The Constitutionality of Non-Specific Pardons

As his peculiar presidency lurches to a reluctant close, Donald Trump is reported to be considering yet another novelty – a parting bouquet of pardons for his family, his associates, and perhaps even himself.

As I have explained elsewhere, a self-pardon is very probably unconstitutional. But the point is untested, and it seems quite likely Trump will try it. As for family and friends, a president can pardon any and all federal (but not state) crimes. There is no requirement that the offense have already resulted in a conviction, or even an indictment. It is sufficient that the criminal conduct occurred before the issuance of the pardon.

A pardon issued for provably corrupt reasons – for example, in return for a bribe or for the express purpose of obstructing investigation of wrongdoing by the president himself –  might itself be a crime.  But handing out a pardon to a relative or political supporter just because you like them, or as a political reward, though sleazy and perhaps a proper ground for impeachment, is neither a crime nor a basis for revoking the pardon. Indeed, although others disagree, I think a pardon, once issued and accepted, is irrevocable.

It seems, therefore, that nothing stands in the way of a Trump pardon spree except for the doubtful limit of his own gall. However, it has recently been suggested, most persuasively by Professor Aaron Rappaport, that there exists a technical restriction on the scope of pardon grants that could put a stick in the spokes of Trump’s pardon wheel. The claim is that the facially unlimited language of Article II – the president “shall have Power to grant Reprieves and Pardons for Offences against the United States, except in Cases of Impeachment” – contains an implicit requirement of specificity. Broadly, the idea is that, to be valid, a pardon must specifically identify the crimes or acts being pardoned. Unfortunately for those looking for limits to the pardon power, this theory finds scant backing in law, history, or practice.

An Implausible Implied Specificity Requirement

At least according to Trump’s most ardent critics, he and those around him have been involved in a wide array of private and public conduct that might subject them to criminal liability. The objective of a Trump pardon spree would be to block federal prosecution on all conceivable grounds. If there is a specificity requirement for pardons, it might preclude issuance of a blanket pardon for “all violations of federal criminal law” or even for general categories of crime, and require instead a detailed enumeration of all the crimes or acts intended to be covered. Drafting a pardon under such a specificity constraint would pose a dilemma for Trump. If it were insufficiently specific, it would leave open avenues of prosecution. If it were truly comprehensive, it could constitute a roadmap for state prosecutors, federal civil regulatory authorities, congressional investigators, and private litigants to all of Trump World’s vulnerabilities – and, perhaps equally important for Trump, undermine his public narrative of “perfect” conduct both in and out of office.

Sadly, I do not believe there is a meaningful constitutional specificity requirement for pardons. The argument for such a requirement runs something like this: Despite the unrestricted language in the constitution’s pardon clause, the Framers nonetheless understood there to be a specificity limitation because they meant to incorporate British pardon practice into the American constitution and their shared understanding of that practice included a specificity limit.

There is nothing inherently wrong with this form of argument. I employ something like it to reject the constitutionality of self-pardons: There is no record in British history of a self-pardon; there is strong evidence from things the Founders said or wrote that they never imagined a president could pardon himself; and the Framers included three constitutional remedies for serious presidential misconduct – quadrennial elections, impeachment, and the possibility of prosecution in ordinary criminal courts – the last of which would be effectively nullified if a president could pardon himself.

But in the case of pardon specificity, there is no evidence in the record of the Constitutional Convention or subsequent ratifying debates that anyone thought the grant of pardon power was qualified by a specificity limit. The point was never raised directly or by implication.

The idea that, to be valid, a pardon must specifically identify the crimes or acts being pardoned … finds scant backing in law, history, or practice.

The History of a Sovereign Prerogative

Likewise, there is no evidence for a general specificity requirement in British pardon practice prior to the American founding. In Britain, the pardon power was an aspect of royal prerogative. That is, in ancient times, the law itself was said to be an emanation of the will of the king. Accordingly, the king could, as an act of grace, release his subjects from the operation of the law.

English kings and queens routinely issued general pardons covering all crimes, or subsets of them.  Magna Carta contains a broad pardon from King John to his rebellious barons. Elizabeth I proclaimed multiple general pardons. As over the centuries Parliament took an ever-greater share of authority, it began playing a larger role in general pardons, periodically passing bills proposed by the monarch proclaiming amnesty for various classes of offenses. For example, in 1750, George II proposed and Parliament passed “An Act for the King’s Most Gracious, General, and Free Pardon” (which nonetheless contained pages of exceptions to its coverage).  But even in the middle of the eighteenth century, commentators like William Blackstone were at pains to emphasize that all pardons derived their legal force from the crown’s sovereign prerogative of mercy.

For our purposes, the point about these general pardons, whether by royal proclamation or parliamentary bill, is that they were inherently non-specific. They pardoned classes of offenses and offenders, not particular people and not particular instances of particular crimes. Anyone who fell within the broad terms of the pardon could claim it.

The Crown could also issue special “charters of pardon” to individuals. Here there does seem to have been a rule of specificity, but of a special kind for a particular reason. The basic rule (which was really more a presumption than a rule) was that a pardon using general language might not cover a particular crime if the king was unaware of it at the time the pardon was issued, the theory being that the pardon would not have issued had he known. This is the proper context in which to understand Blackstone’s observation, referenced by Rappaport, that “General words have . . . a very imperfect effect in pardons.” The presumption favoring specificity emerged to prevent inadvertent overbreadth in pardons.  It was never meant to address a situation where the Crown intended to pardon a person for a wide array of offenses, and I am aware of no case in which it was applied in that way.

Moreover, careful study of British pardon practice up to the time of the American founding reveals it to be an almost impenetrable tangle in which theoretically unconstrained royal prerogative, struggles for supremacy between Crown and Parliament, and judicial interpretation produced a shifting array of hyper-technical limitations, exceptions, and exceptions to the exceptions. The suggestion the Framers had a clear common understanding of one, and only one, particular aspect of this historical jumble and meant that understanding to be imported sub silentio into the pardon clause is very hard to credit.

American Pardon Practice

Far more important than speculations about the Framers’ unexpressed understandings of unclear British law is the evidence of how American presidents have used the pardon power and the reaction of American courts to those uses.

Presidents have issued general pardons applying to classes of people and broad categories of offenses many times. Between 1795 and 1977, presidents issued at least thirty amnesties, among them multiple proclamations by Presidents Lincoln and Andrew Johnson during and after the Civil War pardoning Confederates, and clemency for Vietnam-era deserters and draft evaders granted by Presidents Ford and Carter.

Consider in particular President Madison’s 1815 pardon of followers of the pirate and smuggler Jean Lafitte who aided American forces at the Battle of New Orleans during the War of 1812.  Madison’s proclamation granted to all who assisted in the defense of Louisiana:

a full and free pardon of all offenses committed in violation of any act or acts of the Congress of the said United States touching the revenue, trade, and navigation thereof or touching the intercourse and commerce of the United States with foreign nations at any time before the 8th day of January, in the present year 1815, by any person or persons whomsoever being inhabitants of New Orleans and adjacent country, or being inhabitants of the said island of Barrataria and the places adjacent . . .

The language of this proclamation is obviously not specific, either as to persons, conduct, or statutory violations. It was self-evidently designed to let the Barrataria brigands off the hook for all their prior maritime skullduggery without inquiring too carefully into the details. Critically, its issuer was James Madison, revered as the father of the Constitution. If the Framers intended a specificity limitation on pardons, Madison somehow missed the point or, once he became president, was happy to violate the restriction.

In over two hundred years, no court has ever voided a presidential pardon, individual or group on the ground of insufficient specificity. Indeed, in a series of cases following the Civil War, the Supreme Court repeatedly affirmed the validity of group pardons, and observed that the pardon power “extends to every offence known to the law, and may be exercised at any time after its commission, either before legal proceedings are taken or during their pendency or after conviction and judgment.”

One might, I suppose, argue that pardons of groups can be general, while pardons of individuals must be specific.  But it would be a very odd constitutional rule that held a president could issue a non-specific pardon to thousands of people, but could not do so for one person.

The Flynn Pardon

In recent days, many have expressed unease with the broad language of Trump’s pardon of Michael Flynn. But the phrasing of that pardon is explainable due to its unusual procedural posture. The vast majority of individual pardons are awarded to persons who have already been convicted and sentenced for some particular crime. Therefore, they seek (preferably through the formal procedures set up by the U.S. Pardon Attorney) a pardon for the crime(s) of which they stand convicted. If a pardon issues, it will be limited to the offense(s) of conviction.

Moreover, because the vast majority of individual pardons for actual convictions are awarded long after those convictions, and even longer after the crimes that led to those convictions, there will almost never be any practical worry that the government might try to prosecute the clemency recipient for some related crime not technically covered by the pardon. Either for double jeopardy reasons or because the statute of limitations will have run, the pardon puts a period on the matter.

By contrast, Flynn pleaded guilty to one crime – lying to the FBI – but he had not yet been sentenced, and moreover remained theoretically subject to prosecution on other charges for his conduct before and after the 2016 election. The objective of Trump’s pardon was to relieve Flynn of liability for any past crime that might, for example, have fallen within the remit of the Mueller investigation.

Another relatively recent example of a similarly broad pardon – on which Bill Barr advised – is President George H.W. Bush’s pardon of those involved in Iran-Contra. The pardon proclamation extended to: “all offenses charged or prosecuted by Independent Counsel Lawrence E. Walsh or other member of his office, or committed by these individuals and within the jurisdiction of that office.

Even if one thinks (as I do) that the Flynn pardon is disgraceful, and the Iran-Contra pardons were questionable, there is nothing inherently improper about drafting pardon language broadly to embrace all potential liability for a complex set of events.  After all, that is what James Madison did for the Barrataria pirates.

As Judge Emmet Sullivan emphasized in his order dismissing Flynn’s indictment due to the pardon, a president’s decision to issue such a pardon is “political.” Absent proof of prosecutable corruption in the issuance of a pardon, the remedies for abuse of the pardon power are likewise political, not legal.


I harbor one lingering reservation on specificity. President Gerald Ford pardoned Richard Nixon for any and all violations of federal law from the date of his inauguration to the date of his resignation. The breadth of that pardon was unique in American history, but was never challenged in court. I do wonder whether the Constitution grants a president quite that much authority.

It is one thing to pardon a group or a person for participation in (or evading participation in) a continuous event like a war, or for a class of crime that perhaps ought not to have been prosecuted in the first place, or for all crimes prosecutable in connection with even a broadly defined series of events. But it seems categorically different to say that a president can wipe a person’s slate entirely clean for any and every federal crime, known and unknown. Still, I cannot find a constitutional basis, or a useful rule of decision, for distinguishing between a pardon covering all crimes (bad) and a pardon covering some broadly defined subset of less-than-all crimes (OK). Perhaps Trump will provide an occasion to explore that problem in the coming weeks.

Photo by Chip Somodevilla/Getty Images.


About the Author(s)

Frank O. Bowman, III

Frank O. Bowman, III, is University of Missouri Curators’ Distinguished Professor and Floyd R. Gibson Missouri Endowed Professor of Law at the University of Missouri School of Law, as well as Dean’s Visiting Scholar at Georgetown University Law Center. He is a former federal and state prosecutor. His most recent book is High Crimes & Misdemeanors: A History of Impeachment for the Age of Trump (Cambridge University Press 2019). Follow him on Twitter (@FOBowman3).