Friday, January 19, 2018

Is there a difference between non-prohibition and authorization?

by Michael Dorf

A new National Constitution Center (NCC) podcast hosted by NCC President Jeffrey Rosen and featuring Cato's Ilya Shapiro and yours truly addresses federalism issues arising out of three controversies in the news: the pending Supreme Court case of Christie/Murphy v. NCAA; the rescission by Attorney General Sessions of the Cole memo regarding federal enforcement of the Controlled Substances Act with respect to marijuana in states that have legalized medical marijuana (and related actions regarding enforcement with regard to state-legal recreational marijuana); and the Trump/Sessions policy with respect to sanctuary cities.

The common thread running through each of these controversies is the constitutional doctrine forbidding the federal government from "commandeering" state legislative and executive officials. Because I have already commented on the marijuana and sanctuary cities issues, in this column I'm going to expand on a position I articulate in the podcast regarding the Supreme Court case, which involves the interaction of federal and state law on sports gambling.

A  good summary of the issues in (what I'll call) the NCAA case can be found on Cornell's Legal Information Institute website. To over-simplify, a provision of the federal Professional and Amateur Sports Protection Act (PASPA) makes it illegal for a state to "authorize" sports gambling. The law made exceptions for states that had on the books or enacted laws within a grace period, but New Jersey changed its law well outside the grace period. The question in the case is whether the application of PASPA to block NJ's repeal of its prior prohibition amounts to unconstitutional "commandeering" of the NJ legislature on the ground that it requires the state to maintain on the books a law it doesn't want.

Why is commandeering impermissible? Because the SCOTUS said so in 1992, in New York v. United States. The Supreme construed the 10th Amendment to forbid Congress from enacting legislation that requires a state to enact its own legislation regulating third parties. As I said during the NCC podcast, I agree with the dissenters in New York and the subsequent case involving commandeering of state and local executive officials that the anti-commandeering principle is not readily traceable to the 10th Amendment or to the constitutional structure, but I also don't have a problem with it as a policy matter. A well-written constitution might well contain an anti-commandeering principle, and it appears settled for now that the SCOTUS will treat our Constitution as containing one. The issue in NCAA is not whether to retain the anti-commandeering rule but how far it goes.

As the Court made clear in 1999 in Reno v. Condon, the anti-commandeering rule does not forbid Congress from regulating the states themselves, especially where, as in that case, the states are acting in much the way that private parties might. (Condon upheld a prohibition on the disclosure to third parties by the states of private driver license info.) Applying Condon, it is or should be clear that insofar as PASPA forbids the states themselves from conducting sports gambling, it does not run afoul of the anti-commandeering principle.

In the briefs and oral argument, it also seemed to emerge that, insofar as PASPA tells a state like NJ that it cannot completely repeal its prohibition on sports gambling, that is forbidden commandeering. After all, telling a state to keep a law regulating third parties on the books is hardly distinguishable from telling a state that it must pass a law regulating third parties.

The argument over PASPA is mostly a matter of characterization, but it's an odd sort of argument. The premise of the NCAA's argument seems to be that a state can't constitutionally be forbidden from repealing a prohibition on third-party conduct outright (because that would be commandeering) but that it can be forbidden from repealing it part-way. As the NCAA brief puts the point, NJ's "effort to channel the state’s preferred forms of sports gambling to the state’s hand-picked venues for lawful gambling was an authorization dressed up as a 'partial repeal' . . . ."

During the podcast, Ilya noted that the NCAA position doesn't make a whole lot of policy sense: Why, he asked, should a state be required to completely deregulate some practice in order to be shielded from commandeering, when the state might wish to mostly deregulate but retain some safeguards? I agree with that policy point, but I would add a further--and in my view, more fundamental--point: the NCAA's position makes no conceptual sense.

Let me explain what I mean by that. During the argument, at least some of the justices seemed to take for granted that there is such a thing as state authorization of private conduct, apart from state non-regulation of that conduct. But I think that if one uses language precisely, one sees that this is not a coherent category.

Consider the following hypothetical state laws:

L0: The state has no law regarding gambling on sports

L1: Gambling on sports in this state is forbidden.

L2: Gambling on sports is permissible only at state-run casinos.

L3: Gambling on sports is permissible only at state-licensed casinos.

L4: Gambling on college and professional basketball and football is forbidden.

One of the things that PASPA does (in states that did not enact authorizations within the grace period) is to forbid states from participating in sports gambling. Under Reno v. Condon, such a federal prohibition on the states acting themselves in ways similar to private actors is not commandeering. Thus, the application of PASPA to override the state permission for gambling at state-run casinos under L2 is unproblematic. However, under the SCOTUS state action cases, state regulation is not sufficient to convert otherwise private conduct into state conduct, so an attempt to invoke PASPA to override a law like L3 would not fall within the Condon exception to the anti-commandeering principle.

But would PASPA be justified in overriding L3 on the ground that L3 doesn't just fail to regulate sports gambling but affirmatively authorizes it? Would PASPA be justified in overriding L4 insofar as L4 is not simply a prohibition on gambling on basketball and football but a tacit authorization of gambling on other sports? I don't see why one would think either of those things.

The state can take three approaches to private conduct: It can do nothing with respect to the conduct; it can forbid the conduct; or it can require the conduct. There is no fourth category of state authorization of conduct. Authorization is simply a confused way of describing the state doing nothing.

Wait, you say. Always? Suppose a state has legalized marijuana for anyone who obtains a prescription from a licensed physician. Isn't the state thereby authorizing medical marijuana for people with scripts from licensed physicians?

Well, yes, in a colloquial sense, but no in a fundamental sense. The anti-commandeering doctrine draws a distinction between Congress forbidding the state from acting in various ways (permissible preemption or regulation of the states per Condon) and requiring the states to regulate (impermissible commandeering per New York and/or Printz v. United States).  In order to make sense of that distinction, one needs a way to describe the baseline against which to measure whether the state is acting or not acting, and the baseline can't simply be whatever the law happened to be before the state passed its latest law, because a federal law either commandeers or doesn't commandeer state action; it doesn't depend on what state law happened to be in each state prior to the enactment of the federal law.

Therefore, in my hypothetical marijuana prescription case, one shouldn't think of the baseline as some prior state law that categorically forbids marijuana use. Rather, the baseline has to be no law--the marijuana equivalent of L0. Seen in that way, the medical marijuana license is not a state authorization to use marijuana. Rather, we should understand the state law as forbidding marijuana use (or possession or whatever) by people who lack valid scripts. People who have valid scripts for medical marijuana are effectively unregulated. A federal law that said (for example) that the state can't authorize medical marijuana use and applied to override such a state law would thus be effectively commandeering the state to regulate medical marijuana as to those people who previously were not forbidden from using it.

Bottom Line: As I said during the podcast, whatever else the SCOTUS does in the NCAA case, it would be very confusing and unfortunate if it announced a new wrinkle in anti-commandeering doctrine distinguishing between state failure to regulate private conduct and state authorization for private conduct.


Joe said...

I think "authorization" here means something in more than a "colloquial" sense and that might be part of the confusion. This is separate from it mattering for the purpose of commandeering and 10A rules. But, in general, doing nothing and having regulations are different things, including I gather legally in various ways.

Jim said...

Doesn't this argument effectively nullify the "authorize" language of the statute? Apart from the narrow circumstance of your L2, I don't see how a state law could be construed as "authorizing" gambling under your understanding of this term. The background state of things will always be that everything is permitted unless a state law prohibits it, so no state law ever "authorizes" any conduct? Am I missing something?

Michael C. Dorf said...

Jim: No, you’re not missing anything. I think that state law that “authorizes” private conduct by providing an exemption from what would otherwise be a prohibition is simply a convoluted way of saying that state law does not forbid that private conduct. Therefore, a federal law forbidding such authorization is a federal law requiring the state to forbid the conduct—which is constitutionally forbidden commandeering.

I am not saying that there are no contexts in which we might sensibly want to distinguish between authorization and failure to prohibit. I am saying that, given the premises of the anti-commandeering doctrine, this is not such a context.

Michael C. Dorf said...

I received the following interesting comment via email:

You say: “The state can take three approaches to private conduct: It can do nothing with respect to the conduct; it can forbid the conduct; or it can require the conduct. There is no fourth category of state authorization of conduct. Authorization is simply a confused way of describing the state doing nothing.”

Licensing a business, for example, is none of those three things. The state obviously hasn't forbade or required the conduct. But nor has it "do(ne) nothing." It has affirmatively granted an authorization slip that benefits the business by using the coercive power of the state to bar competitors who don't have a similar slip.

The same dynamic would be true for lots of other ways that states encourage private conduct. If, for example, a state gave a private business all sorts of unique and valuable services - e.g., heightened police protection, dedicated utilities, etc - that would seem like a lot more than "doing nothing," and a lot like "authorization" to me.

Here’s my response:

I agree that there are contexts in which licensing is sensibly conceived of as authorization, but I don’t think it’s true in a context, like anti-commandeering doctrine, where the default is no state law. E.g., a driver’s license or a barber’s license is simply an exception to the state’s otherwise-blanket prohibition on driving or haircutting.

Meanwhile, I see encouragement as either de minimis or state participation, which could take us out of the anti-commandeering area and into Reno v. Condon.

Greg said...

I have to say, I'm inclined to believe that L3 is a case where there is potential authorization.

To put it another way, perhaps the state could enact L3, but under Reno v. Cordon, the state could also be prohibited from issuing any of those licenses. This has the effect of prohibiting the law, even though the action through which it does so is preventing the issuing of licenses which authorize sports gambling.

Implementing L4 would require no such action by the state that would fall under the Cordon exception.