No, the Presidential Records Act Is Not Unconstitutional

A few days ago, the Office of Legal Counsel issued an opinion to the effect that the Presidential Records Act is unconstitutional. Now, I have said that I don't expect to use this newsletter to do commentary on current events very much in the Trump era, because constitutional controversies these days are all so dumb. This one is no exception. The opinion... it's bad. You can find all manner of commentary to that effect elsewhere. But I wanted to write about it, because I think the central error is one that ties directly into what I have just been writing about in Constitutional Perspectives in the past few weeks, about the nature of congressional power generally.

Okay, first, the basics. What's the Office of Legal Counsel? Well, it is, as the name suggests, an office, located within the Department of Justice. It serves as a kind of internal court for the executive branch. When other agencies encounter unclear questions of law, they can bring them to OLC and get advice. What makes it kind of quasi-judicial is that OLC opinions are binding on the entire executive branch. Including OLC itself: in general, OLC does not rescind or overrule its opinions.

OLC has a not-great reputation these days, dating back to when it produced some infamous opinions during the War on Terror endorsing the legality of, among other things, torture. But the idea of having something like OLC is not a bad one. Much of the law that applies to the federal executive is not really susceptible of being enforced by anyone outside the executive branch. It is a good thing for the executive to make an effort to follow the law, even when there is no one to make it do so. It is a good thing for there to be an institutional mechanism for this.

Unfortunately, in practice OLC often operates less like a court and more like an advocate, staking out maximal views of executive power, which are then clothed with the pseudo-authority of an OLC opinion. I say "pseudo" because, of course, OLC opinions have no force of law, not in the way that a judicial decree would. They are only statements of the executive branch's view of the law; for the purposes either of the courts or of Congress, they might as well not exist. (Or, to be a bit more precise, they are not materially different from a legal brief filed by the Department of Justice: perhaps successful in advocating their position, but of no authority in themselves.)

Okay, next question: what's the Presidential Records Act? It's a law passed by Congress in 1978 imposing a number of requirements relating to, well, presidential records, papers and the like. Most notably, the PRA says that these papers are the property of the United States, not of the president who made them. It also imposes certain affirmative obligations on the president to keep records, and creates a process for archiving presidential records after the president has left office. Eventually, these records are, with some exceptions, to be made public.

Of course it is not a coincidence that this law was passed in 1978, in the wake of Richard Nixon's presidency: irregularities in the keeping of presidential records had been a constant theme in Nixon's various scandals, and he had tried to assert ownership of all his papers as a way of keeping them from ever seeing the light of day. Congress wasn't having it. Carter signed the PRA into law, and to my knowledge, no president before Trump had ever particularly taken issue with any of its requirements. Trump, however, had advanced an essentially-delusional "interpretation" of the law on which it gave him total control over his papers, and therefore he couldn't be prosecuted for improper retention of classified materials.

And now he has gotten OLC to call the whole law unconstitutional. Fun times!

Reading this opinion was a strange experience for me, because the whole time I was waiting for OLC to address what I saw as the main issue, and they just... never did. Explaining the error in OLC's analysis has, therefore, a strange, kind of shadow-boxing quality to it: I don't know exactly what they would say about what I see as the central question. Instead, the bulk of the opinion is devoted to what I see as sideshows. And I don't exactly disagree with some of the conclusions OLC reaches about those sideshows, though they say an awful lot of aggravating and plainly false bullshit along the way. I'll say a little about some of that aggravating bullshit, but I mostly want to discuss the issue as I perceive it.

The opinion spends a lot of time discussing the history of congressional requests for presidential records. Starting with George Washington, just about every president has received such requests, and starting with Washington, many if not most presidents have asserted a power to decline them, at least in part. OLC, looking upon this history, sees a requirement, evidently of constitutional stature, that Congress negotiate for access to presidential records, rather than just claiming a right to them.

This part of the opinion repeatedly refers, albeit a little obliquely, to something James Madison once said about how, where the meaning of the Constitution is ambiguous, subsequent practice can "liquidate" its meaning. Now, you might notice that there is something strange about this. There is no provision in the text that can be thought to speak to this, certainly not with any specificity. Whatever the relevant constitutional principles are, they must necessarily be inferred from general structural features; it is not really a matter of "meaning."

I think it's quite clear what OLC is up to. This line of Madison's is a favorite of originalists, because it allows them to perform a methodological sleight-of-hand (which is what they're all about). Practice subsequent to ratification, more or less by definition, is a matter of precedent. It is, in other words, a doctrinal argument. But Madison's concept of "liquidation" allows originalists to transmute subsequent practice into "original meaning." To convert, in other words, a doctrinal argument into a historical one.

And this is utterly essential for what OLC is trying to do here! Because, as I have observed before, if we're doing doctrinal argument, the later precedent prevails over the earlier one. This is the nature of any system of decisional law: decisions can be overturned! Accepting, then, that the pre-Nixon arrangement was as OLC describes it, then what happened is that a new circumstance (Nixon's criminality) caused the relevant authority (Congress) to change its mind! Doctrinal argument more or less never allows you to invoke an earlier authority to undermine a later one. Historical argument, on the other hand, is all about that: subsequent developments are deemed invalid in light of the original meaning. The use of Madison's "liquidation" language is meant to transmute the one into the other, and therefore create a kind of chimerical argument.

Also, I think the way OLC characterizes the pre-Nixon precedent is questionable. Essentially every time a president refused to turn over records to Congress, it was because those specific records were particularly sensitive. Those assertions need not be in conflict with the principle that, in general, Congress is entitled to these records. And the history of negotiation between the two branches likewise need not imply that Congress could not force the issue if it wanted to. Rather, it looks an awful lot like what lawyers call "bargaining in the shadow of the law." Both sides acknowledged the other side's valid claims, and negotiated in good faith on that basis. There was, therefore, no occasion for Congress to get the heavy artillery out, until Nixon came along. Nothing about this suggests the absence of a coercive power on Congress's part, only that it forbore to use the big guns because the presidents weren't forcing them to.

In any event, although the opinion spends a lot of time on this history, and does seem to suggest that it amounts to a binding constitutional rule, it does not actually rest the whole weight of the matter on this point. Instead, after canvassing the historical practice, the opinion proceeds to examine a handful of possible powers of Congress that might justify the PRA. Of course it concludes that none of them can.

This is where the aggravation sets in in earnest. For the most part, OLC analyzes the PRA in terms of what it calls the "oversight" or "investigatory" power. This is the power that lets Congress compel testimony, or the production of documentary evidence, in a hearing – in other words, to issue a subpoena. (Of course you will note that the PRA is not a subpoena! Bear this in mind; it will be important later.) There were a bunch of disputes about the scope of Congress's subpoena power during the first Trump administration, resulting in a couple of Supreme Court decisions that, in my view, absolutely butchered the relevant constitutional principles.

OLC's opinion relies heavily on one of these cases, Trump v. Mazars (2020), which had held that Congress has only a limited power to issue subpoenas directed toward the president. The idea behind Mazars is that a congressional subpoena must be related to a "valid legislative purpose." In other words, Congress can only compel testimony or evidence insofar as that information would be helpful in crafting legislation. This power is a necessary byproduct of the legislative power itself, and, the Court said, it is the only investigatory power Congress possesses. It cannot just seek to expose wrongdoing for the sake of exposure. That is not a legislative function, but an executive one; Congress is not in the business of law enforcement.

Now, you might be saying, doesn't Congress also have the power to impeach the president? And doesn't that suggest that it might need to procure information related to presidential misconduct? Indeed, and technically Mazars acknowledges this. But it says that Congress can only issue subpoenas under its impeachment power once it has formally opened an impeachment inquiry. This is arrant nonsense. Of course the existence of the impeachment power means that Congress needs to be able to monitor for potential impeachable offenses, even before it already has reason to believe one has been committed.

More broadly, the existence of the impeachment mechanism confirms that one of the core functions of Congress is overseeing the conduct of the executive branch. This oversight might result in legislation, where executive misconduct shows that one of the departments created by Congress needs to be reformed. It might result in a formal impeachment. Or it might result in something like forcing a Cabinet secretary to resign in disgrace, presumably with some suggestion that, had they not done so, Congress would have forced their hand. It doesn't really matter. The point is that Congress has general oversight authority, and Mazars erred badly in holding otherwise.

One pretty conspicuous problem with Mazars is that all of the cases it relied upon concerned subpoenas of private parties. Indeed, you can really tell something is screwy because of the way the Court relied on Barenblatt v. United States (1959). That case concerned a subpoena from, I kid you not, the House Un-American Activities Committee. Yeah, Joseph McCarthy's Communist-hunting operation. And what's worse, the Court upheld the subpoena! The very case from which the Mazars Court derived the "legitimate legislative purpose" standard said that rooting out Communism counted.

Of course Barenblatt should have gone the other way. It really is not the legitimate function of Congress to expose wrongdoing by private individuals, merely for the sake of exposure. (To say nothing of the First Amendment issues.) But things are totally different when the misconduct in question is committed by a public official. Exposing that kind of misconduct absolutely is a natural function of a legislature. (Indeed, John Stuart Mill saw this kind of oversight as one of the main things a representative assembly is good for!)

All of this was infuriating back when Mazars was decided, and it was aggravating to read OLC relying on it here. At the same time, I can't exactly blame them for relying on a recent Supreme Court decision. And honestly, I think they're right, even as a matter of first principles, that the PRA cannot be understood as an exercise of Congress's investigatory powers – even including its power of oversight. the way the PRA works, presidential papers only become public some twelve years after a president leaves office. It is not really about accountability; by that time, it is too late to impeach, or anything like it. Rather, the point is to create a historical record. This is distinct from the oversight power; OLC is not wrong about that.

But that doesn't mean the law isn't constitutional. The powers of Congress, after all, are disjunctive: a law is valid if it falls within any one of them. There is no such thing, I daresay, as a law that could be seen as an exercise of every single enumerated power all at once, even confining ourselves to the eighteen powers listed in Article I, Section 8. This was a point of confusion when the Supreme Court decided NFIB v. Sebelius (2012): people would say that the Court held the ObamaCare individual mandate "unconstitutional under the Commerce Clause" but valid as a tax. This is nonsense! The Commerce Clause forbids nothing. It authorizes some things; in all other matters, it is silent.

Similarly here. The PRA is not a valid exercise of the power to coin money, either. It is immaterial. The question is whether any power can be found to support it. And the answer is "yes": the Necessary and Proper Clause. This is, of course, why I wanted to write about all of this, because I have just been talking about the Necessary and Proper Clause in my module on the question of implied powers. Those of you who have been following that module may recall that the Necessary and Proper Clause confirms the existence of these implied powers. Or, in other words, it confirms that Congress has a broad choice of the means by which it will carry its powers into execution.

But that is not all it says. Read the Clause again, with care:

The Congress shall have Power ... To make all Laws which shall be necessary and proper for carrying into Execution the foregoing Powers, and all other Powers vested by this Constitution in the Government of the United States, or in any Department or Officer thereof.

Ah. Right. There's a whole second part of the Clause. Congress can make laws necessary and proper for carrying out, not only its own powers, but those of any other branch of the federal government!

This part of the Necessary and Proper Clause is often overlooked. But it is really, really theoretically important! It means that Congress has a broad power to decide the way in which the other branches will carry out their powers. This is unavoidable. One way to understand what the Clause does, with regard to the other powers of Congress, is to imagine that it had been inserted after each other provision of Section 8. Thus, we can imagine the Taxing Clause to read:

The Congress shall have Power To lay and collect Taxes, Duties, Imposts and Excises, to pay the Debts and provide for the common Defence and general Welfare of the United States, and to make all Laws which shall be necessary and proper for carrying this power into Execution.

This makes it clear enough what is going on. It is making explicit what would have been implicit anyway: that Congress has a broad choice of the means by which the power to tax will be carried into effect.

Okay, now imagine doing the same thing to Article II.

The President shall be Commander in Chief of the Army and Navy of the United States, and of the Militia of the several States, when called into the actual Service of the United States. Congress shall have the power to make all laws necessary and proper for carrying this power into Execution.

That Congress does, in fact, have this power is quite clear! It has this power as to every power vested in any department or officer of the government. Is the president's Commander-in-Chief power one of these? Of course it is. Therefore Congress has the power to make all laws necessary and proper for carrying it into execution. And this means that Congress has a broad choice of the means by which the power will be carried into effect. Hell, we can run the same play with the Article II Vesting Clause itself:

The executive Power shall be vested in a President of the United States of America. Congress shall have the power to make all laws necessary and proper for carrying this power into Execution.

We aren't used to thinking about the provisions of Article II this way, but, well, that's wrong! The Necessary and Proper Clause really doesn't leave this to the imagination! And if it were written this way, if every provision in Article II immediately reminded us that Congress has the power to make the laws necessary for carrying it into execution, then I think it would be plain as day that this means Congress, not the president, gets to make the choices about how exactly this power will be carried out.

This is not a fringe theory. It is this provision that allows Congress to structure the executive departments. These were some of the first laws Congress passed, creating and organizing the Treasury Department, the State Department, what was then the Department of War. These organic laws have no other basis in the Constitution than this. Indeed, if the Necessary and Proper Clause did not so clearly give this power to Congress, we might well imagine that it would have been seen as part of the president's executive power to organize the executive branch. That has literally never been the understanding, and even the most ardent proponents of executive power do not dare suggest that the business of organizing the executive branch does not fall to Congress.

OLC discusses the Necessary and Proper Clause, but only very briefly. It gets literally two paragraphs, at the very end of this section of the opinion. (The remainder is spent on the question of severability.) The power is referred to, in those two paragraphs, as "Congress’s power to enact legislation to assist in the execution of the President’s powers." The key word here is "assist," and this is basically why they conclude the PRA is not a valid exercise of the power:

Congress may not superintend the Presidency on the theory that Congress has its coordinate branch’s interests more at heart than does the President himself. ... Far from aiding the President in executing his constitutional duties, the Act threatens to impede his performance ... A statute that burdens rather than facilitates the exercise of executive power is not “necessary and proper” to carry that power into execution.

Horseshit. Congress is under absolutely no obligation to have the president's interests at heart when it organizes the executive branch. This is the whole point of the Necessary and Proper Clause: that the choice of how to carry out these powers, yes even the president's powers, belongs to Congress.

OLC's premise is just the reverse: that the president gets to make these choices, and Congress must respect them. But this is just not consistent with the text, with how we know the Necessary and Proper Clause to operate. Recall from the module on implied powers that the question of expediency, of whether a particular measure is truly needful under all the circumstances, belongs to Congress. It must, because it is Congress that decides what laws to pass under this Clause.

The same is true when Congress acts to organize the executive branch. Why do we have a Department of Justice? Why, because Congress decided this was a useful means for carrying out the president's powers as chief law enforcement officer. Did the president think it was expedient to organize this portion of the executive power in this way? Doesn't matter. It's not up to them. The text is as clear on this point as we could reasonably ask it to be.

Of course this decides the question of the Records Act. Congress has no less power to require the president to take notes than it has to create the State Department. OLC notes that Congress has a kind of plenary power over the agencies it creates, and that this does not apply to the presidency, which is true enough, so far as it goes. But it neglects that the power to create executive departments is itself a kind of power over the president! Yes, the president's office comes directly from the Constitution. And this does impose certain limits on how Congress can organize the government. It cannot divest the president of the executive power, and vest that power instead in, say, a prime minister, chosen by Congress itself and responsive to it. But it can say an awful lot about how the president will go about exercising that power. The Constitution gives it this power, in no uncertain terms.

And, you know, it could not really have done otherwise, because Congress is the one who funds the entire government. OLC touches upon this, on whether something like the PRA could be imposed as a condition on funding. Which isn't exactly relevant, because it isn't structured that way. But still. What OLC has to say on the matter is truly shocking. They seem to suggest that Congress literally does not have the power to withhold funding from the other branches of government as a way of compelling those branches to behave a certain way. It does have this power as to private individuals – the receipt of federal funds can and does have conditions attached to it – and even for the states, see South Dakota v. Dole (1987). But not, apparently, as to the president.

I honestly can't even understand what kind of rule OLC has in mind here, because it is just... madness. Congress is under no obligation to appropriate any funds at all! The Appropriations Clause could not be clearer about this: no money is to be drawn from the Treasury unless Congress says so. Again, that means Congress gets to decide what monies should be drawn from the Treasury. The one writing the laws gets to choose what those laws say.

OLC raises the horrid thought that Congress could use this same power of the purse to pass a Judicial Records Act similarly asserting public ownership over the papers of Supreme Court Justices. Of course it could do this! Probably through the Necessary and Proper Clause power to organize the judiciary, but certainly as a funding condition. About a century ago, Congress appropriated funds for the construction of what we know today as the Supreme Court building. Without a doubt, it could say that any Justice who refused to allow public access to their papers could not make use of that building. There is simply no theory on which it could not impose such a requirement.

OLC sees this as negating the independence of the other branches from Congress. I see it, instead, as simply confirming what the Necessary and Proper Clause says in so many words. Yes, those other branches are independent of Congress. Their powers are held by individuals who do not owe their titles to Congress, and Congress can't change that. But in another sense, they are absolutely "dependent" on Congress. They depend on it for their funds, if nothing else! But also they depend on Congress, more broadly, to provide – and therefore to choose – the means by which their powers will be carried into effect. The whole scheme coheres.

At least, it did until OLC got its hands on it.