It will surprise no one to learn that William Barr, who made it clear when Donald Trump picked him to succeed Jeff Sessions as attorney general that he favored strict and uniform application of federal pot prohibition, and John Walters, who ran the Office of National Drug Control Policy during George W. Bush’s administration, think “legalizing recreational marijuana” has been “nothing short of a disaster.” Reason’s Katherine Mangu-Ward already has ably rebutted their recentFree Press piece making that case. I’d like to add a few points about their approach to the subject, which combines valid concerns with strawman arguments, cherry picking, illogical inferences, reliance on dubious estimates, and tendentious interpretations of contested research.

Barr and Walters complain that marijuana legalization has “created the false perception that the drug is ‘safe.'” They think refuting that false perception is enough to justify a return to prohibition. Because “marijuana is dangerous,” they say, “legalizing it was a mistake.” But the question is not whether marijuana is “safe”; it is whether marijuana’s hazards justify the use of force to stop people from consuming it. Barr and Walters fail to seriously grapple with that question even in utilitarian terms, and they completely ignore moral objections to criminalizing conduct that violates no one’s rights.

It easy enough to show that marijuana, like every other drug, has risks as well as benefits. But that banal observation is not enough to clinch the case for prohibition even if, like Barr and Walters, you ignore the claim that adults have a right to weigh those risks and benefits for themselves.

Alcohol, after all, is assuredly not “safe.” By several important measures, it is substantially more dangerous than cannabis. A lethal dose of alcohol is roughly 10 times the effective dose. Given the dearth of fatal reactions to cannabis among humans, that ratio is difficult to calculate for marijuana. But based on research with laboratory animals, it is more than 1,000 to 1. Alcohol abuse results in potentially lethal organ damage of a kind that is not seen even in the heaviest cannabis consumers. Alcohol is more strongly associated with violence than cannabis, and it has a much more striking impact on driving ability.

Alcohol is nevertheless a legal drug, which reflects a judgment that the costs of prohibiting it outweigh the benefits. It is not clear whether Barr and Walters disagree with that judgment, since they do not mention alcohol at all. In fact, they seem keen to avoid any interdrug comparisons that might undermine the premise that marijuana should be banned because it is especially dangerous.

Barr and Walters warn that “THC, the psychoactive component in cannabis, produces a high by altering brain chemistry and interfering with the nervous system’s normal functioning.” The same could be said of any psychoactive substance. That description tells us nothing about marijuana’s relative hazards.

Back in 1988, Francis Young, the Drug Enforcement Administration’s chief administrative law judge, deemed such comparisons relevant in assessing how marijuana should be classified under the Controlled Substances Act. “Marijuana, in its natural form, is one of the safest therapeutically active substances known to man,” he observed. “There are simply no credible medical reports to suggest that consuming marijuana has caused a single death.”

By contrast, it was well-established that both over-the-counter and prescription drugs could kill people when consumed in large doses. For aspirin, Young noted, the ratio of the lethal dose to the effective dose was about 20 to 1, while the ratio for many prescription drugs, such as Valium, was 10 to 1 or even lower. With marijuana, he said, that ratio “is impossible to quantify because it is so high.”

Barr and Walters would have us believe that Young’s assessment is outdated because today’s “hyperpotent marijuana” is radically different from the drug that had been studied at the time. Yet the Department of Health and Human Services (HHS) recently echoed Young’s basic point.

Explaining its rationale for rescheduling marijuana, HHS noted that “the risks to the public health posed by marijuana are low compared to other drugs of abuse,” such as heroin (Schedule I), cocaine (Schedule II), benzodiazepines like Valium and Xanax (Schedule IV), and alcohol (unscheduled). Although “abuse of marijuana produces clear evidence of harmful consequences, including substance use disorder,” it said, they are “less common and less harmful” than the negative consequences associated with other drugs. It concluded that “the vast majority of individuals who use marijuana are doing so in a manner that does not lead to dangerous outcomes to themselves or others.”

This does not mean increased potency poses no challenges. As anyone who was accustomed to smoking an entire joint or bowlful of crappy pot in college could testify, the high-THC strains and concentrates available in state-licensed pot stores require more caution. For occasional consumers, a few puffs is generally enough. But in a legal market, consumers can make that adjustment based on readily available information as well as personal experience. It is not different in kind from the dosing decisions that millions of Americans make when they consume alcoholic beverages that vary widely in potency.

Instead of considering the typical behavior of cannabis consumers, as HHS did, Barr and Walters focus on problem users. “It’s conservatively estimated that one in three people who use marijuana become addicted,” they aver, linking to a page of information from the Centers for Disease Control and Prevention (CDC). “One study estimated that approximately 3 in 10 people who use marijuana have marijuana use disorder,” the CDC says.

The CDC is referring to a 2015JAMA Psychiatry study based on data from the National Epidemiologic Survey on Alcohol and Related Conditions. The researchers compared survey results from 20122013 to survey results from 20012002. Inconveniently for Walters and Barr, who argue that legalization has led to an explosion in problematic use, the analysis found that “the prevalence of marijuana use disorder among marijuana users decreased significantly” during that period, from 35.6 percent to 30.6 percent. Although the first state-licensed recreational dispensaries did not open until 2014, 17 states and the District of Columbia had legalized medical use by 2013, and some of those laws (such as California’s) were permissive enough that pretty much anyone could obtain the requisite doctor’s recommendation.

Barr and Walters equate the survey-based definition of “marijuana use disorder” with addiction. But the former term encompasses a wide range of problematic behavior, including “abuse” as well as “dependence.”

The JAMA Psychiatry study defined “abuse” as meeting one or more of four criteria: 1) “recurrent substance use resulting in a failure to fulfill major role obligations at work, school, or home”; 2) “recurrent substance use in situations in which it is physically hazardous”; 3) “recurrent substance-related legal problems”; and 4) “continued substance use despite having persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of the substance.”

These are all problems, but they are problems of different kinds, and they do not necessarily signify addiction as that term is generally understood. If someone swam, drove, or hiked a mountain trail while high a couple of times, for example, that could be enough to qualify for the “abuse” label under the second criterion.

The study defined “dependence” as meeting three or more of six criteria: 1) tolerance, 2) taking the substance “in larger amounts or over a longer period than intended,” 3) “a persistent desire or unsuccessful efforts to cut down or control substance use,” 4) spending “a great deal of time” on “activities necessary to obtain the substance, use the substance, or recover from its effects,” 5) forgoing or reducing “important social, occupational, or recreational activities&helli;because of substance use,” and 6) continuing use “despite knowledge of having a persistent physical or psychological problem that is likely to have been caused or exacerbated by the substance.”

Now we are getting closer to the conventional understanding of addiction. But equating any three of these criteria with addiction is still questionable. If a regular marijuana user found that he needed a larger dose to achieve the same effect, sometimes went one toke over the line, and decided to get high instead of going out with friends, for example, he could be deemed “dependent” under this test. More generally, critics of applying psychiatric diagnoses based on survey responses have noted that such data may result in overestimates because they neglect “clinical significance.”

Despite these limitations, Barr and Walters conflate dependence/addiction with a much broader category of marijuana-related problems, and they deem the resulting estimate “conservative.” That one-in-three past-year estimate is much higher than the lifetime dependence risk that a 1994 study calculated based on the National Comorbidity Survey: 9 percent for cannabis, compared to 32 percent for tobacco, 23 percent for heroin, 17 percent for cocaine, and 15 percent for alcohol. It is also at odds with a detailed 2010 analysis inThe 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.

I have just devoted half a dozen paragraphs to one dubious claim out of many in the Barr and Walters piece. As Mangu-Ward notes, they also gloss over the vigorous debate about the nature of the connection between marijuana and psychosis, ignore countervailing evidence regarding the alleged impact of marijuana on IQ, and erroneously equate any level of THC in a driver’s blood with impairment.

Barr and Walters cite the persistence of black-market marijuana in states such as California as evidence that legalization cannot work when it is actually evidence that high taxes and burdensome regulations make it hard for licensed businesses to compete with unauthorized dealers. They likewise blame burglaries and robberies of dispensaries on legalization when the actual problem is the barriers to financial services created by continued federal prohibition, which force those businesses to rely heavily on cash.

Barr and Walters note that marijuana smoke contains “many of the same toxic and carcinogenic chemicals” as tobacco smoke, falsely implying that it is equally carcinogenic. In addition to differences in the composition of marijuana and tobacco smoke, the dose has to be considered: Given typical patterns of use (say, an occasional joint vs. a pack a day), cigarette smokers are exposed to much higher amounts of toxins and carcinogens than marijuana smokers. And Barr and Walters do not even acknowledge smoke-free alternatives such as vaping and edibles.

Barr and Walters cite increases in “marijuana-related ER visits” without considering how legalization might affect people’s willingness to seek treatment or to identify themselves as cannabis consumers. They mention increases in “adolescent cannabis abuse” during “the past two decades” without acknowledging the lack of evidence that legalization has increased underage consumption.

Taking a stab at cost-benefit analysis, Barr and Walters cite a laughably bad Centennial Institute analysis that supposedly showed “every dollar of cannabis-related tax revenue [in Colorado] has been offset by $4.50 in costs due to marijuana-related traffic fatalities, hospital care, and lost productivity.” In assessing the costs of marijuana use, such as health care expenses stemming from “physical inactivity” and lost productivity related to dropping out of high school, that report conflated correlation with causation. It counted tax revenue as the only benefit of legalization, ignoring the expansion of liberty and the boost in consumer satisfaction as well as the criminal justice and law enforcement benefits. Most egregiously, the study did not even attempt to measure how legalization had affected the negative outcomes it tallied.

Barr and Walters likewise see only costs from legalization, which they systematically exaggerate. “Greater marijuana use has contributed to the steady erosion of the civic responsibility, self-discipline, and sobriety required of citizens to sustain our system of limited government and broad personal liberty,” they write. “A doped-up country is a nation in decline.”

As Barr and Walters see it, “broad personal liberty” requires the state to dictate which psychoactive substances people may consume, asserting the authority to control their brains by controlling the drugs they use. That is a counterintuitive view, to put it mildly. Barr and Walters never even broach an issue that is central to this debate: When and why is it moral to deploy the threat and use of violence against peaceful individuals because you disapprove of how they get high?