Courts & Litigation Criminal Justice & the Rule of Law Executive Branch

The Relative Insignificance of the Immunity Holding in Trump v. United States (and What Is Really Important in the Decision)

Jack Goldsmith
Monday, September 23, 2024, 12:52 PM
A Constitution Day speech at the University of Michigan Law School
Former President Donald Trump gestures with a fist pump in October 2018. (Photo by Joyce N. Boghosian, White House, Public Domain)

Published by The Lawfare Institute
in Cooperation With
Brookings

Editor's Note: This is a slightly revised version of a speech delivered on Constitution Day, Sept. 17, at the University of Michigan Law School.

Today I focus on Article II of the Constitution, and in particular on how the interpretation of Article II in Trump v. United States will impact executive branch adherence to law. That decision, as you know, concerned the prosecution of former President Donald Trump for allegedly conspiring to overturn the 2020 presidential election by spreading false claims of election fraud in order to obstruct the certification of the election results. The Court held in that context that a former president is absolutely immune from post-presidency prosecution for presidential actions that fall in his exclusive constitutional authority, and at least presumptively immune for other official presidential acts.

Many critics claim that Trump v. United States opens the floodgates to a “lawless presidency.” Justice Sonia Sotomayor in dissent, for example, said that the majority “effectively creates a law-free zone around the President, upsetting the status quo that has existed since the Founding.” Justice Ketanji Brown Jackson similarly argued that the majority opinion alters the “individual accountability model” that had previously guided presidents and thus “undermines the constraints of the law as a deterrent for future Presidents who might otherwise abuse their power.”

As Adrian Vermeule has suggested, these critics assume what we might call, following Holmes, a bad-man president of the United States—a president, that is, who is oblivious to the norms and other non-legal expectations of the office, and who follows narrow self-interest right up to the point that the effective sanction of law allows it. With the threatened sanction of criminal law gone, the bad-man president is cut loose to do all sorts of awful illegal things that serve the president’s narrow private interests.   

I think arguments of this form rest on a misleading picture of how the criminal law operates on the presidency and that this misleading picture has led some to misjudge how Trump v. United States will impact the presidency. I will make two claims. First, the immunity holding that has been the focus of most attention will not be nearly as consequential for the presidency as the critics claim. Second, the decision’s main significance for the presidency lies in its expansive discussion of exclusive presidential powers, independent of the immunity ruling.

***

To unpack these claims, I begin with Article II of the Constitution. That article vests the president with the “executive Power”; it gives the president a duty to “take Care that the Laws be faithfully executed”; and it requires the president to take an oath to “faithfully execute the Office of President” and to “preserve, protect and defend the Constitution.”

These clauses in various ways require the president to comply with the law. But they also empower the president to enforce the law, which includes a great deal of discretion about whether and how to do so. They empower the president to supervise the running of the executive branch and to direct government policy (including firing certain subordinate executive branch officials). And they empower the president to interpret the law for the executive branch.

Starting with the George Washington administration, presidents have wielded this latter power to interpret the law for the executive branch in self-serving ways, just as courts often interpret the law in self-serving ways. This tendency is consistent with a Madisonian model of interbranch contestation, though sometimes, as we will see, such self-serving interpretations can cross over into abuse.

***

I now turn to explain why I believe the immunity ruling in Trump v. United States will not have a great impact on executive branch compliance with law. Because time is short, I will assume knowledge of the details of the decision.

I begin with the underappreciated truth that for almost every act of presidential lawbreaking, the president must rely on executive branch subordinates who are not immune from criminal liability even if the president is immune. This principle of subordinate criminal liability, combined with executive branch norms of right behavior, are the main determinants of executive branch compliance with criminal law.

Consider Justice Sotomayor’s claim that a president would be immune from criminal liability if he or she ordered the Navy’s Seal Team 6 to assassinate a political rival. This claim does not follow from the Court’s opinion. But assume that it does. Even if the president is immune from prosecution for ordering the assassination, the soldiers could be prosecuted by the next president. This potential criminal liability for subordinates makes it hard for a lawless president to carry out lawless acts.

I know from my time as the head of the Office of Legal Counsel two decades ago that the threat of criminal sanction for subordinates was a very powerful check on executive branch officials. After the Justice Department in May 2004 declared that approximately half of the Stellar Wind warrantless wiretapping and bulk metadata collection program had to cease because it was inconsistent with criminal laws, the president disagreed and, on the advice of the White House counsel, declared the entire program lawful. Nonetheless, very senior executive branch officials were about to resign rather than execute the full program, and at that point the White House backed down. They were about to resign for many reasons, but a main one, as many made clear to me at the time, was fear of criminal law investigation and possibly conviction.        

More recent proof comes from Volume 2 of the Mueller report, which documented 10 instances of then-President Trump seeking to obstruct the ongoing FBI investigation into Russia’s interference in the 2016 presidential election. For reasons I will get to in a moment, it was unclear under executive branch interpretive principles whether Trump as president was capable in these instances of violating the criminal obstruction statute. What is remarkable about the episode is that many of Trump’s closest subordinates refused to carry out his commands to effect the obstruction.

These officials acted in part, probably large part, out of fear of violating the criminal obstruction statute even though Trump himself was oblivious to it. The Mueller report included a White House Counsel’s Office note from March 2, 2017. The note was written soon after Attorney General Jeff Sessions had recused himself from the Russia investigation in the face of pressure not to recuse from President Trump through White House Counsel Don McGahn. The note stated: “No contact w/Sessions” and “No comms / Serious concerns about obstruction.” And Michael Schmidt’s book, “Donald Trump v. The United States,” shows that this obstruction concern permeated the White House Counsel’s Office during this period.

Let me acknowledge that some have read Trump v. United States to weaken the principle of subordinate liability in the Court’s first immunity category—when the president acts in accord with a conclusive and preclusive presidential power. Jeffrey Clark, a senior Justice Department official who helped Trump in his election fraud scheme, has made just such an argument. The Supreme Court ruled that President Trump was absolutely immune from prosecution for discussions with Justice Department officials about a Justice Department letter to state legislatures that would have fraudulently conveyed concerns about the election and instructed the states to craft an alternate slate of electors. Clark was involved in these plans. In his disbarment proceeding, he claims the same absolute immunity as the president.

I think Clark is wrong and that the Supreme Court will reject this argument. The Court’s decision in Trump v. United States turns on the unique position of the president in the constitutional scheme; the Court nowhere hints that it is extending immunity to subordinates; and the Court in its 1982 decision in Harlow v. Fitzgerald rejected the argument in the civil context that absolute presidential immunity entailed absolute immunity for subordinates. But if (as is possible) Clark is right and I am wrong, then Trump v. United States would weaken the principle of subordinate official liability to an uncertain degree.

***

As it stands today, as it has stood for a long while, and as I believe will persist, the principle of subordinate criminal liability is a powerful check on a lawless president. And it applies to almost all, though perhaps not all, of the hoariest hypotheticals advanced after Trump v. United States. Certainly the president cannot effectuate a coup or assassinate political rivals without assistance of people who are fully subject to criminal law. This is the first reason the immunity decision is not as momentous as it might seem.

But there is a deeper reason for this conclusion, a reason that is a sharp qualification to the significance of subordinate liability. It is this: A bad-man president has long had the ability to skirt both subordinate criminal liability and presidential criminal liability in a broad array of cases, wholly independent of the Court’s immunity ruling.

Recall the president’s power of legal interpretation. The president delegates that power to the attorney general, who in turn delegates it to the Office of Legal Counsel, where I used to work. And when the attorney general or the Office of Legal Counsel writes the president or another executive branch official a legal opinion concluding that a course of action is lawful under a criminal statute, that president or executive branch official will not be prosecuted by a subsequent administration.

When I was in the government, the CIA general counsel, John Rizzo, referred to this immunity-conferring power of the Justice Department as a “golden shield.” Rizzo was referring specifically to the effect of the notorious Office of Legal Counsel opinions ruling that the criminal prohibition on torture did not forbid the CIA’s brutal post-9-11 interrogation techniques. Those opinions were sought by the White House in a distorted process that aimed to greenlight the CIA program. They were badly legally flawed on multiple levels. As a very senior Justice Department lawyer told me in government, the main opinion on torture “reads like a bad defense counsel’s brief, not an OLC opinion.”

And yet despite these deep flaws in process and substance, the golden shield worked. President Obama’s attorney general, Eric Holder, believed that the Office of Legal Counsel opinions were appallingly bad and that elements of the CIA program were illegal. But when he ordered an investigation of the Bush-era CIA program, he made clear that he would, in his words, “not prosecute anyone who acted in good faith and within the scope of the legal guidance given by the Office of Legal Counsel regarding the interrogation of detainees.” There are many reasons why he reached this conclusion. But one reason was a legal doctrine called entrapment by estoppel. Under this doctrine, it is a violation of due process to prosecute anyone who relies on an authorized government representative’s claim that an action is lawful.

Entrapment by estoppel is a powerful doctrine inside the executive branch that undergirds the significance of legal advice related to criminal statutes. (There are others as well.) Don’t take my word for it. At oral argument in Trump v. United States, the special counsel’s lawyer, Michael Dreeben, held up entrapment by estoppel as an important protection for the president entirely independent of other immunity doctrines. He told the justices that “it would be a due process problem to prosecute a president who received advice from the attorney general that his actions were lawful absent the kind of collusion or conspiracy that itself represented a criminal violation, which I don't really see as being a realistic option.” Dreeben mentions the limit of entrapment by estoppel, but it is a distant limit. The permissive OLC torture opinions, to take one notorious example, did not rise to the level of a criminal conspiracy and had full legal effect as a golden shield.

It is easy to imagine a prescient bad-man president getting a golden-shield legal opinion from his attorney general or Office of Legal Counsel prior to January 6 to the effect that none of the statutes that former President Trump is charged with violating in the January 6 prosecution apply to the president’s official acts. Such an opinion would rely on a different executive branch interpretive legal principle that was also discussed in the special counsel’s brief. This is the principle that generally worded criminal statutes that “arguably limit the President’s constitutional role” or involve a “possible conflict with the President’s constitutional prerogatives” do not apply to the president absent a plain statement.

Again, we do not have to speculate very much about this. Trump’s attorney general at the end of his presidency, until mid-December 2020, was William Barr. Barr in 2018 wrote a famous memo that argued that the plain statement rule precluded Robert Mueller from applying the main obstruction of justice statute to the allegedly obstructing Trump acts he was investigating in the Russian collusion matter, at least as Barr understood those acts at the time. This is the very same statute that supports two of the four charges against Trump in the January 6 prosecution.

People can and do disagree about how the Office of Legal Counsel plain statement rule best applies to the statutes Trump is charged with violating. The special counsel acknowledged the OLC principle in its brief and read it narrowly. The Court cited several of the relevant OLC opinions as evidence that the government “broadly agrees that the President’s official acts are entitled to some degree of constitutional protection.” Justice Amy Coney Barrett said without explanation that she did not believe the OLC canon “applies in this circumstance.” As noted above, Barr took a robust view of the canon as applied to an obstruction statute. (I have previously laid out my views of the canon in the context of the Mueller investigation.)

While the proper applicability of the canon is contested, what is not contestable, I think, is that a bad-man president’s attorney general could in good faith have concluded that the plain statement rule rendered inapplicable all of the criminal statutes in Jack Smith’s indictment that “possibly” conflicted with Trump’s constitutional prerogatives or “arguably” limited his constitutional role—an interpretation that would likely rule out all charges involving the president’s official acts.

A second very robust presidential power that a bad-man president could use to skirt the principle of subordinate liability is the pardon power. This one is easier to see, so I will be brief. As the Court noted in Trump v. United States, the pardon power is broad. A bad-man president could pardon subordinates who, for example, carried out an assassination or a coup, just as Donald Trump in the two weeks after January 6 could have pardoned everyone involved who was later indicted, thereby precluding the indictments. Also, recall that Donald Trump pardoned cronies who were investigated and convicted by Robert Mueller. Finally, no one knows if a president can pardon himself, but many people believe that he or she can.

***

The point of this analysis of a bad-man president’s robust law-skirting tools is to highlight the relative insignificance of the presidential immunity ruling in Trump v. United States. I have tried to show that executive branch compliance with law is mainly secured by norms and subordinate official liability; that a bad-man president had plenty of ways to weaken the effect of norms and subordinate liability prior to Trump v. United States; and thus that the marginal impact of the immunity decision is unlikely to be significant. This is a very different and I believe more realistic picture of executive branch compliance with criminal law than one that sees post-presidency enforcement of criminal law against the president as a significant check on the president and against executive branch illegality.

There are many other reasons to discount the significance of Trump v. United States on presidential illegality. For one thing, the Court decided very little—basically only that the president was immune in connection with the charge related to Justice Department communications. Jack Smith’s superseding indictment of Trump reads Trump v. United States to permit him to pursue the rest of his original prosecution. Smith is not obviously wrong. If Smith reaches the merits of the charges in his indictment, the dissents will be shown to be wrong even on their own terms.

I am not claiming, let me be clear, that the immunity ruling in Trump v. United States will have zero impact on executive branch adherence to law. Under various assumptions the ruling could embolden a bad-man president to commit more illegal acts than otherwise or to push subordinates harder or more successfully to do so. I have simply tried to suggest that an accurate picture of how criminal law operates on the presidency suggests that the marginal impact of the immunity holding on the presidency is not large.

***

I have discussed the bad-man president because I think it is an implicit assumption in the dissents in Trump v. United States and because I think it is a useful heuristic to explain the redoubtable power that the executive branch in theory has to engage in lawless behavior independent of and prior to the Court’s immunity ruling.

To assess some of the more realistic yet underappreciated impacts of Trump v. United States on executive branch compliance with law, I think it is useful to leave our bad-man president behind and use the heuristic of what I will call the “normal” president. A normal president has internalized the importance of complying with law and norms, and with being perceived to have done so. He or she has an ingrained sense of respect for coordinate institutions. And he or she has a sense of shame.

Normal presidents differ in their attitudes toward the scope of presidential power under Article II and on how they talk about that power. They aren’t going to assassinate political rivals or direct a coup. But they feel compulsion to accomplish vital agenda items and to keep the country safe, and they sometimes rely on aggressive interpretations of their statutory and constitutional powers to achieve these ends. Sometimes, a normal president will push an aggressive legal interpretation so far as to commit what is widely perceived as an illegal act. Almost every president, starting with George Washington up through President Biden, has been a version of a normal president in these broad senses.

I think Trump v. United States will likely have a significant impact on the stylized normal president trying to wield executive power to pursue his or her agenda. The impact will not come from the presidential immunity element of the decision, but rather from the Court’s unusually expansive discussion—much of it dicta—of the president’s exclusive presidential powers independent of their connection to presidential immunity.

Executive branch lawyers in normal presidencies do not interpret the law the same way that courts do. They follow Supreme Court precedents when they are on point, but they also give great weight to past executive branch practice and legal opinions that are often more salient. Especially when there is no controlling judicial precedent, executive branch lawyers “tend[] to rely on general principles embodied in Supreme Court dicta, especially ones that favor presidential power.” On the whole, executive branch lawyers tend to see executive power in broader terms than do courts.

Normal presidents and their executive branch lawyers will use the remarkably pro-executive statements in Trump v. United States to their advantage in the course of taking care to faithfully execute the law. As I mentioned at the outset, this duty entails the power to interpret the law, including a Marbury-like power to interpret federal statutes for their consistency with the Constitution, and especially with Article II. Although Trump v. United States was technically a loss for the government, I am confident that the decision was met with glee in some corners of the Solicitor General’s Office and the Office of Legal Counsel—offices that are typically in the business of defending executive branch power or of explaining it in congenial ways.

Several parts of the decision, I think, will be especially salient for these ends. I will mention just a few.

As several people have pointed out, it is hard to see how Morrison v. Olson—the already-on-the-ropes 1988 decision upholding the constitutionality of the independent counsel statute—survives Trump v. United States. Morrison said Congress could “reduce[] the amount of control or supervision that the Attorney General and, through him, the President exercises over the investigation and prosecution of a certain class of alleged criminal activity.” Trump v. United States says quite clearly that Congress cannot do these things. The Trump decision also described in unprecedentedly broad terms the president’s exclusive power over “[i]nvestigative and prosecutorial decisionmaking,” and it arguably did the same for the president’s removal power.

Executive branch lawyers will use these elements of the decision to stave off efforts by Congress to direct any aspect of prosecution or law enforcement disfavored by the executive branch. They will also use them to reject congressional efforts to constrain the president’s directive power, and to resist statutes that purport to regulate the structure of the government in ways the executive branch dislikes, especially concerning the Justice Department. Some administrations will use the decision to try to remove a broader swath of subordinate officials than has been allowed in the past, including possibly independent agency heads and inferior officers. For example, a second Trump administration, if it comes to pass, will surely invoke these passages to argue that a Biden-era Office of Personnel Management rule that seeks to “Trump-proof” the civil service is unconstitutional. And executive branch lawyers will invoke the decision in support of enhanced executive branch power to exercise discretion in law enforcement, including civil law enforcement.

There are more examples, but the general point should be clear: Trump v. United States is a gift to executive branch lawyers who defend and seek to extend executive branch prerogatives. As Steve Vladeck said, Trump v. United States “is going to be the opinion that launches 1000 executive power ships.”

Vladeck added that the decision is “going to take courts in any number of contexts years, if not decades, to sort out just how much … or how little” the decision means (emphasis added). This statement is true but does not focus on where Trump v. United States will have its greatest impact. For executive branch interpretations that result in actions that are reviewable in court, judges will be unbeholden to the dicta in Trump v. United States and will have the final say in shaping the impact of the decision.

But many executive branch interpretations and applications of Trump v. United States will take place in contexts that are not subject to judicial review. One typical example will come in Office of Legal Counsel bill comments about the unconstitutionality of a new statutory restriction or command that get sifted into presidential signing statements about how the president will under-enforce or not enforce the restriction or command. These decisions often impact the scope of federal law in ways that are not subject to judicial review. My predecessor in the Office of Legal Counsel, Walter Dellinger, accurately described the OLC view that it has an “enhanced responsibility to resist unconstitutional provisions that encroach upon the constitutional powers of the Presidency,” especially in cases beyond judicial review. And that is what will happen.

***

I fear that for many, my talk will seem like a downer on Constitution Day, since it seems like the Constitution is not terribly efficacious in constraining executive power. Some might even see my talk as “anti-constitutional.”

That is not the way I see it. Article II is part of the Constitution and we should be realistic about what it does. It expressed a completely new conception of executive power, one that bound the president to law but at the same time gave the president extraordinary power and discretion to enforce the law and also to interpret the law as a prerequisite to enforcement.

This combination of powers has been central to a vigorous principled executive power that the Framers correctly believed was needed to make the government work. The same powers have always given a president the means to abuse the law, and some unscrupulous presidents have abused it.

At the end of the day, it is not the big-C Constitution, considered narrowly as standalone foundational text interpreted by courts, that has tempered such abuses. Instead, presidential respect for law has depended at bottom on two related things, both having to do with the American people.

First, it has depended on the character and commitments of the president elected by the American people—it has depended, that is, on the election of what I have described as a “normal” president (and on that president’s appointment of “normal” subordinates.) One implication of my talk today is that if the American people elect a bad-man president, there is not much that the big-C Constitution can do to prevent presidential or executive branch illegality.

Second, and relatedly, presidential respect for law has depended on the types of pressure that the American people place on members of Congress and other government actors to resist executive branch abuse. Yes, the courts play an important role in this equation, and their decisions matter. But I think judicial decisions are secondarily important compared to the beliefs, commitments, and actions of the American people through their political and other institutional representatives. “Laws are always unstable unless they are founded upon the manners of a nation,” Tocqueville once said. The laws and norms that govern the presidency work only if a wide swath of the population and their elected representatives believes in them, and their legitimacy, and wants them to work.

Unfortunately, we live in an era in which the customs of the nation are riven. The American people are deeply and increasingly distrustful of federal governmental institutions and perceive the reality and legitimacy of presidential action in starkly different and contradictory ways. The issue is in part a loss of public vigilance about norms, but it is also in part about wildly polarized perceptions of the actual facts bearing on those norms. It is not at all clear how well small-c constitutional constraints on the presidency can work effectively in this milieu.

Arthur Schlesinger Jr. expressed these points well in the closing lines of his magisterial book, “The Imperial Presidency.” The book was in the fall of 1973, “in the latter days, hectic and ominous, of the Presidency of Richard M. Nixon,” as Schlesinger later put it. Here is what Schlesinger said:

[W]hat kept a strong President constitutional, in addition to checks and balances incorporated within his own breast, was the vigilance of the nation. Neither impeachment nor repentance would make much difference if the people themselves had come to an unconscious acceptance of the imperial Presidency. The Constitution could not hold the nation to ideals it was determined to betray. The reinvigoration of the written checks in the American Constitution depended on the reinvigoration of the unwritten checks in American society. The great institutions—Congress, the courts, the executive establishment, the press, the universities, public opinion—had to reclaim their own dignity and meet their own responsibilities. As Madison said long ago, the country could not trust to “parchment barriers” to halt the encroaching spirit of power. In the end, the Constitution would live only if it embodied the spirit of the American people.

I think that is a poignant and appropriate place to stop. Thank you.


Jack Goldsmith is the Learned Hand Professor at Harvard Law School, co-founder of Lawfare, and a Non-Resident Senior Fellow at the American Enterprise Institute. Before coming to Harvard, Professor Goldsmith served as Assistant Attorney General, Office of Legal Counsel from 2003-2004, and Special Counsel to the Department of Defense from 2002-2003.

Subscribe to Lawfare