Can a federal agency's policy pre-empt state law?

Justices Thomas and Gorusch write that an executive-branch policy is not "'Law' for purposes of the Supremacy Clause."


Today the Supreme Court denied review in Lipschultz v. Charter Advanced Services (MN). This petition considered whether state law could be pre-empted by a federal agency's policy. Here, the Federal Communications Commission adopted a "policy of nonregulation" of Voice over IP services.

Justice Thomas, joined by Justice Gorsuch, concurred in the denial of certiorari. They flagged an issue that the parties did not brief: is an executive-branch policy "Law" for purposes of the Supremacy Clause.

First, Thomas and Gorsuch explain that a "policy of nonregulation" should not be considered "Law," because it is not final agency action:

At the time of the founding, this Clause would have been understood to pre-empt state law only if the law logically contradicted the "Constitution," the "Laws of the United States," or "Treaties."

It is doubtful whether a federal policy—let alone a policy of nonregulation—is "Law" for purposes of the Supremacy Clause. Under our precedent, such a policy likely is not final agency action because it does not mark "the consummation of the agency's decisionmaking process" or determine Charter's "rights or obligations."

Second, even if the policy resulted from a final agency action, it is still not necessarily "Law."

Even if it were final agency action, the Supremacy Clause "requires that pre-emptive effect be given only to those federal standards and policies that are set forth in, or necessarily follow from, the statutory text that was produced through the constitutionally required bicameral and presentment procedures." Wyeth v. Levine, 555 U. S. 555, 586 (2009) (THOMAS, J., concurring in judgment).

Third, they explain that the allowing preemption in this case permits the aggrandizement of executive and judicial power, at the expense of the states:

Giving pre-emptive effect to a federal agency policy of nonregulation thus expands the power of both the Executive and the Judiciary. It authorizes the Executive to make "Law" by declining to act, and it authorizes the courts to conduct "a freewheeling judicial inquiry" into the facts of federal nonregulation, rather than the constitutionally proper "inquiry into whether the ordinary meanings of state and federal law conflict," Wyeth, supra, at 588 (THOMAS, J., concurring in judgment) (alteration and internal quotation marks omitted).

Alas, this issue was not raised in the cert petition. Therefore, the question was not before the Court:

Because this petition does not clearly challenge the underlying basis of the pre-emption theory, how- ever, I concur in the denial of certiorari.

Note to practitioners: raise this issue in the future. You have at least two votes.

A similar issue was raised in Arizona Dream Act Coalition v. Brewer. The Ninth Circuit concluded that DACA preempted an Arizona law that denied drivers licenses to deferred action recipients. Arizona's cert petition raised the exact question Justice Thomas flagged:

Did the Ninth Circuit err in assuming that the Deferred Action for Childhood Arrivals (DACA) program, an executive-branch policy of non- enforcement, was valid "federal law" capable of preempting a state police power regulation?

Cert was denied in March 2018.

NEXT: All Courts Should Abolish the Need for Amici to Seek Leave of the Parties

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. I am conflicted. On the one hand, I strongly agree with Thomas and Gorsuch that an administrative agency’s “policy” decisions ought not to carry the force of law and should not be sufficient to preempt state law. On the other hand, the states have a truly miserable track record of mismanagement, extortion and abuse of pretty much every technology issue. (See, for example, the article earlier today about NIMBYism in the 5G rollout.) VOIP does seem like the sort of issue that Congress ought to stick their noses into and formally preempt the states from mucking it up.

  2. “Can a “federal agency’s policy can pre-empt state law?”

    Can you get an editor?

    1. I think you mean Can you “can get an editor?

  3. This is exactly why Arizona should have prevailed in Arizona v United States: Arizona’s law did NOT conflict with federal law: It merely conflicted with a federal policy of not enforcing the law.

    And the supremacy clause only makes federal “law” supreme, not federal “policy”.

    Thomas took that position back then, and lost. The majority in that case are still in place. I suspect that, had the Court granted cert, he would have lost again.

    1. “Arizona’s law did NOT conflict with federal law”

      If the Constitution isn’t federal law, what kind of law is it?

  4. Baffles me. Seems like if the requirements of the Administrative Procedure Act have been followed, then administrative polices duly arrived at ought to be treated as law. Why are they different than law? Is it only because two justices prefer a policy of empowering states?

    Of course, plenty of folks want to go to war with the notion that administrative agencies can be delegated any policy-making power at all. Gorsuch and Thomas may be among those. I hope this is not a straw in a wind which will later bring news that Gorsuch is the kind of vulgar originalist who thinks his mission on the court is to pronounce unconstitutional on originalist grounds a bunch of laws he dislikes, regardless of what the historical record does not have to say on the subject.

    1. The problem with VoIP is that the FCC never defined VoIP as an “Information Service” (title 3, very lightly regulated) or a “Telecommunications Service” (title 2, heavily regulated). There seems to be several reasons for this – both technical and regulatory, but the bottom like is the FCC hasn’t, at least through a rulemaking, classified VoIP one way or the other. Instead, it has done rule making in a piecemeal fashion, usually with a “policy” of non-regulation, but not always (e.g., VoIP services are subject to CALEA, unlike typical “Information Services”).

      So, this is a bit of a different issue – I’m pretty sure Thomas and Gorsuch would not push this argument if there was a rulemaking under the APA which classifed VoIP as an “information service”. But since it’s just a “policy” their argument makes a lot more sense.

    2. Baffles me.


      Seems like if the requirements of the Administrative Procedure Act have been followed, then administrative polices duly arrived at ought to be treated as law.

      If they issue a regulation, that is law (to the extent regulations are law). This case is about them failing to act one way or the other.

    3. As a purely logical matter, I think everyone can imagine the problem with federal agencies, who are not directly elected, creating rules that are treated identically to laws passed exclusively by elected officials.

      1. Er, Article III courts are not directly elected, and they make law that just as much law as statutes passed by Congress. Your argument might need some refinement.

  5. The headline left me wondering what a “policy can” might be, and whether federal agencies keep them in the basement with the old civil defense supplies.

  6. Heck of a job, Keith.

  7. “This shall be the regulation: Nothing, and states may not add additional qualifiers.”

    Would be a nice start. Apply also to arguments “Congress considered then chose not to regulate something” implying states may not either by the dormant commerce clause. Not having enough votes to pass something is not a majority for law (much less a majority + signed bill) stating nothingness is imposed on the states for that issue.

    Put elected politicians all on the hook with actual votes then let it fall out in the courts.

  8. I agree that just as a congressional decision not to pass a law does not bind states not to pass one, a regulator’s decision not to regulate or a prosecutor’s decision not to prosecute doesn’t bind states either.

    That said, it seems to me that DACA, whatever its other legal issues, was a bit more than just a decision not to prosecute or not to regulate. It’s being more might make it inconsistent with the statute. But it also means, if it’s otherwise valid, that states aren’t operating in a vacuum.

    I think in general that to avoid constitutional issues, Congress should give all pending proposed regulations a pro-forma, up-or-down ratification vote before they go into effect if they are to be binding on states or impose penalties on imdividuals. This could be done as a routine matter of business perhaps once a quarter. I think this would solve a lot of problems. If controversial regulations are occasionally taken out by amendment and put on a slow track, this would simply be Congress exercising its proper oversight role.

    I think this approach is especially important after Chadian. Especially since Congress can’t undo a regulation once done without a veto-proof majority, it should set itself up in a role to act affirmatively, by a simple majority, before regulations go into effect in the first place.

    The vast majority of the time, this should be a simple, routine, pro-forma matter.

Please to post comments