What is the pardon power?

The pardon power is vested in the President of the United States by Article II, Section 2, Clause 1 of the Constitution: the President 'shall have Power to grant Reprieves and Pardons for Offenses against the United States, except in Cases of Impeachment.' Three structural features of this clause define the outer limits of the power. First, it extends only to offenses against the United States -- federal criminal offenses and contempts of federal authority. It does not reach state criminal convictions, state criminal proceedings, or civil liability of any kind; a presidential pardon cannot free a person convicted of a state crime, and the fifty state governors hold their own independent pardon powers under their respective state constitutions. Second, the power may be exercised at any stage: before indictment, before trial, before conviction, or after conviction and entry of judgment; a reprieve delays punishment without erasing guilt, while a full pardon extinguishes both guilt and punishment. Third, the 'except in Cases of Impeachment' exception removes the pardon power entirely from the constitutional removal and disqualification mechanism that Article I allocates to the House and Senate: the President may not pardon a federal officer to prevent the House from impeaching, to abort a Senate trial underway, or to undo a disqualification judgment already entered by the Senate under Article I, Section 3, Clause 7. The constitutional design reflects a deliberate Founding choice to vest the power in a single executive rather than a council or a court, on the premise that mercy and discretion in individual cases require a defined locus of accountability and that pardons should function as acts of grace for which the executive is politically answerable. Hamilton in Federalist No. 74 (1788) argued that the power of pardoning should be entrusted to the executive because seasons of insurrection often demand a conduct precisely the reverse of that which a constitutional interdiction of pardons would impose -- a legislature lacks the flexibility to respond with the necessary promptness -- and that the responsibility in the executive is a sufficient security against the abuse of this trust.

Ex parte Garland, 71 U.S. (4 Wall.) 333 (1867), remains the Supreme Court's most expansive description of the pardon power's scope. Augustus Hill Garland was an Arkansas attorney admitted to practice before the Supreme Court before the Civil War who subsequently served in the Confederate Congress. After the Confederate surrender President Andrew Johnson issued a full pardon to Garland covering all offenses related to his participation in the rebellion. Garland then sought to resume practice before the Supreme Court, but a federal statute required attorneys to take a loyalty oath swearing they had never engaged in acts of hostility against the United States -- an oath Garland could not truthfully take because of his Confederate service. He argued that the presidential pardon erased the disqualification. The Supreme Court agreed, 5-4, with Justice Field writing for the majority. Field's description of the pardon power's effect is the most-quoted passage in American pardon-power jurisprudence: a pardon reaches both the punishment prescribed for the offence and the guilt of the offender; when the pardon is full it releases the punishment and blots out of existence the guilt, so that in the eye of the law the offender is as innocent as if he had never committed the offence. The pardon, once granted and accepted, cannot be revoked; it is a completed act of executive grace. The disqualification in the loyalty-oath statute was, in the majority's analysis, a punishment for the past offense and therefore subject to erasure by the pardon. The four dissenters -- Justices Miller, Swayne, Davis, and Chief Justice Chase -- would have held that the loyalty oath set a qualification for the practice of law, not a punishment for past conduct, and therefore lay outside the pardon power's reach. Garland himself later served as Attorney General of the United States under President Grover Cleveland from 1885 to 1889.

Burdick v. United States, 236 U.S. 79 (1915), resolved two questions about the legal nature of a pardon: whether it must be accepted by the recipient to be effective, and what a pardon implies about the recipient's guilt. George Burdick was the editor of the New York Tribune who was subpoenaed before a federal grand jury investigating a newspaper story about customs house corruption. Burdick refused to answer questions, invoking his Fifth Amendment privilege against self-incrimination. President Woodrow Wilson then tendered Burdick a full and unconditional pardon for all offenses he had committed or might have committed in connection with the matters before the grand jury. Burdick refused the pardon and continued to decline to testify. He was held in contempt. The Supreme Court held unanimously, in an opinion by Justice McKenna, that Burdick's refusal of the pardon was legally effective: a pardon is a deed, to the validity of which delivery is essential, and delivery is not complete without acceptance by the recipient. Because Burdick refused it, the pardon was not operative and could not strip him of his Fifth Amendment privilege. More significantly for subsequent constitutional debates, the Court articulated a fundamental characteristic of a pardon: the grace of a pardon carries with it an imputation of guilt; acceptance is an admission of the criminal act that the pardon addresses. A person of honor may prefer to endure the civil consequences of refusal rather than have imputed to them, and thereby admit, the commission of the criminal offense. This principle -- that a pardon implies a confession -- carries structural implications for the self-pardon debate: a President who would pardon himself not only acts as a judge in his own cause but, under Burdick's logic, simultaneously admits to federal criminal conduct while escaping accountability for it.

The question of whether the President may pardon himself has never been decided by any federal court and remains an open constitutional question. The Article II, Section 2, Clause 1 text does not include any restriction on who may receive a pardon beyond the Impeachment exception; it speaks of the President's power to grant pardons without specifying that the recipient must be a person other than the grantor. Those who argue a self-pardon is constitutionally permissible point to this textual silence: the Framers expressly limited the power in the Impeachment exception when they intended a limit, and they imposed no additional express restriction. Those who argue a self-pardon is constitutionally impermissible advance two principal structural arguments. First, the principle that no person may be a judge in their own cause -- nemo judex in causa sua -- is one of the foundational axioms of Anglo-American law and predates the Constitution; a President who pardons himself exercises judgment in a proceeding in which he is the subject, which the common-law constitutional tradition has consistently regarded as incompatible with the rule of law. Second, Article II, Section 3's Take Care Clause requires the President to faithfully execute the laws; using the pardon power to immunize oneself from criminal liability for one's own conduct may be inconsistent with the duty to execute the laws faithfully rather than to manipulate their application for personal benefit. During the Watergate crisis of 1974, the Department of Justice's Office of Legal Counsel examined the self-pardon question and prepared a legal opinion concluding that the constitutional structure did not support a self-pardon: under the no-man-may-be-his-own-judge principle and the separation of powers underlying the pardon clause, the President cannot pardon himself. That opinion is not a judicial decision and is not binding on federal courts; it represents a legal interpretation by one executive branch office. President Nixon resigned before the question required a definitive resolution. No subsequent President has sought a judicial ruling on the constitutional permissibility of a self-pardon, and the question remains unresolved.

For the 2028 election, the pardon power will intersect with at least four categories of political and legal debate during the presidential term beginning January 20, 2029. First, pardons of federal officials from prior administrations: incoming administrations face recurring decisions about pardoning officials from prior administrations who were convicted or who face pending federal charges; the political and legal consequences of such pardons -- including the Burdick implication-of-guilt principle, the possibility that such pardons serve factional loyalty rather than mercy, and the limit that a pardon does not reach independent state charges -- are certain to be contested. Second, the self-pardon question: the constitutional validity of a self-pardon is unresolved; whether the incoming President asserts a self-pardon theory, whether it is challenged in court, and what a court would hold if it reached the merits are all contingent on the specific facts of any federal criminal proceedings involving the President. Third, congressional oversight constraints: Congress possesses no enumerated authority to regulate the President's exercise of the pardon power in advance; statutes purporting to require reporting, impose waiting periods, or establish substantive criteria for pardon decisions would raise serious separation of powers concerns under the Article II vesting clause and the structural commitment of the pardon power to the President's discretion. Political accountability -- through elections, public opinion, and congressional hearings after the fact -- is the primary institutional check on pardon decisions. Fourth, the federal-versus-state boundary: the limit of the pardon power to federal offenses means that state prosecutions proceed independently of any presidential pardon; the question of whether state charges duplicate or supplement a pardoned federal prosecution involves both substantive law and the principle of dual sovereignty under which federal and state governments are each sovereign within their respective spheres. The 33 Class II Senate seats in the 2028 elections will determine which party controls the Senate and which committees may conduct oversight hearings, subpoena witnesses, and generate the political record that constitutes accountability for pardon decisions even when judicial review of the pardon itself is unavailable.