America’s Marijuana Debate Is Flawed

Public policy analysts and lawmakers are often asked to determine whether the societal benefits of legalizing recreational marijuana nationwide would outweigh the harms. The short answer is: Who knows and who cares? The question is not only impossible to answer empirically with any sort of precision or credibility, but also beside the point. Moreover, as framed, it invites readers to embrace a number of rhetorical fallacies that further undermine the prospects of shedding any useful light on this discussion. 

To appreciate why the question presented misses the mark, we can ask the same question with respect to various activities that do not carry the social stigma of drugs, or the decades of regulation that tend to imbue drug laws with a false patina of legitimacy. Here are some examples: Do the societal benefits of legalizing recreational gun possession and use outweigh the harms? Do the societal benefits of legalizing recreational swimming outweigh the harms? Do the societal benefits of legalizing recreational sex by people under 25 outweigh the harms? Do the societal benefits of legalizing medical care for retired people outweigh the harms? 

As one of the lawyers who brought the Heller case to the Supreme Court in 2008, I have debated the question of gun ownership countless times, and I’m thoroughly familiar with the arguments on both sides. And guess what? A plausible case can be made that a universal gun ban (if successfully implemented, which it never could be; more on that below) would in fact bring significant societal benefits, including substantially lower rates of suicide and homicide, fewer nonfatal gun injuries, and a tremendous reduction in the amount of anxiety that many people feel about the possibility of falling victim to a mass shooting or other act of gun violence. Does that tell us whether we should legalize or criminalize “recreational” gun ownership? No—and not simply because of the Second Amendment, which did not exist when the colonists used privately owned firearms to run the British out of America. Armed self-defense is a preexisting natural right, not a mere privilege conferred by the Constitution or any other human law. The same may well be true of consuming intoxicants, which has been a consistent practice throughout all human cultures and history.

The next two questions—whether “we” should allow other people to swim or have sex just for fun may seem silly, but they illustrate an important point, which is that some things that other people do for fun are entirely their business, even if those activities produce demonstrable societal harms. According to the Centers for Disease Control and Prevention’s National Center for Health Statistics, 4,790 people died by drowning in America in 2023 (excluding water transport), mostly in swimming pools and natural water environments (as opposed to, say, bathtubs or sinks), which suggests that the majority of those entirely avoidable deaths were associated with recreational swimming. Why on earth do we tolerate such carnage? And one scarcely needs to enumerate the many ways in which allowing young people to have purely recreational sex imposes costs on society, including unwanted pregnancies, sexually transmitted diseases, and untold emotional trauma. Shame on us for “legalizing” such a costly activity by people who lack the maturity to handle it responsibly.

The final illustration is perhaps the most powerful because two things are true about health care spending: Old people consume a vastly disproportionate amount of it, and those dollars could absolutely be spent in ways that provide far greater value to society than simply extending the life of some unproductive retiree by a few months or years. But is that the only—or even the most relevant—question to ask in deciding whether to “legalize” health care spending by retired people? Plainly not.

In general, whether another person chooses to own a gun, go for a swim, have sex, or seek medical care is—and I can’t stress this enough, so I’ll use italics—none of your damn business! Moreover, that remains true even if you can show that those decisions, whether individually or in the aggregate, generate substantial societal costs. And the same goes for a person’s decision whether to drink a beer, lick a toad, or smoke a joint. Those are private choices that you have no more business involving yourself with than other people’s sex lives or health care decisions.

Besides being irrelevant, the question whether legalizing recreational marijuana use would generate more societal benefits than harms is impossible to answer empirically, both because it is exceedingly difficult to reliably quantify either the costs or the benefits, and because the relevant values (fiscal, hedonic, therapeutic, aesthetic, etc.) are fundamentally incommensurable.  

Let’s start with the primary rationale of all prohibitory policies, which is that criminalizing a potentially harmful substance like marijuana means less of it will be consumed and it will do correspondingly less harm than would occur without the prohibitory regime. That argument is not entirely baseless, and indeed there is a scholarly consensus that alcohol prohibition from 1920 to 1933 did decrease overall consumption to some uncertain extent. But just how much of a decrease in consumption of a plant that can easily be grown at home (or even in national parks) can we realistically expect from a prohibitory regime that lacks public support, is scoffed at by jurors to the point where it becomes difficult to indict, let alone convict, people for low-level crimes, including even distribution of marijuana? Good luck trying to estimate how much—if any—less marijuana ends up getting consumed in a prohibitory regime versus a permissive one.

Moreover, prohibition doesn’t simply reduce consumption—it creates an entirely separate category of harms that wouldn’t exist under legalization. Black market violence, product contamination, the enormous costs of enforcement and incarceration, the collateral consequences for families and communities affected by drug arrests, corruption of law enforcement—all of these are direct products of the prohibitory regime itself rather than of marijuana use. When dealers can’t resolve disputes through courts, they use violence. When there’s no regulatory oversight, products are adulterated with dangerous substances and become more potent through the so-called “iron law of prohibition.” When you criminalize a widely desired activity, you create lucrative opportunities for criminal organizations while simultaneously corrupting the justice system through selective enforcement. These prohibition-specific harms must be weighed against any benefits the policy might provide; but again, quantifying them reliably is virtually impossible.

The other side of the ledger fares no better. How is an honest researcher supposed to quantify the net harm inflicted by a prohibitory regime? Again, one can do little more than speculate because there are so many relevant data points that are impossible even to estimate, let alone quantify in any intellectually honest fashion. These include the total number of times people who wished to consume marijuana were discouraged from doing so by the prohibitory regime and thus had to forgo the benefits of that consumption, which can include not “just” pleasure, but also companionship, reduction in pain and anxiety, better quality of life, and, for some people, better sleep and greater creativity. 

Even more problematic is the intractable challenge of measuring substitution effects—the ways in which marijuana prohibition may push people toward more dangerous alternatives. Do people drink more alcohol when marijuana is illegal? Turn to prescription drugs for anxiety or pain relief? Use more harmful illegal substances? Some preliminary evidence from legalized states suggests alcohol consumption may decline where marijuana is legal, which would represent a significant public health benefit given alcohol’s well-documented harms. But these effects are nearly impossible to quantify reliably, involve long time lags, and vary dramatically across different populations. Any honest cost-benefit analysis would need to account for these substitution effects, but doing so requires comparing not just the harms of marijuana use against the harms of prohibition, but the harms of marijuana use against the harms of whatever people do instead when it’s illegal—a calculation that quickly becomes prohibitively complex.

Finally, the seemingly neutral framing of the question—whether the societal benefits of legalizing recreational marijuana nationwide would outweigh the harms—fundamentally misallocates the burden of proof. In a free society, the default position is liberty, not prohibition. Those who wish to criminalize peaceful conduct must justify why the state’s police power should override individual autonomy, not the other way around. The question should not be whether marijuana users can prove their freedom is worth the costs, but whether prohibition advocates can meet the heavy burden of justifying why adults should face criminal penalties for private choices about their own bodies and consciousness.

This burden becomes even more demanding when the prohibited conduct is widespread, socially accepted, and demonstrably less harmful than many legal activities. When a clear majority of Americans support legalization, when the activity produces no direct harm to nonparticipants, and when decades of prohibition have failed to meaningfully reduce use while imposing enormous social costs, the presumption must run strongly in favor of individual choice. The state cannot simply point to speculative harms and declare victory—it must make a compelling case that criminal prohibition is both necessary and proportionate, and that no less restrictive alternative could address whatever legitimate concerns exist. That is a burden prohibition advocates have never met, and the utilitarian framing of this debate conveniently allows them to avoid even attempting to do so.

So, do the societal benefits of legalizing recreational marijuana outweigh the harms? I have no idea. My interlocutor doesn’t know either. And neither do you. But at least we can take comfort in the fact that the answer is irrelevant to any philosophically respectable policy discussion.