Many people enjoy the psychological effects of various chemicals. Any chemical can have unwanted side effects, especially when used often, in high doses, or in combination. There is always the risk that a user will lose control over his or her consumption, using too much or too often.

The likelihood of developing what is now called “substance use disorder” varies by person and by drug; except in the case of nicotine, the victims of this disorder are generally a small minority among users. Most people unfortunate enough to develop a drug problem recover without formal intervention, although recovery typically comes after some struggle and several failed attempts.

But an even smaller minority faces graver problems. Their attempts to cut back fail because of withdrawal symptoms or persistent cravings; they have become addicted. Addicts, although relatively few in number, account for most of the damage done by drugs.

Some potentially habit-forming chemicals—including the two biggest killers, alcohol and tobacco—are legal to use and sell. Others are illegal or restricted to medical use by prescription. This tends to reduce the number of people who develop drug problems, but it also worsens the problems of those who do develop them. Making a drug illegal creates illicit markets and the need for enforcement, and can lead to violence.

The United States has a variety of legal and illegal drug markets, and more than its share of the evils of addiction, illicit trafficking, and drug-related incarceration. Two of those markets—those for cannabis and opioids—will force themselves on the attention of the new administration of U.S. President Donald Trump, although for very different reasons.

Cannabis will be on the agenda because of the conflict between state policies and increasingly unpopular federal law. Last October, a Gallup poll found that public support for legalization had reached 60 percent, the highest level since Gallup began asking the question in 1969. In November, four states, including California, voted to allow cannabis sales without a medical recommendation. More than a fifth of all Americans now live in the eight states that issue permits to grow and sell cannabis—actions that federal law still defines as felonies.

This situation leads to absurd consequences. Some state-licensed cannabis businesses pay their state taxes with sacks of cash because money-laundering laws discourage banks from letting them have checking accounts. Respectable law firms file state regulatory applications to enable their clients to commit federal felonies. Somehow, federal law needs to adapt to the new realities.

Opioids—including both illicitly manufactured heroin and fentanyl compounds and prescription drugs such as oxycodone—are on the agenda for a much grimmer reason: the United States is facing a massive epidemic, with the rapidly rising death toll now great enough to contribute to falling overall life expectancies.

Current policies toward cannabis and opioids are equally unsustainable; the opioid problem is both more serious and harder to fix. Better cannabis policies would accommodate the movement toward cannabis legalization without going all the way to alcohol-style commercial availability; the goal would be to shrink the illicit market while damping the growth of cannabis use disorder and avoiding an upsurge in teenage use. Better opioid policies would curb the over-aggressive marketing and prescribing of opioids that helped create the current problem without going back to the days when patients suffered needlessly from untreated or undertreated pain; they would also improve addiction treatment and make it more widely available, and offer better therapy to those who suffer from chronic pain.

The new administration has great political flexibility; Trump has not committed to any specific cannabis or opioid policies. On the campaign trail, he promised to solve the opioid problem by stopping the flow of smuggled drugs and expanding treatment for opioid addicts. But the new administration will struggle to reconcile the latter with its commitment to repeal Obamacare, which greatly increased funding for drug treatment.


In 1992, illegal cannabis sales in the United States totaled about $10 billion; in recent years, that figure has topped $40 billion, making the market for cannabis by far the largest illicit drug market. In 1992, when polled, of those who said that they had used marijuana in the past month, only about nine percent reported daily or near-daily use. Today, that figure is 40 percent, or about eight million people; about half of them report the symptoms of substance use disorder, including failed attempts to cut back or quit.

Despite steady growth in public support for legalization, federal cannabis law has not changed in decades. But there have been dramatic developments at the state level. In addition to the eight states that now permit commercial sales, another 35 allow the sale or use of cannabis on medical recommendation, which also remains illegal under federal law.

Cannabis grower Steve Dillon tends to his plants on his farm in Humboldt County, California, U.S. August 2016.
A cannabis grower tends to his plants on a farm in Humboldt County, California, August 2016.
Rory Carroll / Reuters

The changes in state law have put the federal government in a bind. The states can’t repeal federal laws, but the federal government can’t enforce those laws without help from the states: 4,000 federal Drug Enforcement Administration agents cannot replace 500,000 state and local police. The Justice Department could shut down state-licensed businesses by obtaining federal injunctions. But unless the states were willing to arrest growers and retailers, the federal government would simply be replacing taxed and regulated sales with untaxed and unregulated sales.

Even with the full cooperation of the states, mounting the enforcement effort required to suppress a $40 billion illicit market is hard to imagine, given the overstrained criminal justice system and concerns about excessive incarceration. Even the current level of half a million arrests for cannabis possession every year strains the relationships between the police and the communities they serve, especially in high-crime minority neighborhoods. But that level is too low to seriously deter people from consuming cannabis: the risk of arrest per day of use is below one in 5,000.

Under President Barack Obama, federal agencies reluctantly acquiesced to the state-level cannabis legalization, except when state-licensed activity involved interstate sales, sales to minors, the use of weapons, or links to organized crime or terrorism. Senator Jeff Sessions of Alabama, Trump’s nominee for attorney general, criticized the Obama administration for not enforcing the law; he also asserted that “good people don’t smoke marijuana.” But as attorney general, Sessions will face the same arithmetic that confronted Eric Holder, Obama’s attorney general: his department doesn’t have the manpower to enforce federal laws without help from the states.

Cannabis prohibition has broken down, probably beyond repair.

Cannabis prohibition has broken down, probably beyond repair. But what has replaced it in the legalizing states is far from ideal. The slogan behind the new system—“Regulate marijuana like alcohol”—sounds sensible only to those who ignore how bad U.S. alcohol policy is. Thanks in part to low taxes and aggressive marketing, 16 million Americans suffer from alcohol use disorders, and about 90,000 people die from alcohol-related causes every year. The alcohol industry depends for most of its revenue on the minority of people who drink too much, and the industry’s political clout ensures that public policy doesn’t interfere much with the business of promoting and profiting from alcohol abuse.

Under the current version of legalization, the marijuana industry is likely to follow the same playbook: for-profit businesses will strive to create more and more of the heavy daily cannabis use that accounts for 80 percent or more of cannabis sales.

The right set of policies for marijuana would look less like the current policies on alcohol and more like those on tobacco, where taxes and regulations are designed to decrease smoking. High taxes, restrictions on marketing, and relentless antismoking messages have driven tobacco use down sharply—especially among minors—and it will continue to fall. But current state-level cannabis legalization features relatively low taxes, loose regulations, and minimal restrictions on marketing (except to minors). As legal marijuana production replaces illegal growing, cannabis prices will continue their rapid decline: adjusting for inflation and potency, today’s cannabis produces about four times as much intoxication per dollar as it did a quarter century ago, and legal competition will drive prices lower still. Lower prices make it easier for casual users to slip into heavy use: good for the vendors, bad for the users.

A good alternative to full national legalization would be to change federal law to accommodate state-licensed cannabis sales, but only if the taxes and regulations that replaced state prohibitions were strict enough to prevent an acceleration in the rate of heavy use. The federal government could do this by using “policy waivers,” like those it now uses to allow state-level experiments with other policies. But legalizing cannabis without prompting a large increase in heavy use would require very different polices from those adopted so far in the legalizing states. At a minimum, it would require replacing taxation based on price—which means that taxes fall with market prices—with taxation based on potency. More radically, it might entail replacing a for-profit industry with co-ops, nonprofits, or state-operated retail stores.

For now, the current debate on legalization remains at the level of yes or no, with no intermediate options on the table. Proponents of legalization see no reason to compromise, while the remaining supporters of prohibition are holding out to the bitter end, hoping that the steady growth in support for legalization will somehow miraculously reverse. It’s not that voters or officials have rejected the ideas about temperate cannabis policy developed by the tiny group of academic drug policy analysts; rather, those ideas have never been up for discussion.

If the federal government is ever going to move toward policies that support moderation, the time is now. Once California and the other states where marijuana was recently legalized have created multibillion-dollar commercial markets, potent political forces will resist any radical change.


The costs of inaction on opioid policy would be much higher. An estimated two million Americans suffer from opioid abuse disorders, and in 2015, 32,000 died of opioid overdoses—nearly as many as died in car crashes and more than twice the number killed in homicides.

The abuse of prescription opioids, including hydrocodone (sold as Vicodin or Lortab) and oxycodone (or Percodan, Percocet, and Oxycontin), began to grow rapidly in the early 1990s; the annual count of people reporting first-time nonmedical use of opioids rose from around 200,000 in 1992 to more than 2.4 million a decade later, exceeding the comparable figure for cannabis.

An estimated two million Americans suffer from opioid abuse disorders, and in 2015, 32,000 died of opioid overdoses.

For the most part, those drugs were not smuggled into the country; they were prescribed by physicians and purchased legally from pharmacies. Encouraged by pharmaceutical manufacturers, physicians began to consider pain “the fifth vital sign” that they should monitor routinely, along with body temperature, blood pressure, pulse, and respiration rate, and to overrule concerns that the medical use of opioids would lead to dependency.

Rising supplies of prescribed opioids helped create a black market. Patients exchanged and sold unused pills; burglars stole them. Drug dealers began to recruit people to pose as patients and secure high-dosage prescriptions from as many physicians as possible. Drug-seeking patients learned that they could usually get a prescription just by rating their pain at seven or above on an arbitrary ten-point scale.

Prescription opioids penetrated populations left largely untouched by heroin. Finding heroin required finding a dealer, and dealers clustered in places where heroin was already common; the prescription drugs were available wherever there were physicians and drugstores. In some states, such as Florida, lax laws encouraged so-called pill mills, where doctors prescribed—and sometimes also dispensed—opioids to anyone willing to pay. The pills were less frightening than heroin and therefore more appealing. They came in measured doses in pill bottles, not as white powders of unknown composition in glassine bags. They were typically swallowed like normal medicines, rather than snorted or injected. And they were available at a drugstore, or from an acquaintance who had a prescription, instead of from a dealer in a back alley.

A community health worker visits a former addict who was treated at a suboxone clinic and now lives in supportive housing, Boston, Massachusetts, January 2013.
A community health worker visits a former addict who was treated at a suboxone clinic in Boston, Massachusetts, January 2013.
Brian Snyder / Reuters

But the two markets did not remain separate for long. A person addicted to prescription opioids whose need for the drug outstrips his or her budget may trade down to heroin—which costs about a quarter the dose-equivalent price of prescription opioids on the black market—or to the even cheaper, more potent, and more dangerous synthetics of the fentanyl class. Law enforcement efforts can have the unwanted side effect of accelerating the transition: when the police shut down a local pill mill, they rarely identify the users and help them get treatment, and heroin and fentanyl dealers are quick to move in to exploit the new business opportunity. On the other hand, if the police don’t shut down pill mills, they risk swelling the number of prescription-opioid users who may later graduate to heroin or fentanyl.


Policymakers and health-care providers have several options to tackle the opioid crisis. None offers a miracle cure, and each involves either spending money or imposing and enforcing regulations.

The quickest way to save lives is probably to expand access to “antagonist” drugs, which can bring overdose victims back from the brink of death. These drugs, such as naloxone (sold as Narcan), save thousands of lives every year. Naloxone is now available as a nasal spray, and it requires no medical training on the part of the person administering it. Changes in policy have made antagonists easier to obtain legally and have put them in the hands of police and emergency medical technicians, and aggressive public information campaigns have spread the word that an overdose is reversible if first responders (or the opioid user him- or herself, a friend, or a passerby) can administer an antagonist quickly.

But reversing an overdose is only a start; many users overdose more than once. Last April, for instance, naloxone was used to revive the music icon Prince; one week later, he overdosed again, with no one around this time to administer the antidote.

Getting opioid users into treatment and keeping them there requires hard work. Substitute drugs, such as methadone and buprenorphine, can relieve withdrawal symptoms and prevent overdoses, but regulatory barriers and a lack of trained clinicians have made them hard to obtain. Methadone clinics, for example, are mostly located in big cities, where they sprang up in response to the last heroin epidemic; today, however, most users live in the suburbs, exurbs, small towns, or rural areas, far from the nearest clinic. Too much of the criminal justice system still insists on strict abstinence and rejects substitution therapy, despite overwhelming scientific evidence that it works. Many drug courts and probation and parole agencies, and most prisons and jails, refuse to let their clients and inmates use substitute drugs. And the substitutes alone aren’t nearly as effective as substitution accompanied by high-quality psychosocial treatment, which not every prescriber of the substitutes is able or willing to provide.

Prescription opioids have penetrated populations left largely untouched by heroin.

Recent advances in substitution therapy, such as implants that avoid the need for daily dosing, are promising but expensive, and expanded access to treatment would have to be paid for. The same antagonist drugs that reverse overdoses can also be administered in long-acting formulations; a monthly injection can prevent a user from getting high even if he or she relapses, greatly reducing the risk of relapse. But these drugs, like the long-acting substitutes, cost more than $1,000 per month.

Under the Affordable Care Act, drug treatment is one of the “essential health benefits” that public and private insurers are required to cover. Subsidies for private insurance through the ACA exchanges and the expansion of Medicaid have provided health coverage, including drug treatment, to about 20 million people who had previously been without it. Keith Humphreys, a professor of psychiatry at Stanford University, has called the ACA “the largest expansion of drug treatment in U.S. history,” and the official estimate is that it has improved access for 60 million people.

Trump and congressional Republicans have pledged to “repeal and replace” Obamacare. Last year, Representative Tom Price of Georgia, Trump’s nominee for secretary of health and human services, put forward an alternative that removed the requirement for insurers to cover a specific set of benefits. Since people with drug problems are expensive to insure, under such a plan, insurers would presumably revert to their previous practice of driving them away by offering no coverage, or inadequate coverage, for drug treatment. In addition, Price proposed cutting federal funding to subsidize private insurance and reversing the Medicaid expansion. That approach would make it hard to expand access to high-quality opioid treatment.

While objections to public spending are one barrier to expanding treatment, objections to government regulation—embodied in the Trump campaign’s promise to repeal two old regulations for every new one adopted—are a barrier to reducing the supply of diverted prescription pills. The current crisis is partly the result of inadequate regulation.

Much of the necessary power lies at the state, rather than the federal, level. State medical boards should be more aggressive in revoking the licenses of pill-peddling practitioners, instead of leaving the problem for the police to handle. Databases of opioid prescriptions (called Prescription Drug Monitoring Programs, or PDMPs), which states are increasingly using, can help physicians and pharmacists spot pill-seeking patients, shrinking the supply of pills on the illicit market. But those databases are full of personal information that needs protecting; designing databases that are both secure and easy to use is difficult and expensive. Consulting a state’s PDMP also takes up clinicians’ scarce time, and without regulations or incentives to encourage their use, PDMPs won’t work.

The current crisis is partly the result of inadequate regulation.

None of these moves would address the availability of heroin and fentanyl. Indeed, if physicians deny users opioids, or if the price of illicit prescription opioids begins to rise as the supply falls, demand for heroin and fentanyl will rise, possibly raising death rates, at least in the short run. In 2014, deaths from overdosing on prescription opioids fell, but deaths from fentanyl overdoses almost doubled.

As long as there is demand, preventing those cheaper drugs from entering the country will be almost impossible. More than a million cargo containers cross the United States’ borders every month; any one of them could hold enough heroin to supply the country for that month or enough fentanyl to supply it for a year. Cracking down on the retail supply has become much harder since drug dealers started connecting with customers by cell phone rather than by loitering on street corners. Policing is expensive: annual police budgets nationwide total more than $100 billion. Ramping up operations against opioids would require either spending more money or doing less of something else: enforcing other drug laws or suppressing predatory crime, for example. Imprisoning more dealers would require letting other offenders out or reversing the widely desired decrease in the U.S. prison population, which now stands at five times its historical level and seven times the average rate of other rich democracies.

Cracking down on opioid prescribing could also make it much harder for people in genuine pain to receive relief. Opioids are often not the best way to manage pain, especially chronic, nonterminal pain: patients often need help changing patterns of work, stress, exercise, and diet. But too few health-care providers understand these approaches, and many insurers will not pay for them. Prescribing some pills is much cheaper than providing physical therapy.

A long-term solution would require better clinical practice and new drugs on the market both for pain relief and for opioid-dependency treatment. Buprenorphine, for example, a fairly cheap generic drug used in substitution therapy, can also relieve pain, and it carries a very low risk of overdose. But it is currently packaged and marketed primarily for treating opioid addiction and severe chronic pain; internists are more likely to prescribe the more dangerous hydrocodone or oxycodone. A drug company that wanted to make buprenorphine a routine pain drug would have to put a new formulation through a long, expensive regulatory process at the Food and Drug Administration, with no guarantee of regulatory success or sufficient clinical acceptance to recoup its investment.

The same is true of several promising drugs and formulations for drug treatment: someone has to pay to develop them, and right now there isn’t enough financial reward to justify the gamble. The federal government could fill that gap, funding not only basic research (as it currently does) but also the clinical-trial process for drugs with high social value but limited profit potential.

Ultimately, the opioid epidemic, like all epidemics, will burn itself out: as the grim joke shared among medical residents goes, “All bleeding stops, eventually.” But how many lives the epidemic takes, and how many it ruins, will depend on choices made today and tomorrow. The worst of the problem is almost certainly still to come.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • MARK A. R. KLEIMAN is Professor of Public Policy and Director of the Crime and Justice Program at New York University’s Marron Institute of Urban Management. He is the author (with Jonathan Caulkins and Beau Kilmer) of Marijuana Legalization: What Everyone Needs to Know. He is also Chair of BOTEC Analysis, a consultancy with clients that include the Washington State Liquor and Cannabis Board. Follow him on Twitter @MarkARKleiman
  • More By Mark A. R. Kleiman