Wednesday, February 5, 2014

Marijuana rescheduling and the "potential for abuse" factor

President Obama's comments about marijuana last week prompted some renewed discussion about the status of marijuana as a Schedule I substance.   

A few days ago, Mark Kleimen weighed in, saying that marijuana rescheduling is essentially beside the point.  Since marijuana still would not have FDA approval, Kleiman argued, growing marijuana "would still be the illegal manufacture of a Schedule II controlled substance."  Although overstated (and full of odd and off-base personal attacks against Jacob Sullum), Kleiman's basic point is valid and often overlooked: rescheduling marijuana would not solve the conflict between state medical marijuana laws and the federal Controlled Substances Act.  (Unlike Kleiman, I do not think rescheduling would have "zero" practical effect--it would have a significant political impact and could provide space for litigation on the legality of distributing marijuana without FDA approval, for example as an herbal supplement.)

Putting the question of what impact rescheduling might have aside, however, I just saw an update to Kleiman's post that struck me as misguided.  In the update, Kleiman claims that marijuana could not be moved below Schedule II because "more than 2 million people in the U.S. meet diagnostic criteria for cannabis abuse or dependency at any one time."

Kleiman's position stems from the federal Controlled Substances Act's three scheduling criteria, one of which is a substance's relative "potential for abuse."  The law provides that substances in Schedules I and II are those with a high potential for abuse.  Schedule III substances have a potential for abuse less than those in Schedules I and II and so on.

The trouble is, the CSA does not define the term "potential for abuse."  (In fact, the only term in the CSA's scheduling criteria that is expressly defined, is "United States.")  The result--as anyone with a basic familiarity with administrative law can guess--is that the DEA has enjoyed incredibly broad discretion to interpret and define "potential for abuse" and other scheduling criteria.  

And here's where Kleiman's position is not as air-tight as he seems to think it is.  Currently, the DEA defines "potential for abuse" in a way that equates, roughly, to overall use rates.  And if we apply this definition (as the DEA does), Kleiamn is right: marijuana's abuse potential would place it in Schedules I or II.  

But there are plenty of other reasonable ways to define "potential for abuse."  And the only thing stopping the DEA from adopting a different definition of "potential for abuse" is, well, the DEA.  Instead of focusing on the total number of users, for example, we might define "potential for abuse" based on the percentage of users who become addicted to a substance or based on the ancillary harms that come from regular use.  (Indeed, many people seem to think idea that marijuana's abuse potential is the same as heroin's is pretty ridiculous.  Presumably, folks in this category think that there are other measures of abuse potential than Kleiman's/the DEAs.)     

This is not to say that marijuana would necessarily end up with a lower abuse potential rating if the DEA decided to revise its definition of the term.  My point is only that it could and that there are certainly reasonable definitions of "potential for abuse" in which it almost surely would.  Kleiman's position that marijuana's abuse potential means it must remain in Schedule I or II misunderstands the way administrative law works and the DEA's power to interpret "potential for abuse." 

I examined the DEA's definition of "potential for abuse" in some detail in this article for the Albany Government Law Review last year.

Federal Marijuana Laws, Policies and Practices, Medical Marijuana Commentary and Debate | Permalink


Hi Alex--As always, really enjoy reading your posts (and feel like I learn a lot from them). Can you say a bit more about what you mean by "Currently, the DEA defines "potential for abuse" in a way that equates, roughly, to overall use rates." (I know you referred me to your article, but I was wondering if you could just give me a sneak preview). Again, really interesting stuff--give me a shout if you're ever in the Bay Area!

Posted by: W. David Ball | Feb 5, 2014 1:41:03 PM

Good to see your comment! I'll actually be up in the bay a couple of times in March, I'll email you re: the details.

Sure thing re: the potential for abuse issue. I was possibly being a bit generous to say the DEA has a single definition for the term. Probably more accurate is that the agency relies on factors in analyzing the issue that mean widely used substances will be found to have a high abuse potential. the DEA's analysis of marijuana in a 2011 denial of rescheduling is a good example. The DEA cited the following factors in finding marijuana has a high abuse potential relative to substances in Schedules III and lower: "the prevalence and frequency of use in the general public and in specific sub-populations, the amount of the material that is available for illicit use, the ease with which the substance may be obtained or manufactured, the reputation or status of the substance 'on the street,' as well as evidence relevant to population groups that may be at particular risk."

The first three of those mean that a widely-used substance like marijuana will be found to have a high abuse potential, even if the percentage of users who become addicted is low (relative to other substances) or the harms that come with regular use is low (relative to other substances), etc.

This approach helps explain how Marinol (synthetic THC) can be Schedule III while marijuana is Schedule I. At first glance, one might understandably think that the abuse potential for synthetic THC would have to be the same as for marijuana. But the DEA's way of interpreting abuse potential means they classify the abuse potential for Marinol to be lower than for marijuana simply because lots of people use marijuana and not that many use Marinol (or at least so it appears--the DEA administrative decisions are far from models of clarity.)

Another interesting related note, a first circuit decision in the 1980s on the scheduling of MDMA (ecstasy) found that the DEA had not articulated any standard for analyzing relative abuse potential. Nevertheless, the court upheld the DEA's finding on the grounds that the agency could reach a conclusion simply by comparing a substance to others that are already scheduled. Court decisions like that help explain why the DEA's interpretation of "potential for abuse" (and other CSA terms) is so open-ended, even more than 40 years after the law's passage.

Posted by: Alex Kreit | Feb 5, 2014 3:35:25 PM

Fascinating. So it seems like it doesn't matter what the numerator is, as long as the denominator is sufficiently large. As long as millions of people use it, then the potential for abuse is large--even if a small percentage of them abuse it. That seems to get the potential for abuse entirely backwards, in that it's not about relative numbers, but about absolute numbers. Thanks for clarifying. One last question--have you (or anyone else) written anything on Chevron and the DEA potential for abuse standard? Or is there a good case about it? (I'll respond to the travel stuff offline). (The other) DB

Posted by: W. David Ball | Feb 5, 2014 4:32:18 PM

One additional point, just to make sure I have the logic right. If there were a single person addicted to a substance (let's call it Juanamari), and more and more people started using it without getting addicted, then as the numbers of non-addicted people rose (and the observed likelihood of addiction fell), then it seems the DEA's potential for abuse would actually increase. Is that right?

Posted by: W. David Ball | Feb 6, 2014 8:59:12 AM

I think that is right. Though the DEA's analysis in scheduling decisions is so open-ended that it's not easily predictable. Ultimately, I think the DEA is very results oriented in its scheduling and that it has left its definitions of the 3 scheduling criteria as general as possible to give itself maximum flexibility (Marinol's status as Schedule III vs. marijuana's as Schedule I is again a good example of this, I think.) There hasn't been too much written about the CSA and Chevron (on the potential for abuse standard or otherwise). The most detailed treatment that comes immediately to mind was from the 1st Circuit in DEA v. Grinspoon in the late 1980s. For a statute that's been in place for 4 decades, the CSA's regulatory structure in general is surprisingly under analyzed (both in courts and, especially, by academics.) Most of the commentary about the CSA's scheduling process focuses on marijuana's status--especially the question of accepted medical use--rather than the bigger picture issues. It would be great to see a thorough exploration of Chevron and the CSA (I've thought of taking a crack at it myself but so far haven't put it far enough up in the project queue to really get anything going on it.)

Posted by: Alex Kreit | Feb 7, 2014 4:10:09 AM

Post a comment