Detailing the 'Real Problem With Legal Weed,' a Critic Grossly Exaggerates the Prevalence of Pot Addiction

The Manhattan Institute's Charles Fain Lehman misleadingly equates a survey's measure of "cannabis use disorder" with "compulsive" consumption that causes "health and social problems."

Why is marijuana legalization such a mess in New York? The answer seems clear: The rollout of licensed dispensaries has been "a disaster," as Gov. Kathy Hochul puts it, because of misguided policies and bureaucratic ineptitude. But Manhattan Institute Fellow Charles Fain Lehman is unsatisfied by that explanation. He prefers one that makes little sense but gives him an excuse to discuss "the real problem with legal weed": that marijuana addiction is more common and severe than people tend to think.

In a recent New York Times Magazine article, Lehman argues that Americans generally do not take the problems caused by excessive or reckless marijuana use seriously enough. While he may be right about that, his own take errs on the side of hyperbole by equating "cannabis use disorder" (CUD), a broad concept encompassing a wide range of behavior, with addiction, an especially severe subset of CUD.

Lehman notes that "ongoing litigation" provoked by New York's licensing preferences has obstructed legalization, that "programs meant to give disadvantaged license holders a head start have struggled to secure funds and storefronts," and that "the state's main marijuana regulator, the Office of Cannabis Management, was given almost no enforcement power in the initial law." While "these setbacks can't be helping," he says, "there are flaws in every policy rollout."

New York's pot predicament "seem[s] worse" because "marijuana is addictive," Lehman argues. "Combining addiction with the profit motive creates perverse incentives, letting corporations compete to help people ruin their lives. Once you understand these dynamics, New York's weed problems are no longer confusing—they're obvious." The state's legislators and regulators were "blinded by excitement," he says, seeing marijuana as "a great opportunity with no downside." But "as they are slowly finding out, they were wrong."

As an explanation for the problems that New York has faced in trying to displace the black market, Lehman's thesis is puzzling. After all, marijuana is no more addictive in New York than it is in other states that have managed legalization better. And even if it were, how would that account for the agonizingly slow pace at which legal pot shops have been opening? But Lehman's dubious analysis of the situation in New York is just a pretext for his argument that capitalism and cannabis are a dangerous combination.

Lehman says "around 30 percent of users" consume marijuana "compulsively," thereby "harm[ing] themselves and the people around them." For obvious reasons, pot prohibitionists like to cite that seemingly authoritative figure. But this CUD estimate is highly misleading, especially when it is taken to mean that nearly a third of marijuana users are addicts.

To support that striking claim, Lehman cites two sources. "The Centers for Disease Control and Prevention notes that roughly three in 10 marijuana users qualify as having a 'cannabis-use disorder,' known as CUD," he writes. The CDC, in turn, cites a 2015 JAMA Psychiatry study based on data from the National Epidemiologic Survey on Alcohol and Related Conditions. Lehman also cites the National Survey on Drug Use and Health (NSDUH), which found that "19 million Americans"—about 30 percent of past-year marijuana users—"suffered from cannabis-use disorder" in 2022.

By contrast, a 1994 study based on the National Comorbidity Survey estimated that 9 percent of cannabis consumers experience "dependence" at some point in their lives, compared to 32 percent of cigarette smokers, 23 percent of heroin users, 17 percent of cocaine users, and 15 percent of drinkers. Lehman's annual estimate implies a much higher lifetime risk. It also seems to be at odds with a detailed 2010 analysis in The Lancet, which found that the dependence risks for marijuana and alcohol were similar while rating the overall harm attributable to alcohol more than three times as high.

Although Lehman does not explicitly address the huge gap between his estimate and earlier calculations, he suggests a possible explanation. "According to the National Institute on Drug Abuse," he notes, "average THC concentration has risen from around 4 percent in the mid-1990s to 15 percent in 2021. Legalization has also permitted the production and sale of high-potency concentrates, with THC levels as high as 80 percent. As a result, addiction has become more common, even as public perception has lagged."

A closer look at the NSDUH numbers suggests a likelier explanation. CUD, especially as measured by the survey, is a much broader category than what psychiatrists used to call "dependence," covering a wide range of marijuana-related problems that do not necessarily fit the conventional understanding of addiction.

NSDUH defines CUD based on the latest edition of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-5), which combines what used to be two separate labels: "dependence" and "abuse." Lehman glosses over that change when he says "modern psychiatry tends to characterize addiction—today usually called 'substance-use disorder'—as continued use of a substance in spite of negative consequences." He thereby implies that CUD is the same as "addiction," which is not accurate.

A CUD diagnosis requires "clinically significant impairment or distress," manifested by at least two of 11 criteria:

1. The marijuana user consumes cannabis "in larger amounts" or "over a longer period" than intended.

2. He has a "persistent desire" to reduce his consumption and has unsuccessfully tried to do so.

3. He spends "excessive time" acquiring cannabis, using cannabis, or recovering from its effects.

4. He has "cravings" for marijuana.

5. He neglects "social obligations" as a result of "recurrent use."

6. He continues to use marijuana "despite social or interpersonal problems."

7. His marijuana use leads him to forgo "important social, occupational, or recreational activities."

8. He continues marijuana use "despite physical harm."

9. He continues marijuana use "despite physical or psychological problems" associated with it.

10. He experiences tolerance, requiring larger doses to achieve the same effect.

11. He experiences withdrawal symptoms when he abstains.

NSDUH includes questions that reflect those criteria. It grades CUD as "mild" (two or three criteria), "moderate" (four or five), or "severe" (six or more). In the 2022 survey, 30 percent of past-year marijuana users qualified for the CUD label, which is the number that Lehman highlights. But just 5 percent of past-year users qualified for the "severe" category, compared to about 8 percent with "moderate" CUD and 17 percent with "mild" CUD.

The "mild" category, in other words, accounted for 55 percent of the marijuana users who were deemed to have experienced CUD in the previous year. Given the range of answers that could put someone in that category, it is misleading to say, as Lehman does, that people with "mild" CUD "use compulsively," let alone that their cannabis consumption "harms themselves and the people around them."

If a respondent said that he spent "a great deal" of his time using marijuana and that he sometimes used more than he planned, for example, that would be enough to qualify for the CUD label. Likewise if he reported that he sometimes had "a strong urge" to use marijuana and that he increased his dose to compensate for tolerance. Such answers do not necessarily indicate that someone uses marijuana "compulsively," and they say nothing about whether he harmed himself or anyone else.

Lehman repeatedly implies otherwise. DSM-5, he says, "defines CUD in part as 'an inability to stop using marijuana even though it's causing health and social problems.'" But CUD, as measured by NSDUH, does not necessarily entail "health and social problems" or "an inability to stop using marijuana." Lehman likewise says marijuana "creates a health issue for as many as 30 percent of its users," which is misleading for the same reason.

Equating the NSDUH measure of CUD with addiction is also problematic because the survey does not assess whether respondents experienced "clinically significant impairment or distress," which is a prerequisite for the diagnosis. Critics of applying psychiatric diagnoses based on survey responses have noted that such data may result in overestimates because they do not measure clinical significance.

None of this means that NSDUH respondents who qualify for the CUD label have not experienced marijuana-related problems. But those problems cover a broad range and are typically "mild," contrary to the impression that Lehman leaves.