[Cross-posted from the Washington Monthly’s Ten Miles Square]
Public opinion about marijuana has moved sharply over the past few years; about half now support legalization. Tuesday voters in Washington State and Colorado took the leap of approving commercial production and sale for non-medical use, under more or less the same rules that apply to alcohol.
Ideologues on both sides are claim to know with certainty what the results of legalization would be; all good in the view of the legalization advocates, all bad in the view of those who support the current laws.
But those of us who try to study the issue scientifically find ourselves in a world of doubt. How much lower would legal prices be than current illegal prices? If there were heavy taxes, how much evasion would there be? Would buyers in a legal market favor possibly more dangerous high-potency varieties, or would lower-strength products dominate the marijuana market as beer dominates the alcohol market? Would legalization greatly increase problem marijuana use? Use among teenagers? (That might depend on the price.) Would there be an increase in auto accidents due to stoned driving? Would problem drinking decrease â€“ or increase â€“ as result?
All of those questions matter. None of them can be answered by abstract reasoning, or by studying small variations in marijuana policy such as decriminalization of possession for personal use. The only way to find out how legalization would work in practice is to actually try it.
But actually trying it on a national basis carries heavy risks. If it goes badly â€“ if, for example, heavy use and use among teenagers quadrupled â€“ it would be very hard to put the genie back in the bottle. All those new users would become potential customers for an expanded illicit market if the drug were re-prohibited.
So the obvious way to learn something about marijuana legalization would be to try it out one state at a time: relying on what Justice Brandeis called â€œthe laboratories of democracy.â€ If Coloradoâ€™s legalization went badly, that would be a much easier problem to correct than if the mistake had been made on a national basis.
Of course, itâ€™s not really possible to legalize in a single state with the federal prohibition in place; making growing and selling marijuana legal under state law doesnâ€™t make it any less of a federal crime. But in practice most enforcement is done at the state and local level, so removing the state law would allow us to learn quite a lot about the consequences of full legalization.
Virtually no one advocates completely unregulated trade in marijuana: for example, there is near-universal support for a ban on sales to minors, for product-labeling requirements, and for a substantial tax to prevent a drastic price decrease and help deal with state budget crises. All of those are written into both the Washington and Colorado laws, and one of the things we could learn from their experience is how well that regulatory process works and how hard it is to collect those taxes.
But the federal government could shut down both of those experiments, if it were determined to do so. Everyone who applies for a license to grow or sell marijuana is, in effect, asking the state for permission to break the federal law, and that list of applicants could become a list of targets for federal drug-enforcement agents.
That approach would please the drug warriors. But it would make it impossible to learn anything useful from the Colorado and Washington experiments.
So why shouldnâ€™t the federal government cut Colorado and Washington some slack? As long as those states prevent marijuana grown under their laws from crossing state lines and thereby subverting marijuana prohibition in the rest of the states, the Justice Department could step back and let the consequences of the new policies play themselves out. They might succeed, or they might fail. In either case, the rest of us could learn from their experience.