Free-market capitalism unleashes awesome forces. The quest for ever greater profits stimulates innovation in products and production processes, yielding a wider range of cheaper and more effective products in which consumers can indulge — and sometimes over-indulge. That is a blessing in the case of 99 percent of products, but not all of them. We do not allow corporations to sell human organs, sexual favors, or performance-enhancing steroids for non-medical use, and some harbor misgivings about for-profit prisons and universities.
With little consideration of alternatives, the country is deciding whether to create a for-profit industry for marijuana and other cannabis products. I think we should not.
The usual debate concerns whether to legalize marijuana. That framing distorts, because legalization is not a binary choice; there are many ways to legalize, with sharply varying pros and cons. In addition to legalizing use — which is what most polls ask about — there are many options on how to legalize supply. At the cautious end is just letting individuals grow a little for themselves at home, as in Vermont and Washington, D.C. At the far end is legalizing commercial production by for-profit industry. Colorado and Washington State pioneered that riskier approach, which Oregon and California have emulated.
There are many intermediate options, including restricting production to nonprofit organizations. Right now, the Trump administration could steer the country toward those better options at modest cost, but that window of opportunity will soon close forever. Once established, a for-profit industry would be next to impossible to unwind.
No modern nation has ever allowed large-scale commercial production for non-medical use — although Canada is expected to do so on July 1. The Netherlands allows only retail sale, not production or wholesale trade. Uruguay has taken a different but similarly cautious approach. Elsewhere, cannabis reform is restricted to granting rights to individuals, not corporations.
Legalizing a for-profit industry is an irreversible leap into the unknown. Prudence suggests first taking smaller steps. The horse is not yet out of the barn because the rapidly growing, state-licensed marijuana businesses are still 100 percent illegal under federal law. The federal government could shut them down in a heartbeat. The 1980s and 1990s showed law enforcement’s limitations in the face of illegal drug markets, but that does not extend to state-licensed businesses. It is easy to enforce laws against licensed companies that operate out of brick-and-mortar buildings at fixed addresses that they report to regulators. The Justice Department can simply send a letter warning that if the company does not shut down within a reasonable time, its owners will be arrested and its assets seized. There is no reason to actually put anyone in prison — except those who don’t take the hint and persist beyond the specified grace period. This would not require an act of Congress. The Obama administration opened the era of commercial cannabis production when its Justice Department issued a memo (the “Cole memo”) promising non-interference with businesses that are legal at the state level and meet certain minimal requirements. The Trump administration could phase out those for-profit businesses in favor of nonprofit organizations by issuing a new memo. Solidifying the policy with changes in federal law would be good but can come later. Shutting down state-licensed businesses could feed the black market, but not if nonprofit organizations are allowed to take their place.
Below, I describe two versions of this idea. The first would preserve existing state rules and regulations and would add one: All licensed cannabis enterprises must be nonprofit organizations with boards constituted to protect public health and with charters that define their mission as meeting existing demand (to undercut the black market) but not promoting greater consumption. The second would restrict production to co-ops that supply only their own members.
Before getting to the details, it is worth explaining why cannabis merits this cautious approach, as opposed to being treated as a regular article of commerce, albeit one subject to special taxes and regulations, the way we treat alcohol, tobacco, gasoline, tanning beds, and rental cars.
Cannabis is a dependence-inducing intoxicant, but a relatively safe one. Overdoses — particularly from edibles — prompt many thousands of people to seek care in emergency rooms every year, but overdose deaths are all but impossible. Even long-term use doesn’t cause much organ damage. Yes, cannabis smoke contains carcinogens, but not enough to make excess cancers visible in epidemiological studies. Cannabis intoxication impairs reaction time, memory, and one’s ability to perform tasks that require attention, but it does not produce reckless or aggressive behavior the way alcohol does.
Adults’ using a few times a week when not at work, school, or minding children is pretty harmless, and that describes almost half of cannabis users. Alas, it describes only a tiny share of cannabis use, because the smaller number of daily or dependent users consume so much more per person.
We may ask: “What proportion of cannabis consumption is accounted for by adults with no history of substance-use disorder and who use fewer than ten times a month?” (The last clause proxies “use only on weekends”; the data do not record day of week.) Anyone can work out the answer with public data from the National Survey on Drug Use and Health, plus the rule of thumb that daily users consume about three times as many grams per day of use as do occasional users.
Such moderate, adult use is engaged in by about one in three cannabis users, but accounts for only 2 percent of consumption and so a trifling share of sales and profits. Likewise, many kids use, but most do not use daily, and there are some adults who use ten to 20 times per month.
Yet it is daily and near-daily users who account for about 80 percent of consumption. As policy liberalized, cannabis transformed from a weekend party drug to a daily habit, becoming more like tobacco smoking and less like drinking. The number of Americans who self-report using cannabis daily or near-daily grew from 0.9 million in 1992 to 7.9 million in 2016.
Here are a few other pertinent statistics. Just under half of consumption is by people who report either having been in alcohol or drug treatment or suffering enough current problems to meet medical criteria for substance-use disorder. (Since denial is a hallmark of addiction, this proportion is likely conservative.) About 60 percent of consumption is by people with a high-school education or less, a group with lower disposable income and greater sensitivity to falling prices.
Prices have indeed fallen, and sharply. From August 2014 to August 2017, prices (with taxes) in Washington’s state-legal market fell by 70 percent — from $23.50 to $7.25 per gram. The THC potency — the key measure of how strong the marijuana is — now averages 22 percent, so the cost per hour of intoxication has fallen below $1, cheaper than beer or going to the movies.
Product variety has exploded, including THC-infused candies and edibles, oils that can be vaped (akin to e-cigarettes), and chunks of 70-plus percent THC that are suitable for flash-vaporization (“dabbing”). The increase in average daily dose has been startling. Until 2000, the average potency of seized cannabis never exceeded 5 percent, and 4 percent was typical. Someone consuming one 0.4-gram joint each weekend night was consuming 0.032 grams of THC per week, or 4.6 milligrams per day. Daily users now average about 1.3 grams per day. At 20 percent potency, that is 260 milligrams per day — nearly 60 times as much.
People develop tolerance to cannabinoids, so perhaps higher doses might have few behavioral consequences. On the other hand, the dose might make the poison. Consider two sharply different patterns of cocaine consumption. A member of the indigenous populations in the Andes who drinks a cup of coca tea per day (a habit analogous to coffee drinking) gets about 4 or 5 milligrams of cocaine, while the average “chronic” cocaine user (before the U.S. cocaine market went into decline) consumed about 260 milligrams of cocaine per day.
To put it in more familiar terms: My daily 20-ounce Diet Coke contains 76 milligrams of caffeine. Multiplying that dose by 60 would require drinking 30 Starbucks 16-ounce grande cappuccinos. Although humans also develop tolerance to caffeine, escalating from one Diet Coke to 30 grande cappuccinos would affect my health and behavior. So the consequences of the dominant mode of cannabis consumption today might differ from what Baby Boomers experienced on college campuses in days of yore. Unfortunately, most laboratory studies investigate doses of 20 to 50 milligrams of THC, not 200 or 300 milligrams.
I am not saying that consuming 260 milligrams a day of THC is unhealthy. I am saying there is little historical experience with those doses, and even less laboratory evidence. That makes me leery of creating a for-profit industry whose raison d’être is getting people to buy more and more of its products, which effectively means getting more and more people to use heavily.
I suggest that we pause for a decade and restrict legal supply to nonprofit organizations. One option would require organizations applying for a state license to be nonprofit groups whose governance structures focus them on serving the public interest. I suggest two conditions. First, the majority of governing-board members must come from the child-welfare and treatment communities. Second, the organization’s charter must define its mission as meeting existing demand, in order to undercut the black market, but not promoting greater consumption.
Two arguments for legalization are the need to eliminate the black market and the desire to reduce criminal-justice costs. The approach I have described would achieve those goals. Also, although zero profits implies zero corporate income taxes, nonprofits pay excise taxes, and their employees pay individual income taxes, leaving most of the potential tax revenue intact.
The primary argument against legalization is concern about greater use and abuse. Evidence to date suggests that decriminalizing or legalizing use has modest effects, but commercialization does increase use. Use in the Netherlands rose not with the creation of its policy, but later, when the number of coffee shops rose. Likewise, use in Colorado rose not when its medical-marijuana law passed in 2000, but when dispensaries proliferated in 2009.
Sharp increases in use are less likely if the only legal suppliers are low-key, eschewing advertising, promotions, and lobbying, than if a for-profit industry is actively promoting greater use. Still, the nonprofit option makes problematic promotion only less likely, not unimaginable. Some nonprofit hospitals compete as aggressively as their for-profit counterparts. Nonprofit universities do, too. Even government monopolies can promote vices, as with state lotteries. Large organizations with ambitious, paid, professional leaders can manifest a drive to grow even when that growth is not profit-driven.
An alternative approach would be more conservative, restricting legal supply to small co-ops that produce for members but not outsiders. These “cannabis clubs” are common in Spain and some other countries, including Belgium and Uruguay.
Co-ops essentially pool the own-growing privileges of their members. A club with 30 members, each permitted one plant, could have up to 30 plants, and all 30 could be at the residence of whichever member had a green thumb. This would preserve artisanal production and avoid economies of scale that radically drive down prices. Spanish clubs allow sale to members at cost, to reimburse gardeners for materials and time. A more cautious version would permit only gifting.
Admittedly, the co-op model would not displace the entire black market. Most high-frequency users would probably join and so take most of the market away from criminals, but youth, low-frequency users, and some frequent users might still purchase in the black market. That is not a fatal flaw: After all, in the lauded “Dutch model” of legalizing only retail sale, 100 percent of production and wholesale distribution is black-market activity, so a plan that protects against greater use and eliminates 60 percent of the black market should not be dismissed.
Nor are the co-op and nonprofit models mutually exclusive. A state could permit co-ops to grow for members tax-free while also allowing other nonprofits to supply taxed cannabis to those who do not wish to join a co-op. That combination might all but eliminate the black market while tempering somewhat the size and power of the nonprofit organizations.
The pro-marijuana movement celebrates legalization as a triumph following in the footsteps of civil rights, women’s rights, and homosexual rights. That is juvenile. Although most marijuana users have no problems with the drug, most sales and profits flow from people who consume so much that it interferes with their lives. The typical session of marijuana use is part of a bad habit, if not a diagnosable substance-use disorder.
Mark Kleiman, my colleague at New York University, has suggested meeting legalization with “grudging toleration,” not hosannas. Grudging toleration can be practiced in various ways, including high taxes and restricting supply to a government monopoly managed by a public-health bureaucracy. A simpler option that the Justice Department could enforce tomorrow is restricting legal supply to nonprofit organizations.
Maybe ten years’ experience will show that legalization does not increase problem use by enough to fuss over. If so, a for-profit industry could enter the market then. But if we create an industry and discover that it promotes use by people with substance-use disorders, then no matter how much we might regret that cannabis equivalent of the tobacco industry, there would be no turning back.