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

Correcting Presidential Immunity's Original Sin

Alan Z. Rozenshtein
Wednesday, April 10, 2024, 8:00 AM

In both civil and criminal cases, presidents should generally receive qualified, not absolute, immunity for official acts.

Published by The Lawfare Institute
in Cooperation With
Brookings

Later this month the Supreme Court will hear oral argument about whether former President Donald Trump can be prosecuted for his attempt to overturn the 2020 election. In Nixon v. Fitzgerald (1982), the Court held that presidents enjoy absolute immunity from civil litigation for acts within the “‘outer perimeter’ of [their] official responsibility” but explicitly reserved the issue of immunity from criminal liability. Thus, a natural way of framing the question in United States v. Trump is whether the Court should extend Fitzgerald to the criminal context. In denying Trump criminal immunity, the U.S. Court of Appeals for the D.C. Circuit declined to so extend Fitzgerald, arguing that, even if the alleged conduct fell within the scope of Trump’s official acts, there is a heightened public interest in criminal cases, especially in the context of upholding elections.

I agree with the D.C. Circuit that Trump should not be immune for at least most of the conduct alleged in his D.C. federal indictment. And I also agree that Fitzgerald should not apply to the criminal context. But that is not because, as the D.C. Circuit suggested, the considerations behind civil and criminal immunity are so different. The problem lies with Fitzgerald itself, particularly its use of absolute, rather than qualified, immunity. By applying the strongest possible degree of immunity to such a broad range of activity—the “outer perimeter” of official presidential actions—the Supreme Court shielded an excessive amount of presidential malfeasance from civil liability. But the Court also created a dilemma for future decisions on criminal immunity: either apply Fitzgerald’s extreme immunity to the criminal context and functionally put the president above the law or impose an arbitrary distinction—criminal versus civil liability—into the law of presidential immunity.

Fitzgerald is the law of presidential immunity’s original sin, albeit a fixable one. If presidents enjoyed only qualified immunity for official acts—at least for those acts falling outside a small core of powers that the Constitution vests solely in the president—not only would the doctrine of presidential civil immunity be more robust with respect to lawbreaking presidents, but it could also be applied straightforwardly to the criminal context in a way that would respect both what Fitzgerald characterized as the president’s “unique status under the Constitution” and then-judge Ketanji Brown Jackson’s admonition that “Presidents are not kings.”

I’ve divided my argument into several parts. First, I describe the conceptual origins of presidential immunity: doctrines of civil immunity that apply to public officials. Second, I address presidential civil immunity and the Fitzgerald doctrine and argue, using the ongoing civil lawsuit against Trump for his encouragement of the Jan. 6 Capitol riot as an example, that Trump’s losses in court belie a deep vulnerability in the civil immunity doctrine’s supposed safeguards, a vulnerability that a switch to qualified immunity would cure. Finally, I return to the question of criminal immunity, arguing that a qualified immunity standard would both lead to a sensible result in United States v. Trump and bring coherence to the law of presidential immunity in both the civil and criminal contexts.

Basic Immunity Principles

Presidential immunity is a specific case of a more general civil immunity that the law affords public officials. (Technically, immunity sometimes applies to private individuals who are temporarily or in a limited capacity discharging a function—but for simplicity I’ll refer to all enjoyers of immunity, whether formal public officials or not, simply as “officials.”) In the first instance, “immunity” might seem paradoxical, because, presumably, we only want officials doing what we want them to do, and the way we specify what we want them to do is through law. But if an official acts according to the law, they’ve done nothing wrong, and they don’t need additional “immunity.” 

The problem—and where immunity comes in—is that not all violations of the law are equal. An official, even one acting diligently and in good faith, will often not be sure whether a particular action is legal, because either the law or facts are in dispute. If officials are personally civilly liable whenever they act illegally, they will naturally be overcautious, and the public will suffer from official timidity and underenforcement. Immunity doctrine recognizes that, in an imperfect world, the ideal level of official illegality is nonzero, and, for every task we want our officials to perform, we need to add an illegality “fudge factor” for which we will grant them immunity.

The central task of immunity doctrine is to calibrate that fudge factor. Make it too small and you’ll discourage officials from doing their jobs. But make it too large and you’ll encourage lawlessness and deny injured parties relief. A substantial but not infinite fudge factor is known as “qualified immunity,” which generally applies to the discretionary and policy judgments of most officials. The standard for qualified immunity has evolved over American history. Initially at common law, immunity required the official to have acted in “good faith” (or, alternatively, without “malice”). 

The modern rule, established in Harlow v. Fitzgerald (1982), is that officials are civilly immune as long as their “conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.” One can question whether the “clearly established law” test has been calibrated at the right level—especially in the context of policing, there is widespread concern that the way courts apply the “clearly established law” test protects too much police misconduct—while still recognizing that some degree of immunity is unavoidable if we want officials to do their jobs.

But qualified immunity is not the only kind of immunity that’s available to government officials. For a smaller subset of official acts, courts have recognized “absolute” immunity, by which official acts are immune from liability even if the official knew that what they were doing was illegal. Absolute immunity applies to individuals carrying out legislative, judicial, and prosecutorial functions—and, as will be discussed below, to the president.

It’s important to recognize just how extreme absolute immunity is. For example, absolute immunity shielded a judge from allegations that he ordered police officers to forcibly and violently drag an attorney into his courtroom. It protected another judge who, without legal authority or process, ordered a teenage girl sterilized at the request of her mother. It applied to a prosecutor accused of using perjured testimony to convict a defendant of first-degree murder. And it protected both legislators and their aides for disclosing classified information on the floor of Congress.

For this reason, courts have been very careful to circumscribe the scope of absolute immunity, at least in the case of prosecutorial immunity (they have been notably more generous with immunity for their own, judicial actions). This has led to exceedingly “fine distinctions” that are difficult to make sense of. For example, in one case, the Supreme Court distinguished between the immunity afforded to a prosecutor for getting an arrest warrant (absolute) versus filling out the supporting declaration (qualified). In another case, the Court held that a prosecutor’s decision to use evidence would be protected by absolute immunity, whereas his going out to find evidence was protected only by qualified immunity. The Third Circuit has held that failing to turn over exculpatory evidence to the defense is protected by absolute immunity, whereas destroying evidence is protected only by qualified immunity. The court has also held that withholding funds for a state-funded position was legislative (and thus covered by absolute immunity), but that a state legislative caucus firing an employee was administrative (and thus covered by qualified immunity). The Seventh Circuit has held that police officers get absolute immunity when they testify before grand juries, but only qualified immunity when they fail to turn over exculpatory evidence.

The Danger of Fitzgerald’s Absolute Immunity Standard

This is the background against which to understand the major case regarding absolute immunity for the president, Nixon v. Fitzgerald. In Fitzgerald, an Air Force analyst accused President Richard Nixon of unlawfully firing him in retaliation for Fitzgerald’s politically embarrassing congressional testimony about government cost overruns. The Supreme Court rejected Fitzgerald’s civil lawsuit against Nixon, arguing that even qualified immunity would be insufficient “because of the singular importance of the President’s duties, diversion of his energies by concern with private lawsuits would raise unique risks to the effective functioning of government.” Thus, the Court recognized “absolute Presidential immunity from damages liability for acts within the ‘outer perimeter’ of his official responsibility.”

The question since Fitzgerald has been what precisely is the “outer perimeter of [the president’s] official responsibility.” Some cases are relatively easy. In Clinton v. Jones, the Court held that allegations of sexual misconduct predating the presidency were not subject to absolute immunity, since, by definition, actions taken before someone assumed the presidency could not be part of the president’s official duties. And it’s easy to imagine conduct taken by a sitting president that would also be unambiguously outside the scope of his official duties: for example, a physical or sexual assault, a breach of contract in a personal business deal, or, to take another Trump-era example, making falsely labeled reimbursement payments to a lawyer who, prior to the president’s election, paid hush money to an adult film star to not reveal a pre-presidency sexual encounter.

What about the events leading up to and on Jan. 6? Here, despite Fitzgerald’s expansive language regarding the “outer perimeter,” all the courts to have considered the issue have found that Trump’s actions in trying to overthrow the 2020 election fell outside his official actions. The most important such decision is the D.C. Circuit’s opinion in Blassingame v. Trump, upholding the lower court’s denial of immunity. The D.C. Circuit chiefly focused on the form of Trump’s alleged actions. Rather than portray himself as acting in his presidential capacity and using the characteristic tools of his office—official presidential addresses and the authority of the federal bureaucracy—Trump behaved as any candidate for office would, litigating in his personal capacity, speaking at campaign events like the “Stop the Steal” rally, and working through associates outside the federal government.

The problem with the D.C. Circuit’s approach is that it is easy to circumvent. Because the whole point of absolute immunity is to avoid “highly intrusive inquiries into the President’s motives,” the analysis of “whether the President is engaged in official functions or unofficial reelection campaign activity … does not turn on whether the activity was subjectively undertaken in some measure to enhance the President’s re-election prospects or profile” but, rather, on a “context-specific assessment of the nature of the function performed.” If an activity “is clothed in the trappings of an official function based on objective indicia, it more likely constitutes an official act for immunity purposes than if it bears the hallmarks of re-election campaign activity,” and thus “the same essential message or act may be either official or unofficial depending on the circumstances in which it is delivered or performed,” such as when a president, delivering a State of the Union address (an official act), “draw[s] themes and make[s] points with an eye on maintaining his public standing in an election year.”

In other words, for a president seeking to receive absolute immunity for his acts, the playbook is clear: Launder actions taken for personal gain through the many powers and privileges of the office, such as the federal bureaucracy, the military, and the bully pulpit. While other types of absolute immunity—for prosecutors, judges, and legislators—can also be pushed too far, the sheer breadth of the president’s powers and responsibilities and the frequent difficulty in separating his personal interests from those of the office make absolute presidential immunity particularly vulnerable to unwanted inflation.

This does not render entirely pointless Fitzgerald’s limitation of absolute immunity to official acts. It is better to incentivize even the malevolent president to act through official channels, because those channels come with their own constraints: the formal procedures and informal norms by which subordinate officials—the “deep state,” as Trump so frequently complained about—can resist unlawful action and bring increased transparency to presidential behavior.

But Trump was largely stymied in his attempts to overthrow the 2020 election not only because people recognized his malevolent intentions but also because Trump’s entire reign was one of “malevolence tempered by incompetence.” A more skillful president would have sought to work through, not around, the executive branch, in order to avail himself of absolute immunity. For example, as Marty Lederman has recently noted, Trump’s conspiracy with Department of Justice official Jeffrey Clark to encourage state legislators to choose alternate electors was “undoubtedly … conduct undertaken in Trump’s official capacity as President.” A smarter, more disciplined president would have modeled all his actions on the Trump-Clark conspiracy (perhaps aided by the Schedule F plan to weaken civil service protections for federal workers, a plan that Trump and his allies seem eager to implement if he wins the election).

Given the potential for abuse under the Fitzgerald rule, how might the Supreme Court, if it were so inclined, improve matters? The clearest way forward would be to abandon absolute for qualified immunity. Admittedly, this would require at least abrogating, if not explicitly overruling, Fitzgerald. But Fitzgerald itself was hardly uncontroversial: a 5-4 opinion written over a furious dissent by Justice Byron White.

In addition, the Court would likely have to abandon the current “clearly established law” test, since personal lawsuits against the president (compared to, for example, police officers) are so rare that the absence of decided cases, and thus of “clearly established law,” would make qualified immunity functionally absolute. Instead, courts would likely need to go back to something like the common law standard of bad faith (I have argued elsewhere why considering presidential motive is not only consistent with but central to the constitutional design of the presidency).

The Fitzgerald Court rejected qualified immunity because of concerns that any degree of liability “could distract a President from his public duties, to the detriment of not only the President and his office but also the Nation that the Presidency was designed to serve.” More specifically, as Chief Justice Warren Burger noted in his concurrence:

The enormous range and impact of Presidential decisions—far beyond that of any one Member of Congress—inescapably means that many persons will consider themselves aggrieved by such acts. Absent absolute immunity, every person who feels aggrieved would be free to bring a suit for damages, and each suit—especially those that proceed on the merits—would involve some judicial questioning of Presidential acts, including the reasons for the decision, how it was arrived at, the information on which it was based, and who supplied the information. Such scrutiny of day-to-day decisions of the Executive Branch would be bound to occur if civil damages actions were made available to private individuals. Although the individual who claims wrongful conduct may indeed have sustained some injury, the need to prevent large-scale invasion of the Executive function by the Judiciary far outweighs the need to vindicate the private claims.

These are valid concerns, but here are three responses. First, qualified immunity could be tuned to provide a high degree of protection to the president. For example, courts could implement a strong presumption of regularity, meaning that presidential actions are assumed to be lawful unless proved otherwise. In addition, courts could require heightened pleading standards of litigants, requiring plaintiffs to present a clear and compelling case of presidential bad faith before a lawsuit can proceed.

Second, absolute immunity could be preserved for a small subset of “core” presidential actions, where the Constitution grants the president preclusive powers relative to the other branches. These include actions where the president has total discretion, such as making appointments, issuing pardons, and vetoing legislation.

Third, courts could compensate for eliminating or restricting the application of absolute immunity by expanding what constitutes a presidential “official act,” thus granting the president qualified immunity where he currently receives no immunity at all. For example, in Blassingame, the D.C. Circuit rejected Trump’s argument that he should automatically receive immunity for speaking on matters of public concern. In the context of absolute immunity, this makes sense, since accepting Trump’s argument would vastly expand the scenarios in which he would receive absolute immunity. But if the standard were qualified, rather than absolute, immunity, there would be a much smaller downside to presuming that an action by a president was official. For example, under a qualified immunity regime, it would be reasonable to classify Trump’s remarks at the “Stop the Steal” rally as official because, although in the context of a campaign event, they were about matters of profound public concern: whether the 2020 election was in fact stolen. But because Trump would enjoy only qualified, rather than absolute, immunity, he could still be found liable if, as was the case, it could be shown that he acted in bad faith.

A Unified Standard Based on Qualified Immunity for Civil and Criminal Liability

The other major benefit of a qualified, rather than absolute, immunity standard for official presidential acts is that it could be applied to all forms of liability, civil and criminal. A single standard is desirable because the considerations that go into civil versus criminal immunity for the president are not so different. In both cases, the concern is about not excessively chilling the president’s exercise of lawful power while still vindicating litigants’ and the public’s rule-of-law interests. 

In trying to distinguish the two contexts, the D.C. Circuit emphasized the public’s “fundamental interest in the enforcement of criminal laws,” which it argued was heightened in this case, given the “citizenry’s interest in democratically selecting its President.” The D.C. Circuit also downplayed the chilling effect of criminal liability as too speculative, emphasizing the procedural safeguards that exist—such as the prosecutor’s ethical obligations and the role of the grand jury—to protect against abusive prosecutions.

These are perfectly fine arguments in and of themselves, but they don’t successfully distinguish criminal from civil immunity. While the public’s interest may well be greater in criminal prosecutions than in criminal cases, the threat to the president’s life and liberty are correspondingly higher, and thus so is the potential chilling effect of liability on presidential action. And while it takes more to indict a president than to sue him, the protections that the D.C. Circuit cited do not change the fact that, as Justice Robert Jackson famously observed, the “prosecutor has more control over life, liberty, and reputation than any other person in America.” As to the court’s argument that “the prospect of federal criminal liability might serve as a structural benefit to deter possible abuses of power and criminal behavior,” one could say the same about civil liability.

At the same time, it’s hard to fault the D.C. Circuit for not applying Fitzgerald absolute immunity to the criminal context. What ultimately drove the court’s decision was its unwillingness to “accept that the office of the Presidency places its former occupants above the law for all time thereafter,” an unwillingness that I fully share. But it would be better to get at the root of the problem: Fitzgerald’s excessive grant of immunity, not putative distinctions between the chilling effects of criminal and civil liability.

How, then, would a qualified immunity regime work in the criminal context? First, it would likely need to recognize, as noted above, at least some cases in which absolute, not qualified, immunity is the correct standard. As Trevor Morrison has argued, the logic of Article II and separation-of-powers cases like Youngstown suggest that presidents should be absolutely immune from criminal prosecutions in those “circumstances in which the prosecution would punish a former president for the exercise of an exclusive, unregulable presidential power” and that criminal statutes should thus be read not to apply to such powers unless they do so clearly. But such circumstances would be narrow and, at least in Trump’s case, would not be a basis for criminal immunity even if he acted in an official, rather than just private, capacity.

For all other official acts, the president would enjoy only qualified, rather than absolute, immunity. This raises the question of what qualified immunity means in the criminal context, since the standards for criminal conviction are already much higher than for a finding of civil liability. Convictions require proof beyond a reasonable doubt and a unanimous jury, whereas civil liability requires only a preponderance of evidence. In addition, most criminal laws require some level of mens rea (the “guilty mind”), whereas a defendant can often be held civilly liable even in the absence of wrongful intent. 

But this is actually a feature, not a bug. By its very nature, criminal law already encodes the core idea behind qualified immunity: that punishment should be reserved for clear, rather than borderline, cases of malfeasance. Whether one calls this “qualified criminal immunity” or just ordinary criminal liability is irrelevant. Either way, the president would be safe from frivolous prosecution.

* * *

Ultimately, Trump is likely to lose, as he should, his argument for absolute criminal immunity in the Jan. 6 case. But it matters why he loses: not because the laws of presidential civil and criminal immunity are so different, but because, properly interpreted, they apply the same way. I recognize that my argument, with respect to civil immunity, is about the law as it could be, not as it is. But the Supreme Court has an opportunity to both unify the doctrines and, in so doing, fix a large vulnerability in the law of presidential civil immunity. A unified qualified immunity standard, with a limited role for absolute immunity for core presidential actions, would properly balance the need to safeguard vigorous presidential action with making the president, no less so than any other person, answerable to the law.


Alan Z. Rozenshtein is an Associate Professor of Law at the University of Minnesota Law School, a senior editor at Lawfare, and a term member of the Council on Foreign Relations. Previously, he served as an Attorney Advisor with the Office of Law and Policy in the National Security Division of the U.S. Department of Justice and a Special Assistant United States Attorney in the U.S. Attorney's Office for the District of Maryland.

Subscribe to Lawfare