Jonathan Caulkins is Stever University Professor of operations research and public policy at Carnegie Mellon University. Peter Reuter is distinguished university professor in the School of Public Policy and department of criminology at the University of Maryland.
Ever since opioid deaths suddenly peaked and began to decline in summer 2023, plunging by nearly half over the next two years, experts have been puzzling over why this blessed event occurred. Was it improved public health policies? Increased enforcement? The previous deaths of so many fentanyl users? Or something else?
In an article published Thursdayin Science, we make a case, along with several colleagues, for one major contributor: a reduction in the supply of fentanyl, which was associated with 90 percent of all fatal opioid overdoses in 2022.
Three lines of evidence point in this direction.
First, multiple indicators of fentanyl supply turned downward at the same time deaths did. Notably, the purity of fentanyl powder fell in the second half of 2023 after rising steadily for several years; that meant consumers were getting less fentanyl — and thus running a lower risk of death — from any given purchase. The amount of pure fentanyl in counterfeit pills also fell at that time, as did the number and weight of drug seizures, despite intense federal enforcement.
The second indicator is a novel one inferred from social media. We searched Reddit for references to the term “drought” in conjunction with various drugs. Mentions of a “fentanyl drought” were rare before May 2023 but shot up 1,400 percent by November of that year, an increase that did not apply to other drugs.
Third, fentanyl deaths, seizures and purity declined in Canada at about the same time as they did in the United States. The two countries have very different drug policies, with different approaches to treatment, harm reduction and domestic law enforcement. What they share is certain aspects of the fentanyl supply chain, in particular a dependence on precursors imported from China.
These lines of evidence all point to a supply shock, although other factors could have contributed to a certain extent. Competing theories mostly emphasize “demand side” factors, such as the increased availability of the opioid antagonist naloxone or improved access to treatment with buprenorphine. Some suggest that so many people who use fentanyl have died already that there are now simply fewer using the drug. All these might have contributed, but demand side factors would not be expected to affect the two countries in parallel or to create such a sudden decline, both because they were all already in place before the change and because opioid use is so hard to give up. In addition, declining purity points to reduced supply rather than reduced demand.
There are many sorts of supply interventions. Locking up dealers gets the most attention, and it has a deservedly poor reputation because of long sentences being applied indiscriminately not just to kingpins but also to their unskilled staff and to retailers. Over the decades, however, there have been a number of instances when other types of supply disruption — such as closing down a major supply route to Australia in 2000 — have been accompanied by appreciable declines in drug-related harms.
What might have delivered a “supply shock” to North American fentanyl markets in 2023? We do not know, but we can offer a provocative conjecture. The U.S. fentanyl market is supplied primarily by Mexican refiners, who work with precursor chemicals imported from China, home of the world’s largest chemical industry. Canada, by contrast, imports relatively little fentanyl from Mexico (or the U.S.). It relies primarily on its own domestic refiners, who also rely on precursors produced in China.
Because fentanyl overdoses declined abruptly and substantially at about the same time in both Canada and the U.S., one possibility is that the supply of precursors from China was disrupted for refiners in both Mexico and Canada. The Chinese government is hardly transparent on such matters, but we have pieced together evidence that is consistent with a crackdown against producers and exporters of precursor chemicals. Furthermore, the Drug Enforcement Administration noted in April 2025 that its reporting for 2024 “indicates that some China-based chemical suppliers are wary of supplying controlled precursors to international customers, demonstrating an awareness on their part that the government of China is controlling more fentanyl precursors.”
On the other hand, previous efforts to control precursors have not produced such long-lasting disruptions to supply on the street. For example, multiple rounds of controls on various precursors used in methamphetamine production have interrupted supply for only 3 to 12 months (if at all) before producers found another precursor or another source of the old one. The fentanyl reduction has already lasted two years, which is surprisingly long if the origin was disruption of precursors.
Four final notes:
First, supply disruptions are rarely permanent; we should not assume that the reductions will continue much longer.
Second, a supply control success should not change views about the value of demand side interventions. More treatment and better services can help whether supply is expanding, stable or disrupted.
Third is a reminder of the need for robust data systems for monitoring drug markets and supply, not just drug use.
Finally, none of this story has anything to do with blowing up small boats carrying drugs near Venezuela. As far as is known, Venezuela is not involved in the fentanyl supply chain, and the declines began in late 2023 — well before the current administration took office.
The post What’s behind the stunning decline in American opioid deaths? appeared first on Washington Post.
