to people who use drugs in ways that are “as predictable as they are disastrous.” Like their predecessors, Beletsky and Davis cite Prohibition as the prototypical example. Data suggest the potency of alcohol increased 150 percent pre- and post-Prohibition. “There’s really no question that that’s true about alcohol Prohibition,” Davis said. “And then, to me, again, my sense is that that general phenomenon often happens in other sort of criminalized contexts as well.” The iron law arguably helped explain why potent synthetic opioids, such as fentanyl, replaced heroin in North America. Given the barriers to treating opioid use disorder, Beletsky and Davis wrote, it was hardly surprising that crackdowns did not have the intended effect.

Law enforcement officials might assume that getting rid of lethal drugs that are killing people follows the same principle as getting guns off the streets, Davis said, but the 2017 paper argued the opposite: “If your goal is to reduce drug use, reduce drug-related harm, improve people’s lives, I think the evidence is pretty clear that it’s not doing that.” Knowingly pursuing a failed strategy that arguably drove up deaths, he added, was ethically wrong.

In Davis’s view, they had invoked the iron law of prohibition as neither a provocation nor a hyperbolic explanation of it all. “It’s just another example of where prohibitionist policies just generally don’t work,” he said. The paper clearly resonated. Six years later, it’s been cited over 200 times.


Other researchers contend that the iron law of prohibition is rife with inconsistencies. Observational data do not explain every ebb and flow in potency across the board, for example, especially when there is no change in a drug’s legal status. One example, according to Humphreys, is how tobacco has changed over time despite the fact that it is legal. “Why do cigarette manufacturers allow enriched nicotine and sugar to contaminate tobacco?” he wrote. “And the answer is they make more money that way.”

When it comes to opioids, the potency may have increased in recent decades. But Humphreys and other critics of the iron law stress that neither the introduction of high-dose opioid painkillers in the 1990s, nor the subsequent shift from heroin towards illicitly manufactured fentanyl, could be directly attributed to prohibitionist policies alone. (After all, painkillers like Oxycontin were legal; both heroin and illegal fentanyl remained prohibited substances.) Similarly, the concentration of THC, the primary psychoactive compound in cannabis, climbed in the U.S., despite its legalization in many states.

The driving force behind these shifts in potency was pure and simple profit, Humphreys said in an emailed interview, and not the substance’s illegality. “So yes,” he said, “the potency of drugs rises over time, but it’s never had anything to do with prohibition.”

Critics also say that the iron law of prohibition lacks sufficient nuance. Sarah G. Mars, an associate professional researcher at the University of California, San Francisco, cited Beletsky and Davis’s paper in a 2018 report on shifts in the illicit market. In a recent emailed interview, Mars said that the law may be true when it comes to increasing a drug’s potency. Increasing the potency means that more doses can fit in much smaller container, which is easier to ship and to conceal. (Mars’ colleague at UCSF and coauthor, Daniel Ciccarone, for instance, estimated that all the illicitly manufactured fentanyl coming into the U.S., in 2016, could fit into an estimated 10 industrial-sized barrels.) But she stressed that potency alone did not account for the proliferation of fentanyl.

Fentanyl is a synthetic drug which anyone with the right skill set and tools can make in a lab. Unlike heroin or opium, which derive from poppies that are grown out in open fields, fentanyl is not vulnerable to the effects of weather, climate, or disease. As such, it’s much cheaper to produce, Mars wrote, and its production is not geographically constrained. Fentanyl is also more difficult to identify compared to black-tar heroin since it is a white powder lacking any “distinctive aroma or appearance,” Mars wrote, that would distinguish it from other powders.

Packs of menthol-flavored and other types of cigarettes are displayed for sale in a Los Angeles smoke shop in 2022. The iron law of prohibition doesn’t seem to apply to tobacco, which has changed over time despite the fact that it is legal. Visual: Mario Tama/Getty Images
Cannabis products are sold at a dispensary in Portland, Oregon shortly after the state legalized recreational use of the drug in 2015. The concentration of THC, the primary psychoactive compound in cannabis, has climbed in the U.S., despite its legalization in many states. Visual: Josh Edelson/AFP via Getty Images
In September 2016, Eric Schneiderman, then New York Attorney General, speaks during a press conference regarding a record drug bust in which 33 kilograms of heroin and 2 kilograms of fentanyl were seized. According to Keith Humphreys and other critics of the iron law, the shift from heroin towards illicitly manufactured fentanyl cannot be directly attributed to prohibitionist policies alone. Visual: Drew Angerer/Getty Images

Then there are the counterexamples to the iron law of prohibition. In an emailed interview with Undark, Caulkins said that, prior to 2014, efforts to control illicitly manufactured fentanyl likely saved lives. More recently, much of the fentanyl arriving in the U.S. comes as 120 mg pills, which, according to evidence from drug seizure data, contains only an average of about two milligrams of pure fentanyl. If drug trafficking organizations, or DTOs, were moving product with less than 2 percent purity — these pills are almost entirely something other than fentanyl — then the iron law grossly overstated the importance of maximizing potency. “If it was important to the Mexican DTOs to avoid that extra bulkiness,” Caulkins wrote in an email, “they’d press the pills on this side of the border.”

With so few examples where a prohibited substance was made legal and highly regulated, like the 21st Amendment repealing Prohibition, Caulkins maintains the so-called iron law is not comprehensive, focusing on how the intensity of law enforcement drives up potency and price and neglecting other important factors. While the cost of a bag of powder can increase as police make that substance harder to get, the inverse also happens, he said, where stringent enforcement leads to high prices and reduced purity. “So,” Caulkins wrote via email, “one can, in a biased way, pick and choose your examples to create the illusion that there is evidence in support of that weaker form of the Iron Law, but that’s just advocating for a preferred conclusion, not being academically honest.”

Humphreys echoed this deeper epistemological rift, accusing proponents of being intellectually dishonest. “If you’re going to call it an iron law,” he said, “then it has to hold everywhere.”


The differing views of the iron law of prohibition recently spilled into the public at a March 1, 2023 hearing before the House Judiciary Subcommittee on Crime and Federal Government Surveillance. A grieving mother, whose son died of a fentanyl overdose, urged lawmakers to act to pass a law that, she said, would arguably save lives. Jeffrey Singer, a senior fellow at the Cato Institute, spoke next. Singer said it was wrong to label deaths from drug overdoses as an epidemic or an invasion.

“Fentanyl is not a viral pathogen that jumps from host to host,” he said, according to a C-SPAN video. “The influx of fentanyl is a response to market demand. More crucially, fentanyl is just the latest manifestation of what drug policy analysts called the iron law of prohibition. The shorthand version of the law states: The harder the law enforcement, the harder the drug.”

As the hearing progressed, one Congressional representative, Troy Nehls of Texas, responded, directly challenging Singer. “Doctor, I was a law enforcement officer,” Nehls said. “I can tell you, sir, in my humble opinion, you, sir, are wrong.” Projected on a display behind Nehls was a mug shot of a man he described as coming from Mexico, who had been intercepted and arrested with 10,000 tablets of fentanyl, which, Nehls claimed, amounted to 25,000 lethal doses of the drug. “Are you telling me, sir, that we should not have saved those 25,000 lives that this fentanyl would have killed?”

“If you’re going to call it an iron law, then it has to hold everywhere.”

In the months that followed, federal legislators responded with a resounding no. In May 2023, the U.S. House of Representatives voted 289-133 to pass the HALT Fentanyl Act (HALT stands for “Halt All Lethal Trafficking”). The act would step up law enforcement and increase criminal penalties. Diane Goldstein, a retired police lieutenant and executive director of the Law Enforcement Action Partnership, whose brother died of an overdose, said, “I have a ton of empathy for people who have lost their loved ones to drug overdoses, but we have to start talking about it in a way that is meaningful and actually works. We know the modern-day drug war hasn’t worked, otherwise we’d be drug-free America right now.” Lawmakers had once again doubled down on punitive measures, assuming that interrupting the supply of deadly drugs would still have its intended effect: saving lives.

At least on the city scale, a recent study published in the American Journal of Public Health, suggests that’s not the case. Funded in part by the Centers for Disease Control and Prevention, the paper tested the competing claims, and found that efforts to interrupt the supply were associated with an increase in fatalities. The authors looked at data collected over two years from Marion County, Indiana, which includes Indianapolis, and found that when police seized opioids, fatal overdoses spiked in the immediate vicinity. “They increased fatal overdoses at one week, two weeks, and 21 days,” Goldstein said. “So how is that saving lives?” These unintended consequences of drug busts — that is, the resulting spike in drug deaths — persisted for three weeks. Presumably, the authors write, customers could not go to a trusted supplier and those with a dependency took more risks seeking out a less familiar substance.

As one of the study’s co-authors, Jennifer Carroll, a medical anthropologist at North Carolina State University, said, “The deaths we are seeing come not from its potency per se, but from the fact that people cannot dose properly,” she said. “So if you disrupt the drug market, people are going to die because you are also disrupting the only piece of knowledge that helps them stay alive.” The study cites Beletsky and Davis; as Carroll put it, their 2017 paper clearly distilled the argument: “We have fentanyl because of prohibition.”


Some evidence suggests that the so-called iron law of prohibition provides too simple of an explanation for complex social behavior. Beletsky and Davis’s 2017 paper stood as a counterpoint to the status quo, which, the two characterized as the “almost singular focus on suppression of opioid access.”

Data support these claims: More arrests are made for drug offenses than any other category of crime. Meanwhile, according to the 2021 National Survey on Drug Use and Health, less than 10 percent of the estimated 46 million people with substance use disorders receive any treatment. In 2010, the Drug Enforcement Agency warned that reformulating painkillers could contribute to a shift towards heroin use; CDC data show a rise in heroin overdose deaths. One report describes a 45 percent increase in such deaths from 2010 to 2011 alone, the largest annual percentage increase in over a decade.

Authorities cracked down, knowing that such interventions ran the risk of pushing people with drug dependencies to use deadlier drugs. Beletsky and Davis cite one public health official in Pennsylvania who said officials just had to start somewhere. “These interventions are problematic not only because they are often counter-productive, but also because, under the semblance of decisive action, they crowd out evidence-driven measures,” the researchers wrote. “Every dollar spent on enforcement is a dollar not spent on treatment, harm reduction, or prevention. As we failed to invest in what works, the crisis has mutated into something far more deadly.”

Photos of fentanyl victims are on display at The Faces of Fentanyl Memorial at the U.S. DEA headquarters in September 2022. The next year, in May 2023, the U.S. House of Representatives passed the HALT Fentanyl Act, once again doubling down on punitive measures and assuming that interrupting the supply of deadly drugs would still have its intended effect: saving lives. Visual: Alex Wong/Getty Images

As Beletsky put it in a more recent interview with STAT about the “drug bust paradox” study in Indianapolis, politicians had essentially accepted misinformation as gospel truth: “If we crack down on drug dealers, that’s the way that we’re going to reduce overdose deaths.” The market forces under prohibition contradict that, and yet Beletsky has also acknowledged that the reasons people take drugs are more complex than a rigid economic law can describe. “I think, as with all theories, there are sort of inconsistencies. But overall, I think that the rule of thumb holds and it’s a useful frame to apply,” Beletsky said, in an interview in July 2022 on the podcast Crackdown. “It is not an ironclad law.”

For his part, Davis said the theory never pretended that prohibition, in all its forms, was the sole factor driving drug potency — or, for that matter, the only factor exacerbating all the harms associated with illicit drug use. Nor did the researchers claim, in borrowing the name “the iron law,” that it necessarily applied to all the pressures potentiating the strength of all substances, Davis said. “To me, it’s not like a scientific rule that’s inviolable, but it’s just like a common-sense observation about how the world works,” he added.

The rhetorical shorthand of the iron law of prohibition communicated a broader critique. Peter Reuter, a professor of public policy and criminology at the University of Maryland, said he disliked the term, but agreed that, while overly simplistic, it reflects a generalizable phenomenon. “My father was a food scientist,” Reuter said. “I learned early on from him — ‘Good is what you’re used to.’” Similarly, research generally supports the phenomena, suggesting people generally stick with trusted dealers, and, Reuter said, initially there was little apparent demand for fentanyl, but nonetheless potent substitutes replaced these drugs and proliferated across North America.

As Beletsky put it in a more recent interview with STAT, politicians had essentially accepted misinformation as gospel truth: “If we crack down on drug dealers, that’s the way that we’re going to reduce overdose deaths.”

Back at the NIST laboratory in Maryland, Sisco routinely detects fentanyl, most often in combination with xylazine and caffeine, but, in 496 samples from 2021 and 2022, the results were far from uniform. His lab identified 174 unique combinations of various substances. Were all these shifts in the opioid supply the logical result of intensified law enforcement? (Sisco said it was not his place to comment.) Was it some factory unloading veterinary sedatives at a steep discount? Did xylazine make the experience of using fentanyl feel more like heroin?

That last possibility seems the most likely according to a trio of researchers, who spoke with people who use tranq in Philadelphia and Tijuana, and heard anecdotally that xylazine restored “some of the qualities lost when fentanyl’s increasingly displaced heroin.” Xylazine was a cutting agent, diluting fentanyl with a cheaper, legal substitute that likely reduced the risk of a fatal fentanyl overdose and, as such, possibly decreased police attention around overdoses. As such, tranq seemed at best a messy example but also a direct contradiction of the supposed iron law: In an effort to avoid law enforcement, the market offered up a weaker retail product, but one that still harmed consumers.

Sisco’s testing of used syringes has also revealed the anesthetic medetomidine in combination with fentanyl and xylaxine. The drug was found in only a small number of samples, but questions remain: Would outlawing xylazine invariably lead to more potent drug combinations? Did medetomidine — or other new synthetic substitutes, for that matter — follow a deadly path suggested by the iron law of prohibition? “It doesn’t seem like we’re replacing xylazine yet,” Sisco said. “It seems like it’s getting included for another reason.”

Although the iron law suggests that prohibition and the intensification of law enforcement makes drugs more potent, one of the only certainties in the illicit market seemed to be its unpredictability. As Reuter put it, the iron in iron law implies something too rigid with too direct an effect. The concept has its merits, Reuter said, but needs a rebranding, one a bit softer and with a more malleable bent. His suggestion: “It’s a copper law.”

Peter Andrey Smith is a senior contributor at Undark. His stories have also been featured in Science, STAT, The New York Times, and WNYC Radiolab.