We’re Legalizing Weed Wrong

The end of marijuana prohibition is necessary and inevitable. But the referenda facing voters will create new problems, even as they solve old ones.

A customer looks at marijuana for sale at a dispensary in Eugene, Oregon on March 22, 2016.
A customer looks at marijuana for sale at a dispensary in Eugene, Oregon, on March 22.

Andrew Caballero-Reynolds/Getty Images

Voters in five states will decide on Tuesday whether to allow the sale of cannabis to all adults. Another four will vote on legalizing it for medical use. If all of these measures pass, nine states (plus the District of Columbia) will have legalized marijuana for recreational use, and 29 states will have some form of cannabis legalization on the books.

So what do voters need to know in order to make an informed choice? Reasonable people, starting with different values, can come to different conclusions about the measures on the ballots, which also vary somewhat in their particulars. But we should all be reasoning from the same set of facts.

In my view, based on what I know from 30 years of working on cannabis policy in both government and academia—including advising the Washington State Liquor and Cannabis Board on implementing that state’s legalization measure—legalization is not only necessary but virtually inevitable, even at the federal level. But I also believe that, by and large, our approach to legalization contains some consequential flaws and that the necessity of such laws doesn’t mean we should rush out and grab the first legalization proposal we’re offered. Which is why reasonable debate over how to do this ought to start here:

1. Cannabis Prohibition Is Broken, and No One Has a Plan to Fix It

The numbers tell the story: 40 million past-year users, half a million arrests per year, $40 billion in illicit revenues. The arrest rate is high enough in absolute terms to create huge social costs, but not high enough compared with the size of the market to effectively deter use or sales. Cannabis is now by far the single largest illicit market, dwarfing those for cocaine, opiates, or illegal sex work. In the context of concerns about mass incarceration and overly intrusive policing in poor minority neighborhoods, the sort of massive law-enforcement drive that would be needed to force this genie back into its bottle simply isn’t going to happen.

If California voters approve Proposition 64, as seems overwhelmingly likely, more than 15 percent of the U.S. population will live in states that license actions—growing and selling cannabis—that remain felonies under federal law. Even more people live in states that allow medical marijuana, which is just as illegal under federal law as any other kind. This situation is covered by Herb Stein’s postulate: An unsustainable situation won’t go on forever. The practical question now is how to legalize cannabis rather than whether to legalize it.

2. Not All Legalization Proposals Are Created Equal

Acknowledging that cannabis isn’t the demon weed doesn’t mean accepting it as harmless; it isn’t. Like alcohol or tobacco, cannabis can damage its consumers in various ways. A smart legalization scheme would try to minimize those harms. But that’s not what we’re getting from most legalization advocates.

“Regulate cannabis like alcohol” is an ugly policy. Our current alcohol laws fail spectacularly to control the harm alcohol does to drinkers and the harm drinkers do to others; an estimated 90,000 Americans die each year of their own drinking or someone else’s. Why repeat that mistake when we legalize another potentially habit-forming intoxicant? What we want is the sort of “grudging toleration” the law now extends to tobacco; we should be looking for means short of prohibition to limit the number of people whose lives are made worse by cannabis.

If there is a for-profit industry, the kind many legalization schemes create, its interests will be at odds with public health. There’s no evidence that casual cannabis use is harmful, and lots of evidence that chronic heavy daily smoking isn’t good for you. But if you’re in the cannabis business, casual users aren’t much use to you while heavy users are your best customers, accounting for the bulk of your sales. (The same is true of brewers and heavy drinkers.) So while the public interest is served by making cannabis available to those who want to use it responsibly while minimizing problem use, the commercial interest demands maximizing problem use.

Alternatives to commercialization include the “grow-and-give” model currently used in Washington, D.C., or the restriction of sales to consumer-owned co-ops or not-for-profits. (The best solution of all, from a public-health perspective, might be a state monopoly on retailing, but that’s not feasible while cannabis remains illegal federally, since state officials can’t be ordered to break federal laws.)

One reason to take the profit out of the pot industry is to prevent aggressive marketing; another is that a not-for-profit cannabis sector would be unlikely to match the lobbying muscle that the commercial pot industry is already accumulating and that is likely to expand along with the market. You don’t want the people who sell cannabis for a living writing the cannabis laws any more than you want Exxon Mobil guiding fossil-fuel policy.

3. Cannabis Use Disorder Is a Real and Growing Problem

In 1992, only 9 percent of current (past-month) cannabis users reported being heavy users (25 or more days per month). In 2014, that figure had risen to 40 percent. Those “daily/near-daily,” or DND, users consume about three times as much per day of use as less-frequent smokers. Right now, there are about 8 million DND users (up about sevenfold over the past two decades). Collectively, they account for more than 80 percent of cannabis sales, which explains why cannabis stores feature the sort of superstrong pot that’s way too intense for many casual users but that heavy users (who have become tolerant to the effects of THC) need to get stoned.

About half of those DND users—4 million people at any one time—self-report the symptoms of cannabis use disorder (the new diagnostic label for what used to be called “abuse” or “dependency”). That’s some combination of: (a) using more, and more often, than they want or intend to; (b) failing in attempts to cut back; (c) spending so much time stoned that it interferes with their other plans and responsibilities; and (d) coming into conflict with people they care about due to their cannabis use.

A key question to ask about any proposal for legalization is “What’s the plan for stemming the growth of cannabis use disorder?” The current answer, from legalizers and prohibitionists alike, is largely the sound of crickets: The legalizers want to downplay the problem while the warriors don’t want to admit the possibility of moderate and therefore harmless use.

4. Stoned Driving Is a Minor Problem, but the Effects on Juveniles Is a Serious One

Driving stoned is more dangerous than driving sober, but the difference is more like the additional risk of driving while sleepy or angry than it is like the additional risk of driving drunk. It’s nowhere near as dangerous as driving while using a cellphone, even hands-free. Stoned driving should be a traffic offense, not a crime like drunk driving. Traffic risks aren’t a substantial objection to legalization, though of course smart policy would discourage driving stoned, and especially driving with both cannabis and alcohol on board.

On the other hand, some of the data about heavy use by juveniles, and especially about effects on educational performance, are pretty scary. The recent upsurge in heavy cannabis use has shown up far more in adults than in minors, and we should endeavor to keep it from moving down the age distribution. Putting money into “prevention” programs is not a solution. Keeping cannabis expensive and restricting its marketing (not just its marketing to juveniles) would help.

5. Cannabis Prices Have Been Falling—and That’s Not a Good Thing

The price per ounce or per gram is a poor guide to understanding the cannabis market. After all, consumers aren’t buying plant material or concentrate; they’re buying hours of intoxication. Over the past 20 years, the average price of a gram of pot has been relatively stable at about $10 (which, of course, means that it’s been eaten away by inflation to some extent). But the THC concentration in that gram has increased dramatically, from low-to-mid-single-digit percentages to the mid-to-high teens. (Some of that change reflects agronomy; the rest is the shift from selling mostly leaves to selling almost exclusively flowers.)

The effective price of 20 milligrams of THC—roughly speaking, enough to get a user who hasn’t built up a drug tolerance high for about three hours —has fallen from about $5 to about $1. (Not all of that THC gets into the bloodstream; some is lost in the smoking process.) At less than 50 cents per intoxicated hour, cannabis is already a far more cost-effective recreational drug than beer. That doesn’t matter much to someone using a fifth of a gram once a week, but it matters a lot to someone using three grams every day. And it might matter to cash-constrained teenagers: High prices could help prevent them from using more than occasionally.

6. Prices Will Fall Faster Under Legalization

The Netherlands sorta, kinda legalized retail cannabis sales long ago, with only minor impacts on problem use. But while the Dutch system permits the sale of cannabis in “coffee shops,” it continues to forbid and punish growing cannabis. As the Dutch say, the front door of the coffee shop is (almost) legal, but the back door is strictly illegal. That somewhat incoherent but workable system has prevented any significant decrease in price. (Also, Dutch cannabis businesses aren’t allowed to advertise.) Legalization as enacted or proposed in various U.S. states looks very different.

In Washington state and Colorado, legalization didn’t have much immediate impact on prices, because it took time for firms in the newly legal industry to get their permits and ramp up production, and the limited number of retail licenses allowed retail cannabis stores to collect big markups. But in the long run, competitive pressure makes it impossible to charge illegal-market prices for a legal product that isn’t very hard to grow.

The move from energy-intensive indoor growing to outdoor growing creates large potential cost savings, and the outdoor growers have learned light-deprivation techniques that allow them to produce solar-powered pot nearly as potent as the most potent indoor strains. One big retailer in Seattle now offers “budget bud”: sun-grown, coarse-trimmed cannabis flowers, tested at 18 percent THC by weight, for $95 per ounce, which works out to about $3.50 per gram. That gram contains about nine 20-milligram doses, so a dose costs the user something less than 40 cents, or about 15 cents per stoned hour. And there’s no reason to think that $95/ounce is anything like the bottom; under full national legalization, industrial farming could bring the price of getting stoned to less than a nickel. Again, that doesn’t matter much to a casual user, but it’s likely to drive further increases in heavy, daily use.

A marijuana user named Bob smokes marijuana during a 420 Day celebration on "Hippie Hill" in Golden Gate Park April 20, 2010 in San Francisco, California.
A man smokes marijuana during a 420 Day celebration on “Hippie Hill” in Golden Gate Park on April 20, 2010, in San Francisco.

Justin Sullivan/Getty Images

7. Policies Should Foster Moderation, Not Excess

Since there’s scant evidence that casual cannabis use is harmful, “preventing” it shouldn’t be a priority, even if we knew how (which we mostly don’t). On the other hand, preventing casual users from progressing to heavy, frequent use ought to be at the top of the agenda. Pricing matters; so does marketing. Culture matters most of all, but no one really knows how to shape it (which may be just as well). Perhaps there are public-information approaches that could make progression to immoderate use less frequent, or even promote the use of cannabis short of intoxication, which is after all, the most frequent practice with alcohol; most drinking occasions aren’t drinking binges.

But moderation-promotion programs aren’t going to invent or administer themselves, and—even if we can design successful ones—realizing them will meet fierce opposition from a for-profit cannabis industry. And any attempt to rein in the industry’s attempts to promote excessive use will run into the Supreme Court’s “commercial free speech” doctrine.

8. Legalization Will Affect the Use of Other Drugs, Both Legal and Illegal, but Nobody Knows in Which Direction

In many ways, cannabis is less harmful than alcohol. Much of the discussion about legalization more or less assumes that if people smoke more they will drink less, with resulting benefits to public health and safety. That might be true, or it might not; so far, the evidence suggests virtually no effect. There is some evidence that cannabis substitutes for opiates; if that proves true, it could be among the biggest advantages of legalization. On the other hand, if Americans started copying the European pattern of using cannabis and tobacco together, the resulting uptick in tobacco smoking could be a public-health disaster. We need to study these cross-drug effects and try to craft policies to minimize combination use.

9. Taxes Should Reflect Intoxicating Power

One consequence of falling prices will be falling tax revenues, as long as taxes are based on price rather than potency. This will disappoint the voters who imagine that cannabis legalization will give them free public services. But it will also exacerbate the heavy-use problem by eroding the price barrier. The solution is simple: tax according to THC content, just as distilled liquor is taxed according to alcohol content. If the THC tax were set at $75/gram, then a gram of 10 percent THC flowers would carry $7.50 in tax while a gram of 80 percent THC concentrate would be taxed at $60. If 20 milligrams is roughly an intoxicating dose, then getting stoned would cost $1.50 in taxes, regardless of product type. By contrast, if cannabis is taxed as a percentage of price, the tax falls as market prices fall; if it is taxed per ounce, as it is in Oregon, getting stoned on superstrong “skunk” costs less in tax than getting equally stoned on weaker material.

None of the current or proposed state-level legalization schemes uses taxation by THC level; that’s one consequence of the voter initiative process, which puts a premium on policies that are easy to explain rather than policies likely to have good results.

In any case, cannabis tax revenues are unlikely to be high enough to be a major consideration in weighing legalization against the status quo; even at the current level, which is probably near the peak, Washington state and Colorado support only about 1 percent of their state budgets from cannabis taxes.

10. Voter Initiatives Are a Lousy Way to Legislate, but for Now the Alternative Is Inaction

Cannabis legalization in some form now commands a solid majority in public opinion polls. But fierce opposition from local law enforcement makes governors and legislatures reluctant to legalize. The voter initiative process, even when it isn’t captured by commercial interests, offers far less scope than the normal legislative process for careful policy design and for taking into account multiple competing interests. So, as long as the legislative channel is blocked, voters have to choose between the status quo and the imperfect versions of legalization they’re offered by advocates.

That choice would be less uncomfortable if the initiatives are easy to change when new problems arise or administrative glitches appear. That was true in both Washington state and Colorado, which have tweaked their legalization laws in response: incorporating the medical marijuana system in Washington, dealing with an edibles overdose problem in Colorado. In Arizona, by contrast, the proposition on the ballot would write a problematic, crony-capitalist form of legalization into the state constitution. (California is in between, but closer to Arizona.) Additionally, writing rigid state laws now will also make it harder—perhaps impossible—for states to accommodate their policies to whatever form of legalization Congress eventually incorporates into federal law.

We’re still learning how to do this, and we should expect to have to make midcourse corrections. As urgent as it is to rid ourselves of the burdens of prohibition—most of all, those hundreds of thousands of arrests—locking ourselves in to what may turn out to be an unsatisfactory form of legalization would be a huge mistake.

But of course the biggest mistake voters could make Tuesday wouldn’t be either adopting the far-less-than perfect forms of cannabis legalization currently on the ballot or sticking with an increasingly unworkable status quo. The worst mistake would be electing a president incapable of governing himself, let alone the country and its complex laws. (Imagine what Donald Trump’s Justice Department might do to state-legal cannabis industries.)

I’ve spent most of a lifetime thinking about cannabis policy, and I care passionately about getting it right. But first things come first, and preserving the republic matters a lot more than whether it’s legal to sell weed. So whether or not cannabis legalization is on the ballot in your state, remember that only one candidate on the ballot has a shot of getting this policy—and so many policies—right. Please do your grandchildren a favor and vote for Hillary Clinton.