End the War on Drugs and Tax Them Like Alcohol

From the article:

The United States spends an estimated $47 billion a year enforcing drug prohibition. Today, roughly 200,000 people are in state or federal prison for drug offenses, with additional tens of thousands cycling through local jails. The overdose crisis has now killed more than a million Americans since 1999, and the illegal drug market continues to thrive under the management of cartels and street dealers who have never once been asked to submit to a safety inspection. If this were a clinical trial, the institutional review board would have pulled the plug decades ago.

Becker, Murphy, and Grossman formalized the economic theory in a 2006 paper in the Journal of Political Economy. Their key finding: when demand for a good is fairly inelastic (as it often is for addictive substances), a monetary tax on a legal good can achieve a greater reduction in consumption than optimal enforcement against the same good if it were illegal. In plain English: legalize it and tax it heavily, and you may end up with less use than by trying to police it out of existence. The reason is that prohibition is fantastically expensive enforcement with large deadweight losses, while taxation is cheap and generates revenue you can spend on treatment.

The right lesson from Oregon is not "bring back arrests." It's "if you're going to decriminalize, the treatment and harm-reduction infrastructure has to be in place first, or at least built out simultaneously. And your governance has to be competent." Oregon ran the weakest possible version of reform in the hardest possible conditions and then declared that reform itself had failed. That's like prescribing antibiotics, not making sure the pharmacy delivers them, and then concluding that antibiotics don't work.

"Legalization will increase use." Probably (depending on the tax regime), at least modestly. The cannabis evidence suggests legalization does increase use somewhat, especially among young adults. This is a cost. But it has to be weighed against the costs of prohibition: incarceration, criminal records, the violence of illegal markets, contaminated supply, and the absence of quality control. The question is not "will more people use drugs?" It's "does the harm from increased use outweigh the harm from prohibition?" For cannabis, the answer seems clearly yes: the harm reductions from ending arrests and creating quality control are large, while the increases in use are modest. For harder drugs, the calculation is less certain. But even there, the deaths from fentanyl contamination in the illegal supply are so catastrophic that even a significant increase in use under a clean legal supply could still result in far fewer total deaths.

Stop arresting people for drug possession. It doesn't reduce use, it ruins lives, it costs a fortune, and it falls hardest on people who are already disadvantaged. Invest massively in treatment and harm reduction. Build the system before you change the law, or simultaneously, not 18 months later. Legalize and regulate cannabis through adult retail markets with real restrictions on marketing, potency escalation, and youth access. For higher-risk drugs, create supervised and medically integrated access pathways designed to pull users out of the contaminated street supply. Fund the whole thing partly through drug taxes and partly through redirecting a substantial share of what we now spend on enforcement and incarceration. And expect that some use will increase, accept that as a cost, build the health system to manage it, and remind yourself that the status quo is killing 80,000 Americans a year.

The war on drugs has been running for over fifty years. It has never won, and it cannot win, because it was built on the idea that criminal punishment could scare or cage a society out of wanting to alter its consciousness, and that idea is wrong. The only real question is whether we replace it with a system that’s thoughtful, evidence-based, and designed to reduce harm, or whether we keep pretending that one more decade of the same policy will finally produce different results.

I know which one I’d pick.