Across the history of Supreme Court opinions, some cases are known more for a footnote than for a holding (United States v. Carolene Products (1938)). If the nation is lucky, Jack Smith will follow the cue in footnote 2 of Justice Amy Coney Barrett’s concurrence in Trump v. United States, allowing us to forget the mess that is the rest of that opinion.
American constitutional law has long struggled with the scope of implied presidential immunity. Richard Nixon startled David Frost in his post-presidency interview when Nixon asserted that “when the president does it that means that it is not illegal.” But in the narrow context he was describing, Nixon was right — not because of any magical monarchical power held by the President, but because of limits on the power of Congress to regulate the President’s constitutional authority. If Congress can’t regulate an authority, Congress can’t make it “criminal.” This is the core of presidential immunity: the affirmative substantive limits on Congress’s power to regulate the President.
But immunity is not just about substance. It’s also about buffers. The core of immunity may be the constitutional limits on Congress’s power. The practice of immunity, however, extends beyond those limits. Immunity reaches beyond its constitutional core, to give Presidents the breathing room they need to avoid being forced to litigate the constitutionality of every regulation as it might affect others while performing their duties as President. How much breathing room is not determined by the Constitution; it is a judgment, a balance, chosen by the Court.
In the civil context, the Supreme Court has maximized that breathing room, granting the President absolute immunity against monetary damages for acts that extend to the “‘outer perimeter’ of his official responsibility.” Nixon v. Fitzgerald, 457 U.S. 731 (1982).
In the criminal context, however, the Court has apparently now granted the President something less: The President has just a presumptive immunity for acts within the “outer perimeter of his official responsibility.” In any particular case, the prosecutor must demonstrate why that presumption should be overcome. Trump v. United States (2024). (Even this presumptive immunity is just presumptive for now: The Court expressly says it’s not sure whether it is presumptive or absolute. We’ll just have to wait and see…)
Justice Barrett’s footnote two, however, suggests why answering that question — what is the scope of presidential immunity in the criminal context ?— may not have been necessary. Because even assuming the broadest conception of immunity — reaching to the “outer perimeter of his official responsibility,” as the Court in Fitzgerald put it — the trigger is still some “official responsibility.” Or, under Justice Story’s framing of the question, the trigger is his “discharg[ing] the duties of his office.” If he is not acting on his “official responsibility” or discharging the “duties of his office,” his actions are private and not immune.
This point matters because, as Justice Barrett observed, there is no official presidential responsibility over the selection of the next president. Getting reelected is nowhere within the “duties of his office.” As Barrett wrote:
“The Constitution vests power to appoint Presidential electors in the States. … And while Congress has a limited role in that process, the President has none. In short, a President has no legal authority — and thus no official capacity — to influence how the States appoint their electors.”
Trump’s lawyers had suggested the contrary, based on a badly mischaracterized historical precedent. President Grant sent troops to Louisiana and Mississippi in 1876 to deal with the violence in those elections; Trump’s lawyers claimed that demonstrated that “organizing alternate slates of electors falls within the outer perimeter of Presidential duty.” But Grant was not acting on the basis of any power given to him by the Constitution; Grant was executing the Enforcement Act of 1871, which authorized him to use troops to keep the peace in federal elections. No similar statute authorized Trump to oversee the appointment of electors in 2020. Nor does any other federal election law give the President that power today.
No doubt, federal law does police voting in a presidential election, incident to a federal interest in the selection of electors. It is, therefore, completely proper for the Department of Justice to investigate claims of voting fraud or voter suppression in a particular election. But a demand that a secretary of state “find 11,780 votes” is not a demand that fraud be addressed. Nor is the threatening of a federal enforcement action unless a different slate of electors is chosen a good faith application of any federal voting rights statute. Both, instead, are part of a plan to press for the President’s reelection — again, a plan to advance a private purpose, not an “official responsibility.”
Ordinarily, it makes little sense to try to carve with care the scope of the president’s power to “take Care that the Laws be faithfully executed.” Within a federal domain, there are so many laws that it’s hard to see any limit to the President’s authority.
But in this context, there is a constitutional reason to read the scope of the President’s power carefully: Because the Constitution vests in the states the power to appoint electors — as Justice Barrett insisted. They have primary jurisdiction over the process of appointing their electors. No doubt the President has the power to keep the peace. But federalism demands the Court not translate that incidental power into support for the claim that “organizing alternate slates of electors falls within the outer perimeter of Presidential duty.” His duty has nothing to do with “organizing alternate slates.” If federalism means anything, it means that his duty is to keep the peace and prosecute fraud and discrimination (which, of course, will happen long after electoral votes are counted); it does not mean he gets to insist on a remedy — substitute his slate for my slate—that no court could ever uphold.
On this view of the case, the case was pretty simple. Everything Trump did here he did to ensure different elector slates would be counted on January 6. That purpose was private, not official. No immunity should attach to anything he did to advance that purpose — at least where what he did conflicts with the equally important constitutional value of preserving the power to appoint electors in the states.
The Court seemed to get this distinction. But it then muddied it right away. Section II of the opinion distinguishes between “criminal prosecution for unofficial acts committed while in office” and criminal prosecution for “official acts.” It then confuses the point by suggesting that acts that would be official if done in pursuit of an official duty must remain official acts, even if divorced from any official duty.
This, I suggest, is the critical error in the Court’s analysis. And it follows from the Court’s failure to distinguish precisely between immunity’s core and immunity’s buffers.
If Trump had tortured a state governor to coerce him to certify the Trump slate of electors, that act would have been a private act because selecting a slate of electors has nothing to do with the President’s “official responsibility.” Private act, no immunity.
Alternatively, if Trump had ordered his Attorney General to direct the FBI to torture the same governor and then fired the Attorney General when he disobeyed, the Court’s opinion holds that that act (the firing) is absolutely immune (since the removal of such officers is a core presidential power), and that a prosecutor cannot introduce any evidence about that act as a way to prove the President’s (hypothetical) conspiracy to torture a governor.
This is imprecise thinking. Yes, Congress could not make firing the AG a crime. The President has an unregulable power (subject to the Constitution) to remove core executive officers. The AG is one.
Yet even if Congress could not block his ability to remove the AG, there’s no reason s/he should be privileged to hide evidence of an illegal act merely because the evidence involves firing the AG. Forcing the AG or the President to give evidence about the firing does not invade the President’s constitutional authority — so long as the action for which the evidence would be admitted is grounded in a private, non-official purpose.
The point is more general. Acts are only understandable within the context of their purpose. Purpose colors the character of the act. If I kick a dog, it is critical to understand whether I had any purpose to inflict harm, or move the dog, or punish (however stupidly or cruelly) the dog. If I had no such purpose, my “kick” is not a kick. It is an accident.
This is not a hard distinction to make in many contexts of presidential activity. If the President demands to fly to Texas to stand with citizens who have just suffered (yet another) school massacre, there’s no doubt that act is appropriately presidential and no doubt that the government will bear the (not insubstantial) cost of that trip. Likewise, if the President demands to fly to Michigan to attend a political rally, there’s no doubt that that act is unofficial, and therefore, the campaign will have to pay a (seriously subsidized) cost of that trip. And likewise, there’s no doubt that sometimes purposes are mixed, and in those cases, the law requires a campaign to pay part of the cost of the trip, and the government pays the rest. In each case, the purpose determines how that same act — flying on AirForce One —will be treated: official, non-official, or mixed.
There’s no reason the same rule couldn’t apply to acts that involve even core presidential functions. Not every act by an official is an official act — because of this difference in purpose. And because the immunity that the Court is now flirting with extending to the President goes beyond constitutional limits, there is no constitutional obligation to recognize such immunity regardless of the act’s purpose.
Yet this sensible capacity to evaluate actions in light of their purpose is rejected by Chief Justice Roberts. If the actions could relate to an appropriate federal purpose, or more precisely, to an “official responsibility,” then they are immune, even if it is completely obvious that they are actually unrelated to any appropriate federal purpose. “As my Attorney General, I want you to tell Brad Raffensperger to find me 11,780 votes or we’ll sue” is immune because another President might say, “I want you, the Attorney General, to tell the secretary of state that he must respect the Voting Rights Act and allow his citizens to vote or we’ll sue.”
And, so why? Why must this absolute presumption apply in every case?
Because, as the Chief Justice warns us, the opposite rule would render the President “unduly cautious in the discharge of his official duties.” If there was a chance to ask — or even for a Court to determine—that otherwise plainly official conduct was actually devoted to an unofficial end (“Tell the FBI to torture Brad Raffensperger until he comes up with those 11,780 votes”), the mighty presidency of the most powerful nation in the world might fall.
This is the sort of casual reasoning that no other profession gets to engage in. There’s a claim about the consequence of a certain rule, yet there’s no need to demonstrate that that claim is, in fact, true. Yes, we need to ensure that the President is not “unduly cautious in the discharge of his official duties.” But we also need to ensure that the President does not commit crimes. What is the evidence that a rule that tries to assure both would produce undue caution?
Consider an example to make the point. The President has the power to order the targeted killing of terrorist suspects, at least when the suspect presents an imminent threat. If the President orders a killing pursuant to that authority, that is a core executive function, absolutely immune from prosecution (under the theory I’m advancing and the Court’s).
By contrast, we should want to live in a world where the President has no power to order the killing of his political opponent. Under my rule, that purpose—killing a political opponent—would render the otherwise core presidential action not immune. But as I read the Supreme Court’s opinion, that act would be absolutely immune. That makes no sense.
I don’t mean to suggest the obviously naive view that it will always be easy to distinguish between official and unofficial purposes. Nor do I believe the distinction is always clean. Maybe the President has evidence that the opponent is a Manchurian candidate, of an Al Qaeda kind. If that’s true, then the President’s motives would be mixed. I’d be happy to concede that if the motives were genuinely mixed, immunity makes sense. But even then, I don’t see why a prosecutor shouldn’t at least be able to force the President to show a Court why it made sense to kill a political opponent.
Or consider a more timely example. The former President “retruthed” (yes, that is what they call it) the following meme:
Ex parte Milligan would protect Cheney, no doubt. Under the Court’s opinion, however, Trump could not be held liable for initiating such a monstrosity. I can’t see how at least allowing a prosecutor to demonstrate that the purpose of the act was unofficial would undermine the stability of the executive.
Smith may have triggered this weird confusion by insisting, as the Court described it, “that a former President does not enjoy immunity from criminal prosecution for any actions, regardless of how they are characterized.” But again, this ignores the difference between core and buffers, and again, Nixon had explained why that statement is just false: Some acts, when done by the President, cannot be called “criminal” because Congress has no power to regulate the President in the performance of those acts. So yes, a President does enjoy immunity from prosecution for criminal acts that the Constitution does not permit Congress to call criminal.
Yet, that debate should have had nothing to do with this case. The statutes that Smith prosecutes Trump under do not purport to regulate the President’s constitutional powers or any “official responsibility.” Instead, they regulate acts with a purpose beyond any “outer perimeter of his official responsibilities.” That unofficial purpose should mean that the acts taken to advance those purposes, even if by officials, are not official acts. On this understanding, all of Trump’s acts advancing his purpose to affect the slate of electors counted by Congress would be private and regulable acts. And those helping him to perform those acts would be subject to law.
Barrett’s precision, however, may well have helped narrow the burden that Smith will now face. The Court divided the indictment into four categories of activities: (1) those relating to his interaction with the Department of Justice (p19–21); (2) those relating to his interaction with the Vice President (p21–24); (3) those relating to “Trump’s interactions with persons outside the Executive Branch: state officials, private parties, and the general public” (p24–28); and (4) “Trump’s conduct in connection with the events of January 6 itself” (p28–30).
In the last two categories, the question for the district court is whether the activities are part of the “duties of the office,” as Justice Story put it, or whether they qualify as “official or unofficial.” If they are unofficial — and on my read, they clearly are — then they will admit of no immunity at all.
With the second category (Trump’s communications with the Vice President), here again, Barrett’s precision helps. The Court charged the district court with weighing whether immunity in this context is necessary to avoid interfering with the appropriate role between the President and Vice President (whether denying immunity “would pose any dangers of intrusion on the authority and functions of the Executive Branch”). The Court spoke of the President’s need to work with the Vice President “to advance the President’s agenda in Congress.” But the President has no “official responsibility” to advance an agenda of his being reelected. That agenda is private. To allow criminal law to police his interaction with the Vice President while the Vice President is overseeing the counting of the states’ electoral votes is perfectly consistent with the Constitution’s design—at least if you keep the federalism component of that design in plain view. Again, as Barrett emphasized, the framers kept the President out of that process for good reason. His presence has nothing to do with his “official responsibilities” or the “duties of his office.”
The only category the Court has conclusively found absolutely immunity for is the first, involving the Justice Department and its deliberations about how and whether to enforce federal election law. Even here, however, the justification for immunity is incredibly weak: Federal law can certainly punish state officials for (an increasingly small range of) discrimination (increasingly small because the Court has shrunk the VRA); federal law can even reach private individuals who engage in voting fraud. But federal law does not make the President the Election Commissioner in Chief. More importantly, the President cannot constitutionally leverage federal election law to control which slate of electors a state appoints. It certainly cannot threaten federal prosecution unless the state flips the slate of electors that it says was appointed on Election Day. A Court keen to protect the autonomy of the states in the performance of the duty the framers placed on them — to appoint electors — would be much more careful before it immunized the improper application of federal law to displace state autonomy. But the Court has spoken, and this category of activity alone is now off the table for this prosecution.
Yet here’s the point: more than enough to move forward remains. If Smith helps the district court understand the framers’ purpose in insulating the appointment of electors from federal power, it will be easier to see why Trump’s activities in the last two categories are private, not official. He was free, of course, to talk to any election official, as any of us are. But it was not within the “duties of his office” to demand that the Georgia Secretary of State “find 11,780 votes.”
And if the district court takes note of the extraordinary threat of violence that the Vice President actually faced on January 6, it could well be persuaded that there is no reason to immunize this tiny slice of interaction between the President and his Vice President from criminal review. There are 1,461 days in any presidential term. It’s no great loss if the President and Vice President can coordinate on just 1,460 of those, and leaving that one day to the Vice President alone.
Thus, if Smith succeeds in these three categories — establishing that two are private, not public, and overcoming the presumptive immunity that would cover communication with his Vice President — all that would be excluded is evidence from the President and the Justice Department about what the President said. But there is plenty of independent evidence about the purpose of the President’s engagement with electors and state officials. That evidence should be enough to help any jury understand exactly what was at stake here.
Trump v. United States should have had nothing to do with the hard questions of presidential immunity law that Nixon’s theory raises. It should have had everything to do with the easy answers that Nixon’s theory yields. If the clear purposes of the President’s acts are unrelated to any “official responsibility” of the President, acts to advance those purposes should not be immune. Neither the Constitution nor federal law obliges the President to “take care” that the states do their duty in appointing electors. Neither does it involve him in Congress’s counting of electoral votes. His acts as a candidate are beyond the outer perimeter of his official responsibility. He should, therefore, be entitled to no immunity for those acts.
