Eugene Volokh points to this post by Lawrence H. White at Division of Labour about an apparent mistake in elementary economics by Justice Stevens during oral argument on Ashcroft v. Raich, the medical-marijuana case.
Actually, “apparent” isn’t quite correct: Justice Stevens made an actual mistake in elementary economics, but in this case the relevant economic analysis may not be the one in the elementary textbooks.
The background is that the Supreme Court was considering whether Congress had the power to prohibit the sale of marijuana grown and distributed entirely within California under the California medical marijuana law (Prop. 215). One argument on the feds’ side of the case is that even if the marijuana is entirely local as a physical matter, it’s still part of a national market, and therefore properly under Congressional authority over interstate commerce.
Justice Stevens asked Randy Barnett, arguing the case for Angel Raich, what the effect would be on the national marijuana market if Californians were allowed to grow, sell, and use intrastate marijuana for medical purposes. Barnett replied, correctly, that any such effect would be trivial, but Stevens wanted to know whether the national price of pot would be higher or lower, however slightly. Barnett gave the textbook answer, but Stevens initially wasn’t convinced:
MR. BARNETT: Well, it would reduce demand and reduce prices, I think. But -
JUSTICE STEVENS: If you reduce demand, you reduce prices? Are you sure?
MR. BARNETT: Yes.
JUSTICE STEVENS: Oh, you’re right. You’re right. Okay. Yeah. Yeah.
Clearly, that wasn’t the answer Stevens wanted to hear, and obviously it wasn’t the one that Barnett wanted to give. But it was indeed the textbook answer.
That’s not to say, however, that it was the correct answer.
In general, we expect the supply curve to slope upward because low-cost production opportunities are scarce and the least-cost production opportunities get used first: If demand for oil fell, so would the price, because the highest-cost wells could be shut down.
In addition, a surprise reduction in demand for any product will leave producers and resellers with unsold inventory, thus forcing prices down in the short run: When GM discovers that even GM consumers aren’t dumb enough to buy its latest automobile parody, it starts to pile on the sales incentives.
But under some circumstances the cost of producing an item falls as the quantity produced rises, because of “learning by doing” and because larger volumes can spread the fixed costs of engineering and design, and force down middleman markups; think of microchips and other electronic goods, for example.
In the case of marijuana and other illicit drugs, the dominant cost facing any producer is the cost imposed by law enforcement (employees and principals alike need to be compensated for their risks of arrest and imprisonment) and the cost of evading law enforcement.
Roughly speaking, the enforcement risk faced by the average drug transaction depends on the ratio of the volume in that market to the enforcement effort devoted to suppressing it. If there’s one dealer on a street corner and one cop patrolling it, the dealer is much more likely to get busted than if the same cop confronts 100 dealers. That “safety in numbers” principle is why prey animals herd.
So if we take some of the demand out of the illicit marijuana market in a way producers in that market can predict, they will likely reduce the amount they produce. If we leave the enforcement effort constant, each remaining kilogram of pot faces more law enforcement. Thus we would expect the price of illicit pot to rise (trivially, as Barnett noted, because the proportion of total cannabis demand that is “medical,” even under California’s loose standards, is a small fraction of the total) as a result of removing medical demand from the market.
(Of course, that’s only a partial analysis. If having an exemption for medical pot made cases against people growing pot for the non-medical market harder to investigate and prosecute, the result might be to reduce the effective level of enforcement pressure. And if the existence of a large, open medical-cannabis market somehow increased demand for cannabis for non-medical use — for example, by making the drug seem safer to potential users — that might easily swamp any direct effect of removing patients’ demand from the illicit market.)
Shortly after the case was argued, I had an email exchange on this topic with another member of the tiny community of academics who think about the economics of the illicit markets. He pointed out, reasonably, that I should have anticipated that this would be an issue and filed an amicus brief making the argument above.
Why didn’t I? Not because of my preference for clinical research over litigation as a strategy for making cannabis medically available; I’d almost always rather have the courts decide on the actual facts. The reason was simpler: I simply hadn’t been following the case closely enough to figure out that the price effect would be an issue.