Monday, April 06, 2015

The Supremacy Clause and Federal Common Law

by Michael Dorf

One of my favorite games to play when I teach Federal Courts is "what kind of a holding is this?". Let me explain with an example. In Bivens v. Six Unknown Named Agents, the Supreme Court recognized a cause of action for a person alleging that FBI agents violated his Fourth Amendment rights. What was the basis for the Court's holding? The Fourth Amendment? Some general constitutional principle requiring a remedy for every violation of a constitutional right? If so, is that a free-floating principle or is it located in some particular provision, such as the Due Process Clause (here, of the Fifth Amendment)? Or perhaps this is, as later cases indicate, not a constitutional holding at all. But then what is it? The seemingly best answer is federal common law, but true federal common law can be displaced by Congress without Congress providing any substitute. So perhaps Bivens is what is sometimes called constitutional common law, i.e., a set of judge-made doctrines that implement the Constitution and that may not be displaced by Congress unless Congress provides an adequate substitute. I often give students a list of options, typically concluding with "one or more of the above" and then "none of the above." As I said, it's one of my favorite games. I can't say the students enjoy it all that much, especially as I use it when I've cold-called someone.

Nonetheless, I play this game with my students for two main reasons: (1) It exposes how sloppy the Supreme Court itself often is, making pronouncements about what the law requires, often without paying any attention at all to the source and nature of these pronouncements; and (2) analyzing the source and nature of a particular holding is essential to understanding the scope and limits of that holding.

Last week, the SCOTUS played the game with respect to the Supremacy Clause.  In Armstrong v. Exceptional Child Center, the Court, 5-4, held that Medicaid providers could not sue Idaho state officials for violating the federal statute establishing reimbursement rates. The plaintiffs, respondents in the Supreme Court, had said that they had a cause of action under the Supremacy Clause to enjoin the state officials to increase the reimbursement rates to comply with federal law. However, the Court found that provisions in the Medicaid Act authorizing the federal government to withhold money from the state were the exclusive remedy for the sorts of violations alleged by the respondents.

En route to that decision, the Court said that the Supremacy Clause itself does not create a cause of action for parties to sue state officials on grounds that they are acting inconsistently with federal law. True, the majority acknowledged, such a cause of action is generally available, but that's not a function of the Supremacy Clause itself--which is just a priority rule.

Nonetheless, the majority acknowledged that absent congressional displacement, there is a cause of action by private parties suing government officials to enjoin their conduct as inconsistent with federal law. Where, if not from the Supremacy Clause, does this cause of action come from? According to the majority, it is a judge-made remedy given by courts of equity.

The SCOTUS doesn't use the term "federal common law" in Armstrong, but it's pretty clear that the general remedy it describes is best understood as federal common law. I suspect the Court abjures the term for two reasons: (1) The author of the Armstrong majority is Justice Scalia, who is skeptical of the legitimacy of federal common law (except when he isn't); and (2) it's awkward to describe a principle of equity as a species of common law, because traditionally courts of equity were distinct from common law courts.

Nonetheless, awkward or not, the majority characterizes the general rule in terms that sound in federal common law: a judge-made rule that may be displaced by Congress.

The dissent disagrees with the majority's conclusion that Congress in fact displaced the cause of action that is usually available to enforce the Supremacy Clause. I don't have a lot to say on that point. Interested readers should check out Steve Vladeck's critique at Prawfsblawg. I would note, however, that despite the disagreement over the outcome of the case, the Armstrong dissenters nonetheless grudgingly agree with the majority's characterization of the displaced cause of action. Writing for the dissenters, Justice Sotomayor says "the Court is correct that it is somewhat misleading to speak of 'an implied right of action contained in the Supremacy Clause' . . . ."

Despite the Court's unanimity with respect to the characterization, I want to argue that all the Justices are wrong, at least in some of their broader language. To my mind, the majority and the dissent lump together two questions that should be treated separately. The questions are: (1) What does Congress need to do to displace an equitable remedy to enforce a federal statute against government officials? and (2) What, if anything, can Congress do to displace an equitable remedy to enforce the Constitution against government officials?

Making clear that the dissenters (like the majority Justices) see these questions as no different, Justice Sotomayor writes: "A claim that a state law contravenes a federal statute is 'basically constitutional in nature, deriving its force from the operation of the Supremacy Clause,' . . . and the application of preempted state law is therefore 'unconstitutional' . . . ."  The ellipses indicate omitted citations, which shows that Justice Sotomayor did not originate this conflation of statutory and constitutional injunctive actions. Nonetheless, the Court ought not to perpetuate it.

When Congress creates rights and duties by statute it can (subject to limits arising out of the "procedural due process" doctrine) circumscribe the remedies for those rights and duties. It can, for example, make administrative enforcement the exclusive remedy for a violation. Put differently, we might understand the ostensibly judge-made remedy of a cause of action for injunctive relief against government officials as not really federal common law but really a kind of statutory interpretation. And indeed, Justice Sotomayor's dissent reads in this way in some places, when explaining why Congress oughtn't to be understood to have intended to displace the cause of action for injunctive relief. The basic idea is that Congress, as the creator of the substantive rights and duties, also controls the remedies.

But a cause of action to enforce the Constitution stands on a different footing. Congress has no power to immunize violations of the Constitution. True, the Court has held that states and the federal government enjoy sovereign immunity against certain kinds of lawsuits but the Court has also held that such sovereign immunity does not shield government officials from actions for injunctive relief against future violations of the Constitution. That is the doctrine of Ex Parte Young.

So let's play my game. What kind of a holding is Ex Parte Young? The Armstrong dissent (apparently agreeing with the majority on this point) characterizes it as essentially ordinary federal common law, i.e., as a judge-made rule that can be displaced or even eliminated by Congress. To be sure, the dissenters want a much clearer statement of congressional intent to displace an Ex Parte Young action than the majority requires, but for all of the Justices, a sufficiently clear statement will suffice.

That's clearly right with respect to a case like Armstrong itself because Armstrong is the first kind of Young action--one that seeks to enforce a statute, over which Congress is the master. But what about an Ex Parte Young action to enforce the Constitution? By treating all Ex Parte Young actions as "constitutional"--and then by allowing that they are all subject to being displaced by Congress--the Court unanimously implies that Congress may eliminate an Ex Parte Young action to enforce a genuine constitutional provision, even if Congress does not supply an adequate substitute remedy.

I think that's wrong. In my view, an Ex Parte Young action to enforce a genuine provision of constitutional law (as opposed to a statutory obligation masquerading as the Constitution because it applies via the Supremacy Clause) is not ordinary federal common law. It's constitutional common law. That is, it's the kind of judge-made remedy that Congress cannot eliminate unless Congress provides an adequate substitute.

I don't have the space in a blog post to explain the full scope of this notion of constitutional common law. I refer interested readers to this 2000 article that Barry Friedman and I wrote, in which we elaborate in a different context the notion that the Court can create a remedy that Congress can replace with an adequate substitute but that Congress may not simply eliminate. As we note there, the crucial point is not whether one calls this concept "constitutional common law" or something else. The crucial point is that Congress cannot simply eliminate constitutionally required remedies for constitutional violations. (For other subtle elaborations on the scope of congressional power to adjust remedies for constitutional violations, I recommend a 1991 article by Dick Fallon & Dan Meltzer as well as a 2013 article by Carlos Vazquez and Steve Vladeck.)


Shag from Brookline said...

After Erie, it seems to be a bit eerie to distinguish "ordinary federal common law" from "constitutional common law," neither of which terms is referenced in the Cons\stitution. How can the non-fictional "The Originalist" in Armstrong neuter the Supremacy Clause and be true to originalism? Well, he does this under the justice-made power of horizontal judicial supremacy per ... [drum roll] ... the Supremacy Clause.

Michael C. Dorf said...

To be fair to Justice Scalia, even after Erie, EVERYONE recognizes that there is still some federal common law. E.g., in admiralty cases, cases involving contracts with the United States, interstate boundary disputes, etc. And Justice Scalia has resisted the category of constitutional common law (although wrongly in my view).

Joe said...

I think the "resistance" is the telling point.

Hashim said...

"Congress has no power to immunize violations of the Constitution."

Oh really? What about where (as was true for most of this Nation's history) Congress had not conferred on lower federal courts general federal-question jurisdiction, either at all or at least not for claims below a certain amount in controversy?

It's pretty hard to argue that Congress acted unconstitutionally, given 1) the Madisonian Compromise that Congress controls existence/jurisdiction of lower federal courts; 2) the history and tradition of how Congress exercised that jurisdiction; and 3) the neutral nature of Congress' jurisdictional decision, which applies generally to all federal claims rather than specifically targeting federal constl claims for disfavored status.

Congress' valid determination about the limited jurisdiction of lower federal courts would immunize federal officers from injunctive relief for federal constl violations. No other court could issue such relief. The US SCt definitely lacks original jurisdiction over such claims. See Art. III; Marbury. And state courts most likely lack jurisdiction over such claims, as they definitely can't issue mandamus or habeas corpus against federal officials, and the (well-reasoned) weight of authority says they likewise can't issue injunctions against federal officials. See Hart & Wechsler's discussion of Tarble's Case, et al.

Hashim said...

Or, to take a modern example, what if the President *himself* was acting unconstitutionally in some way (not through a lower executive officer). The SCt has held that the President isn't covered by the APA, and that he can't be enjoined in the performance of non-ministerial duties (and likely can't be enjoined at all). See Franklin v. Massachusetts. So Congress' refusal to include the President in the APA has immunized his actions, unless you're going to say that a state court can issue injunctions against the President. (In which case the next two years would be interesting indeed.)

Michael C. Dorf said...

Hash: I raised just that possibility in class today but then noted that the point of the Madisonian Compromise is that state courts are considered adequate fora for the vindication of federal rights. That's why the Supreme Court has never construed any act of Congress as completely stripping both state and federal courts of jurisdiction over some class of constitutional claim or defense. When Congress seemed to do so in the Portal-to-Portal Act, the Second Circuit in Battaglia stated (in very influential dicta) that it acted unconstitutionally.

Your example with the President is much like the immunities I was bracketing. In any event, I did NOT say that there is a judicial remedy for every constitutional violation. What I said was that where the Court says there is a remedy for a constitutional violation, Congress cannot simply make that remedy go away. That is the argument of my article with Friedman in discussing the Dickerson case.

Hashim said...

The Presidential issue isn't an "immunity" like sovereign immunity. I don't read Franklin to be saying that Congress *cannot* expressly authorize injunctive relief against the President. Rather, i think it's merely saying that no such relief is available *unless and until* congress expressly authorizes it. Thus, Congress *can* simply make such a remedy go away. And if it can do it for the President, I don't see why it can't do it for every lower federal officer too.

As for your more general point, if you're right about the purpose of the Madisonian compromise, then state courts would need to have the power to issue injunctions against federal officers. But McCllung v. Silliman and Tarble's Case strongly suggest otherwise.

Michael C. Dorf said...


Saying that Congress can authorize but hasn't authorized a lawsuit against the President is quite different from saying that there is a constitutional right to sue the President (or anybody else) but that Congress can simply take it away.

As for Tarble's etc, reading those cases for all they're worth sits very uncomfortably with both early 19th practice and modern practice regarding constitutional rights. See, e.g., pages 402-06 of the 6th ed. of Hart & Wechsler.

When I teach this material in Federal Courts, I note that one cannot reconcile all of:
(1) Tarble's Case, if read as a constitutional bar;
(2) The "traditional" view of the Madisonian Compromise and the Exceptions Clause that you are apparently espousing (but that has been rejected by Justice Story, Henry Hart, Akhil Amar, and many others);
(3) Boumediene's recognition of a right to habeas.

Perhaps at some future time I'll elaborate fully why something has to give, but for now I'm going to let go of this extended discussion of what I regard as a side point--namely, your broad view of the substantive remedial implications of your broad view of congressional jurisdiction-stripping power.

Hashim said...

All fair enough. For you to ponder, two final thoughts:

1. Sorry if I'm being dense, but given Franklin v. MA and your remedial views, don't you have to say either: (1) there's no constitutional right to sue the President for injunctive relief for ongoing constitutional violations; or (2) Congress has acted unconstitutionally by failing to authorize such relief, and the SCt has acted unconstitutionally by failing to provide it regardless?

If you're saying (2), then that's consistent with your criticism of Armstrong, but it seems to me a remarkable position, especially given the Madisonian Compromise (as it would essentially require as a constitutional matter that state courts be allowed to enjoin the President himself simply because Congress hadn't created lower federal courts).

Whereas if you're saying (1), then it seems inconsistent with your criticism of Armstrong, unless you have some theory to support what seems to me to be a remarkable proposition that the constitution requires remedies against lower federal officers but not against the President himself even when he acts unilaterally.

2. I entirely agree that Boumediene can't be reconciled with a constitutional version of Tarble's Case plus the "traditional" view of Congress' jurisdictional power. For what it's worth, I actually think that Boumediene and Tarble's Case are *both* wrong: the Suspension Clause doesn't cover federal habeas at all (as such coverage makes no sense since Congress never had to create federal habeas to begin with), but the Suspension Clause does protect the ability of state courts to grant habeas against federal custodians (though that wouldn't do foreign prisoners held abroad any good, since no state court should have jurisdiction to issue the habeas writ in such circumstances).

For purposes of our present discussion, though, the key to my view is that the Suspension Clause protects only state habeas over federal custodians, not state injunctions over federal officials more broadly. So I'm still skeptical that there's any textual or historical basis for state courts issuing in personam injunctions against the President or any lower federal official. And thus I'm still skeptical that the Constitution requires that such injunctions be available, since Congress didn't have to create lower federal courts, much less confer jurisdiction to grant this type of relief.

Anyway, fun stuff. Thanks for chatting about it. Of course happy to discuss more, but totally understand if you're moving on for now.