Constitutionality of the Presidential Records Act

• Court: Department of Justice Office of Legal Counsel
• Decided: April 1, 2026
• Status: Published

Opinion

(Slip Opinion)

Constitutionality of the Presidential Records Act The Presidential Records Act is unconstitutional because it exceeds Congress’s enumerated and implied powers and aggrandizes the Legislative Branch at the expense of the consti- tutional independence and autonomy of the Executive.

April 1, 2026

MEMORANDUM OPINION FOR THE COUNSEL TO THE PRESIDENT

You have asked whether the Presidential Records Act of 1978 (“PRA” or “Act”) is constitutional. We conclude that it is not. The PRA is unconstitutional for two independent but interlocking reasons: It exceeds Congress’s enumerated and implied powers, and it aggrandizes the Legislative Branch at the expense of the constitutional independence and autonomy of the Executive. “Every law enacted by Congress must be based on one or more of its powers enumerated in the Constitution,” United States v. Morrison, 529 U.S. 598, 607 (2000), or “such implied powers as are necessary and proper to carry into effect the enumerated powers,” Carter v. Carter Coal Co., 298 U.S. 238, 291 (1936). And congressional attempts to regulate the Presidency directly raise heightened separation of powers concerns. See The Constitutional Separation of Powers Between the President and Congress, 20 Op. O.L.C. 124, 126–29 (1996) (“Separation of Powers”). The PRA exceeds the oversight power because it serves no identifiable and valid legislative purpose. It exceeds any preservation power because Congress cannot preserve presidential records merely for the sake of posterity. It exceeds Congress’s regulatory power over statutory agencies because it purports to regulate a constitutional office—the Presidency— that Congress did not create and that Congress cannot abolish. It exceeds the spending power, because that power allows Congress to incentivize outcomes with federal funding, not to directly regulate coordinate branch- es of government. And it exceeds Congress’s power to assist in the execu- tion of the powers vested in coordinate branches because it restricts rather than empowers the President. Just as Congress could not constitutionally invade the independence of the Supreme Court and expropriate the papers of the Chief Justice or Associate Justices, Congress cannot invade the independence of the President and expropriate the papers of the Chief Executive.

1 50 Op. O.L.C. __ (Apr. 1, 2026)

I.

We begin with history: Congress and the President have long main- tained the separation of powers through a tradition of negotiation and compromise, highlighting bedrock principles of constitutional structure and the interplay between Article I enumerated powers and Article II independence. Over the first two centuries of the American experiment in self-government, Presidents owned and controlled presidential papers, and Congress obtained such papers through political negotiation and inter- branch accommodation, rather than as a matter of right. That historical practice was interrupted by the Presidential Recordings and Materials Preservation Act (“PRMPA”), Pub. L. No. 93-526, 88 Stat. 1695 (1974), which Congress passed in the midst of the Watergate inves- tigation in order to abrogate an agreement that would have allowed former President Richard M. Nixon to direct the destruction of his own presiden- tial records following his resignation. That minor crack became a major fissure with the PRA, which Congress passed in 1978 to regulate all presidential records prospectively for the first and only time in American history.

A.

“The Constitution reflects a fundamental conviction that governmental ‘power is of an encroaching nature, and that it ought to be effectively restrained from passing the limits assigned to it.’” Separation of Powers, 20 Op. O.L.C. at 125 (quoting The Federalist No. 48, at 332 (James Madison) (Jacob E. Cooke ed., 1961)). The Framers’ solution to the perils of government power was simple and elegant: “divide it.” Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2202 (2020). The Constitution thus creat- ed three branches of government, vesting each with its own type of power: “[T]he legislature makes, the executive executes, and the judiciary construes the law.” Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 46 (1825) (Marshall, C.J.). Within this divided system, the Chief Executive “occupies a unique po- sition.” Nixon v. Fitzgerald, 457 U.S. 731, 749 (1982). The Framers considered the President’s “energetic, vigorous, decisive, and speedy execution of the laws . . . constitutionally indispensable” for “good gov- ernment.” Trump v. United States, 144 S. Ct. 2312, 2329 (2024) (citations

2 Constitutionality of the Presidential Records Act

omitted). The President’s powers are of “unrivaled gravity and breadth.” Trump v. Vance, 140 S. Ct. 2412, 2425 (2020). “Quite appropriately, those duties come with protections that safeguard the President’s ability to perform his vital functions.” Id. Article II of the Constitution “guarantees the independence of the Ex- ecutive Branch” in various ways. Id. Some of the President’s responsibili- ties, like the authority to issue pardons and reprieves, are “conclusive and preclusive,” categorically disabling the other branches from examining or regulating the President’s actions. Youngstown Sheet & Tube Co. v. Saw- yer, 343 U.S. 579, 638 (1952) (Jackson, J., concurring); see Trump, 144 S. Ct. at 2327–28; cf. Fitzgerald, 457 U.S. at 755 (recognizing “absolute Presidential immunity from damages liability for acts within the ‘outer perimeter’ of his official responsibility”). Other powers, while not “con- clusive and preclusive,” come with their own set of Article II protections and immunities. See, e.g., Trump, 144 S. Ct. at 2331–32 (cautioning that the President’s official acts should not be “routinely subjected to scrutiny in criminal prosecutions”); United States v. Nixon, 418 U.S. 683, 708 (1974) (recognizing “a presumptive privilege for Presidential communica- tions” that is “fundamental to the operation of Government and inextrica- bly rooted in the separation of powers under the Constitution”). But Article I’s enumeration of congressional powers also secures the coordinate branches’ independence. The Framers “were particularly concerned with the Congress’s potential for improvident or overreaching action.” Separation of Powers, 20 Op. O.L.C. at 126; see The Federalist No. 49, at 315–16 (James Madison) (Clinton Rossiter ed., 1961) (“[T]he tendency of republican governments is to an aggrandizement of the legis- lat[ure] at the expense of the other departments.”); Bowsher v. Synar, 478 U.S. 714, 727 (1986) (“The dangers of congressional usurpation of Execu- tive Branch functions have long been recognized. The debates of the Constitutional Convention, and the Federalist Papers, are replete with expressions of fear that the Legislative Branch of the National Govern- ment will aggrandize itself at the expense of the other two branches.” (cleaned up)). By limiting the Legislature’s authority to only those powers “herein granted,” U.S. Const. art. I, § 1, the Constitution sought to extin- guish the “propensity of the legislative branch to invade the rights of the Executive,” Separation of Powers, 20 Op. O.L.C. at 126 (internal quota- tion marks omitted); see also Gibbons v. Ogden, 22 U.S. (9 Wheat.) 1,

3 50 Op. O.L.C. __ (Apr. 1, 2026)

195 (1824) (“The enumeration presupposes something not enumerat- ed . . . .”). Articles I and II of the Constitution thereby operate in tandem. “The more the former expands, the more the latter shrinks.” Haaland v. Brack- een, 143 S. Ct. 1609, 1653 (2023) (Gorsuch, J., concurring). The separa- tion of powers ensures equilibrium. By limiting Congress to the exercise of enumerated powers, Article I safeguards “the autonomy and independ- ence of the Presidency” required by Article II and protects the powers vested in the President from encroachment by Congress. Testimonial Immunity Before Congress of the Former Counsel to the President, 43 Op. O.L.C. 108, 125 (2019) (“Testimonial Immunity”). And Article II cabins the Legislature’s exercise of its Article I powers when Congress directs those powers against the President. Thus, it cannot be lightly assumed that Congress’s expressly enumerat- ed Article I powers imply further powers that reach the President to the same degree that they reach other entities and individuals. On the contra- ry, given the President’s “unique position in the constitutional scheme,” Fitzgerald, 457 U.S. at 749, the Supreme Court has repeatedly made clear that “special considerations control when the Executive Branch’s interests in maintaining the autonomy of its office and safeguarding the confidenti- ality of its communications are implicated,” Cheney v. U.S. Dist. Ct. for D.C., 542 U.S. 367, 385 (2004); see, e.g., Trump, 144 S. Ct. at 2330–32; Trump v. Mazars USA, LLP, 140 S. Ct. 2019, 2034 (2020) (“[C]ongr- essional subpoenas directed at the President differ markedly from congressional subpoenas we have previously reviewed . . . .”); Congres- sional Oversight of the White House, 45 Op. O.L.C. __, at *2 (Jan. 8, 2021) (“Congressional Oversight”) (“Even when Congress operates within the appropriate scope of its oversight authority, the Constitution places additional separation of powers constraints on inquiries directed at the White House.”).

B.

Although Congress has no expressly enumerated constitutional power to request information from the President, it generally enjoys an implied “power of inquiry” as “an essential and appropriate auxiliary to the legis- lative function.” McGrain v. Daugherty, 273 U.S. 135, 174 (1927). When Congress invokes that power to seek information from the President, it

4 Constitutionality of the Presidential Records Act

has the potential to set the two political branches on a collision course, pitting the Legislature’s Article I prerogatives against the Executive’s Article II independence. “Congress and the President,” the Court has explained, “have an ongoing relationship that the Framers intended to feature both rivalry and reciprocity.” Mazars, 140 S. Ct. at 2026. Because this “distinctive aspect” of congressional requests for the President’s documents “necessarily informs our analysis” of the separation of powers, id., we must understand the ways in which longstanding historical practice has liquidated the limits on Congress’s power of inquiry, see Zivotofsky ex rel. Zivotofsky v. Kerry, 576 U.S. 1, 23 (2015) (“In separation-of-powers cases this Court has often ‘put significant weight upon historical prac- tice.’” (citation omitted)). “History, custom, and usage indicate unequivocally that, prior to [1974], Presidents exercised complete dominion and control over their presidential papers.” Nixon v. United States, 978 F.2d 1269, 1277 (D.C. Cir. 1992); Title to Presidential Papers—Subpoenas, 43 Op. Att’y Gen. 11, 17 (1974) (“[T]he principle of Presidential ownership of White House materials has been acknowledged by all three branches of the Government from the earliest times.”). Consistent with this principle of presidential control, Congress has had to resort to the give-and-take of the political process rather than legislative fiat to obtain a President’s records, both while he is in office and after he leaves it. And presidential “refusals to disclose information to Congress” have “by no means been unprecedent- ed.” History of Refusals by Executive Branch Officials to Provide Infor- mation Demanded by Congress: Part I—Presidential Invocations of Executive Privilege Vis-à-Vis Congress, 6 Op. O.L.C. 751, 751–52 (1982) (“Refusals to Provide Information”). Yet, for the first two centuries of American history, “disputes over congressional demands for presidential documents [did] not end[] up in court,” nor did Congress seek to compel the production of such documents by statute. Mazars, 140 S. Ct. at 2029. Instead, Congress and the President have long sought to understand and accommodate each other’s interests with respect to presidential records. Begin with George Washington. In 1792, a congressional committee requested documents in President Washington’s possession about military campaigns against Indian Tribes. See Refusals to Provide Information, 6 Op. O.L.C. at 752. President Washington called a Cabinet meeting on the subject, aware that his response could “become a precedent.” Id.

5 50 Op. O.L.C. __ (Apr. 1, 2026)

(quoting 1 The Writings of Thomas Jefferson 303 (Andrew A. Lipscomb ed., 1903)). Those present included Alexander Hamilton, Thomas Jeffer- son, Edmund Randolph, and Henry Knox. All were of “one mind” that Congress could request papers from the President, but the President could “exercise a discretion” in deciding whether and to what extent to respond, “communicat[ing] such papers as the public good would permit” and “refus[ing]” the rest. Id. (emphasis and citation omitted). President Washington sent Secretary of State Jefferson on his behalf to negotiate with the congressmen, and the House committee narrowed its request, which President Washington fulfilled. See id. at 752–53; Mazars, 140 S. Ct. at 2030. Washington continued to exercise discretion in his disclosures to Con- gress. In 1794, for instance, the Senate requested correspondence involv- ing the United States Minister to France. See Refusals to Provide Infor- mation, 6 Op. O.L.C. at 753. “President Washington submitted certain of the correspondence requested, but withheld ‘those particulars which, in [his] judgment, for public considerations, ought not to be communicat- ed.’” Id. (quoting 1 J. Richardson, Messages and Papers of the Presidents 152 (1896)). Two years later, the House sought from the President the instructions he had given to the United States Minister who negotiated the Jay Treaty, as well as correspondence and records pertaining thereto. See id. “President Washington denied the House’s right to demand and receive any of the papers requested,” even though implementation of the Jay Treaty “required an appropriation which the House was called upon to vote.” Id. Because the requested materials included records of foreign negotiations, Washington feared that supplying such documents to Con- gress “might have pernicious influence on future negotiations; or produce immediate inconveniences, perhaps danger and mischief, in relation to other Powers.” Id. Representatives debated Washington’s refusal on the House floor. James Madison, although taking issue in some respects with the President’s refusal, acknowledged during those debates “that the Executive had a right, under a due responsibility, also, to withhold infor- mation, when of a nature that did not permit a disclosure of it at the time.” Id. at 754 (quoting 5 Annals of Cong. 773 (1796)). The House ultimately took no action upon Washington’s refusal. “The practice of refusing congressional requests for information . . . was employed by many Presidents in the ensuing years.” Congressional

6 Constitutionality of the Presidential Records Act

Requests for Confidential Executive Branch Information, 13 Op. O.L.C. 153, 155 (1989) (“Congressional Requests”). In 1798, the House request- ed from President Adams diplomatic material concerning U.S. representa- tives to France. See Refusals to Provide Information, 6 Op. O.L.C. at 754. President Adams transmitted only some of the requested documents to Congress and “omitted ‘some names and a few expressions descriptive of the persons’ involved.” Id. (citation omitted). When Jefferson later assumed the Presidency, he, too, “followed Wash- ington’s precedent.” Mazars, 140 S. Ct. at 2030. In 1807, the House requested that President Jefferson provide any information in his posses- sion about a conspiracy to invade Spanish territory in North America with private forces. See Refusals to Provide Information, 6 Op. O.L.C. at 754–55 (citing 16 Annals of Cong. 336 (1806)). Jefferson declined to furnish the “voluminous” materials on the matter, id. at 755 (citation omitted), instead providing Congress “particular documents and a special message summarizing the conspiracy,” Mazars, 140 S. Ct. at 2030 (citing 16 Annals of Cong. 39–43). Further precedents abound. In 1825, President Monroe refused to give Congress documents related to charges against certain naval officers. See Refusals to Provide Information, 6 Op. O.L.C. at 755–56. In 1832, Presi- dent Jackson declined a congressional request for correspondence be- tween the United States and the Republic of Buenos Aires. Id. at 756–57. On another occasion, he withheld certain documents related to negotia- tions with Great Britain over the Northeastern Boundary, although he reconsidered his position nearly two years later. See id. at 757. In late 1833, President Jackson withheld a document requested by the Senate “relating to the removal of the deposits of the public money from the Bank of the United States and its offices.” Id. He stated that the Legisla- ture had no authority to “require of me an account of any communication, either verbally or in writing, made to the heads of Departments acting as a Cabinet council . . . [nor] might I be required to detail to the Senate the free and private conversations I have held with those officers on any subject relating to their duties and my own.” Id. (alterations in original) (citation omitted). In 1835, President Jackson similarly refused to produce documents related to his removal of the Surveyor General. Id. at 758. He explained, “[t]his is another of those calls for information made upon me by the Senate which have, in my judgment, either related to the subjects

7 50 Op. O.L.C. __ (Apr. 1, 2026)

exclusively belonging to the executive department or otherwise en- croached on the constitutional powers of the Executive.” Id. (citation omitted). He continued, “[s]uch a result, if acquiesced in, would ultimate- ly subject the independent constitutional action of the Executive in a matter of great national concernment to the domination and control of the Senate.” Id. (citation omitted). President Tyler, too, refused to furnish documents at several points in his Presidency, including when, in his view, the congressional request was unrelated to any valid legislative objective. See id. at 759–61. To name but one example, the House requested from the President and heads of departments information regarding executive appointments. See id. at 759. President Tyler repudiated this request, because the disclosure of such documents “could serve no ‘useful object connected with a sound and constitutional administration of the Government in any of its branches.’” Id. (citation omitted). He elaborated, “I can not perceive anywhere in the Constitution of the United States any right conferred on the House of Representatives to hear the reasons which an applicant may urge for an appointment to office under the executive department, or any duty resting upon the House of Representatives by which it may become responsible for any such appointment.” Id. (citation omitted). President Tyler also generally refused to accede to Congress’s requests for information relat- ing to the treaty to suppress the slave trade, evidence regarding steps taken to obtain Mexico’s recognition of claims made by American citi- zens, negotiations regarding the Northwestern Boundary, foreign corre- spondence regarding the ownership and occupation of the Oregon Territo- ry, and so on. See id. at 759–62. Presidents Polk, Fillmore, Buchanan, Lincoln, Johnson, Grant, Cleve- land, Harrison, McKinley, T. Roosevelt, Coolidge, Eisenhower, Kennedy, Johnson, Nixon, Carter, and Reagan, to varying degrees, each refused to provide requested information to Congress—including when Congress lacked what the President determined to be a valid legislative end. See id. at 762–81. In 1876, for instance, the House—attempting to embarrass President Grant for spending the hot summer at Long Beach—requested information to show “whether any executive acts or duties had been performed away from Washington, the lawfully established seat of gov- ernment.” Id. at 767. President Grant refused, emphasizing that “the Constitution did not give the House of Representatives authority to in-

8 Constitutionality of the Presidential Records Act

quire of the President where he performed his executive functions,” and that any congressional power to seek documents is “limited to information necessary for the proper discharge of its powers of legislation or im- peachment.” Id. Or consider President Cleveland, who declined to provide documents pertaining to the dismissal of a United States District Attorney (the 1886 equivalent of a United States Attorney), since “the documents related to an act (the suspension and removal of an Executive Branch official) which was exclusively a discretionary executive function.” Id. Upon leaving office, Presidents throughout the nineteenth and twentieth centuries continued to exercise total dominion over even their official papers. See Nixon, 978 F.2d at 1277. President Franklin Roosevelt, for one, commented that “every President since Washington had regarded their presidential files as their personal property and had always taken them from the White House at the expiration of their terms.” Id. (citation and internal quotation marks omitted); Jonathan Turley, Presidential Papers and Popular Government: The Convergence of Constitutional and Property Theory in Claims of Ownership and Control of Presidential Records, 88 Cornell L. Rev. 651, 657–66 (2003) (collecting examples); see also Folsom v. Marsh, 9 F. Cas. 342, 345 (C.C.D. Mass. 1841) (Story, J.) (“[C]ongress . . . actually purchased these very letters and manuscripts, at a great price, for the benefit of the nation, from their owner and posses- sor under the will of Mr. Justice Washington, as private and most valuable property.”). Or, as President (and later Chief Justice) Taft put it, “[t]he Executive office of the President is not a recording office. The vast amount of correspondence that goes through it . . . does not become the property or a record of the government . . . .” William Howard Taft, Our Chief Magistrate and His Powers 34 (1916). Until President Nixon, every President treated his official documents “as private papers to hold, give away, withhold from others, transfer for consideration or bequeath as he saw fit.” Nixon, 978 F.2d at 1280. This tradition of presidential control came at a recognized, but accept- ed, cost: Historical records—even official documents of public interest— were inevitably lost. As Taft observed, “there is lost to public record some of the most interesting documents of governmental origin bearing on the history of an administration.” Taft, supra, at 34. Yet for two centuries, Congress “acquiesced in this tradition” of presidential control and its attendant consequences. Nixon, 978 F.2d at 1282. Throughout the

9 50 Op. O.L.C. __ (Apr. 1, 2026)

1800s, in lieu of legislation, “Congress routinely bargained for and purchased presidential papers for ‘fancy sums.’” Id. In the 1900s, when the Library of Congress was tasked with collecting and purchasing presi- dential papers in the twentieth century, it had to either purchase those papers outright or accept them as gifts subject to conditions. See id. at 1283. “Indeed, by statute, Congress ratified the practice by mandating that the Library of Congress comply with any restriction placed on deposited presidential papers.” Id.; see also Title to Presidential Papers— Subpoenas, 43 Op. Att’y Gen. at 13 (“[The] 1955 Presidential Libraries Act, which serves as the permanent basis of the Presidential Library system, constitute[d] clear legislative acknowledgment that a President has title to all the documents and historical materials—whether personal or official—which accumulate in the White House Office during his incumbency.”). Congress continued its recognition of presidential owner- ship when it enacted the Federal Records Act of 1950 (“FRA”), Pub. L. No. 81-754, tit. V, 64 Stat. 583. That statute “distinguished between presidential papers and other government records.” Nixon, 978 F.2d at 1283. “While the FRA made it clear that Congress regarded the ownership of agency records to be in the United States, it specifically excepted presidential materials for different treatment.” Id. (emphasis in original). In sum, starting at the Founding and continuing for the next 200 years, Congress obtained access to the President’s papers through political negotiation or interbranch accommodation, not legislative fiat. Official presidential papers were regarded as the President’s private property, subject to his exclusive control. While holding office, Presidents exer- cised discretion in deciding whether and which documents to share with Congress—including by considering whether Congress had expressed, in the President’s view, a valid legislative need for the records. No statute purported to mandate preservation or congressional access, either while holding office or after, even though Presidents throughout history often declined to fulfill Congress’s requests in full. Through this “regular course of practice,” the separation of powers principles at issue in congressional requests for information have been “liquidate[d] & settle[d]” into a process of negotiation and compromise between coordinate branches. NLRB v. Noel Canning, 573 U.S. 513, 525 (2014) (quoting Letter for Spencer Roane from James Madison (Sept. 2, 1819), in 8 The Writings of James Madison 450 (Gaillard Hunt ed.,

10 Constitutionality of the Presidential Records Act

1908)). Today, that process is known as “accommodation.” See Congres- sional Oversight at *37 (“The manner of that compliance is determined by the operation of the accommodation process mandated by the Constitu- tion, recognized by the Judicial Branch, and practiced by the Executive and Legislative Branches.”); id. at *38 (“This long-standing ‘tradition of negotiation and compromise’ stands at the heart of the accommodation process.” (citation omitted)). Accommodation requires the political branches to engage in “good faith negotiations over their respective inter- ests” when Congress requests information from the Executive Branch. Id. at *38. But that is all it requires. Until 1974, Congress never purported to regulate by statute records belonging to the head of a coordinate branch.

C.

Presidents’ complete control over their papers persisted until the Wa- tergate scandal, when Congress for the first time sought to regulate presi- dential records by statute. President Nixon resigned from office on Au- gust 9, 1974. Upon resignation, he instructed that some 42 million pages of documents and 880 tape recordings be sent to his home in California. See Nixon v. Adm’r of Gen. Servs., 433 U.S. 425, 430 (1977) (“Nixon v. Administrator”). The Watergate Special Prosecutor, however, “advised President Ford of his continuing need for the materials.” Id. at 431. Nixon and the Administrator of General Services then entered into an agreement specifying that the materials would be deposited in a government facility near Nixon’s California home. See id. The agreement also provided that “the Administrator shall destroy such tapes as Mr. Nixon may direct,” and that all remaining tapes “shall be destroyed at the time of (his) death or on September 1, 1984, whichever event shall first occur.” Id. at 432 (altera- tions and citation omitted). Congress responded shortly thereafter with the PRMPA, which was de- signed to abrogate the agreement between Nixon and the Administrator. See id. In relevant part, the PRMPA directed the Administrator to receive or retain all tape recordings made by a federal government employee that (1) involved Nixon or other federal employees, (2) were recorded in the White House or other Executive Office buildings, and (3) were recorded between Nixon’s 1969 inauguration and his resignation. See id. at 433–44. The PRMPA also made those recordings and other preserved material available in response to legal process, giving priority of access to the

11 50 Op. O.L.C. __ (Apr. 1, 2026)

Watergate Special Prosecutor, but subject to “any rights, defenses, or privileges which the Federal Government or any person may invoke.” Id. at 434 (citation omitted). Nixon challenged the constitutionality of the statute, raising separation of powers and executive privilege claims, among others. See id. at 439– 55. The Court analyzed and rejected only facial challenges to “the provi- sions of the Act requiring the Administrator to take [Nixon’s records] into the Government’s custody subject to screening by Government archi- vists.” Id. at 439. As to the separation of powers claim, the Court con- cluded that the PRMPA did not “prevent[] the Executive Branch from accomplishing its constitutionally assigned functions.” Id. at 443; see also id. at 444–45 (“Thus, whatever are the future possibilities for constitu- tional conflict . . . , nothing contained in the Act renders it unduly disrup- tive of the Executive Branch and, therefore, unconstitutional on its face.”). The Court emphasized that under the statute, the “Executive Branch remains in full control of the Presidential materials,” and that such “materials can be released only when release is not barred by some appli- cable privilege inherent in that branch.” Id. at 444. As to the executive privilege claim, the Court concluded that “adequate justifications are shown for this limited intrusion into executive confiden- tiality.” Id. at 452. The Court pointed to the Act’s legislative history, which revealed that “Congress acted to establish regular procedures to deal with the perceived need to preserve the materials for legitimate historical and governmental purposes.” Id. Here, too, the Court found it important that Nixon’s privilege claim was asserted against the current Executive Branch rather than against some other person or branch of government. “An incumbent President,” said the Court, “should not be dependent on happenstance or the whim of a prior President when he seeks access to records of past decisions . . . .” Id. Congress acted “legiti- mately” in enacting the PRMPA “to rectify the hit-or-miss approach that ha[d] characterized past attempts to protect these substantial interests.” Id. at 453. The Court also underscored the importance of “restor[ing] public confidence in our political processes by preserving the materials as a source for facilitating a full airing of the events leading to [Nixon]’s resignation and Congress’ need to understand how those political process- es had in fact operated in order to gauge the necessity for remedial legis- lation.” Id. The Court thus concluded that “the scheme adopted by Con-

12 Constitutionality of the Presidential Records Act

gress for preservation of [Nixon]’s Presidential materials cannot be said to be overbroad.” Id. at 454.

D.

One year after Nixon v. Administrator upheld the PRMPA, Congress went further and enacted the Presidential Records Act of 1978, Pub. L. No. 95-591, 92 Stat. 2523 (codified as amended at 44 U.S.C. §§ 2201– 2209). “By all accounts, before the [PRA], Presidents were never subject to any such specific, express legal duty to create or maintain their papers.” Nixon, 978 F.2d at 1276 (citation omitted). The PRA was watershed legislation, and we discuss its relevant provisions here. The Act governs the creation, retention, and disposition of all “Presi- dential records.” 44 U.S.C. § 2203. Presidential records are “documentary materials” that are “created or received by the President, the President’s immediate staff, or a unit or individual of the Executive Office of the President whose function is to advise or assist the President, in the course of conducting activities which relate to . . . the constitutional, statutory, or other official or ceremonial duties of the President.” Id. § 2201(2). Presi- dential records do not include “official records of an agency” or the Presi- dent’s “personal records.”1 Id. § 2201(2)(B). Departing from the unbroken understanding that the President’s papers are his private property, the PRA vests title to all presidential records in the United States. See id. § 2202. And it requires the creation of such records where none would otherwise exist: The President must ensure that the “the activities, deliberations, decisions, and policies that reflect the performance of the President’s constitutional, statutory, or other official or ceremonial duties are adequately documented and that such records are preserved and maintained as Presidential records.” Id. § 2203(a). While in office, the President is “exclusively responsible for custody, control, and access” to presidential records. Id. § 2203(f); see also id. § 2205 (setting forth access exceptions under which presidential records must be made available).

1 Personal records are those “of a purely private or nonpublic character which do not

relate to or have an effect upon the carrying out of the constitutional, statutory, or other official or ceremonial duties of the President.” 44 U.S.C. § 2201(3).

13 50 Op. O.L.C. __ (Apr. 1, 2026)

The Act allows the President to dispose of presidential records in cer- tain circumstances. An incumbent President may dispose of his presiden- tial records “that no longer have administrative, historical, informational, or evidentiary value,” but only if he “obtains the views, in writing, of the Archivist” on the matter, and so long as “the Archivist does not intend to take any action under” section 2203(e). Id. § 2203(c). Subsection (e) requires the Archivist to “request the advice” of certain congressional committees “with respect to any proposed disposal of Presidential rec- ords” if the Archivist believes that “these particular records may be of special interest to the Congress,” or if consultation “is in the public inter- est.” Id. § 2203(e). Subsection (e) notwithstanding, the President may dispose of records “if copies of the disposal schedule are submitted” to the relevant congressional committees at least 60 days of continuous session of Congress before the records’ destruction. Id. § 2203(d). Before leaving office, the President must decide whether documents falling into certain categories “shall be restricted” from disclosure. Id. § 2204(a). The President may, for example, restrict access to “confidential communications . . . between the President and the President’s advisers.” Id. § 2204(a)(5). But even when a President decides to withhold such documents, his section 2204 designation can shield them from disclosure for only a limited duration, “not to exceed 12 years.” Id. § 2204(a). The Act generally requires the public disclosure of all presidential rec- ords no later than 12 years after the President leaves office. Upon the President’s departure, the Archivist assumes “responsibility for the custo- dy, control, and preservation” of the President’s papers. Id. § 2203(g)(1). The Archivist must “make such records available to the public as rapidly and completely as possible,” consistent with the terms of the Act, includ- ing the provisions allowing the President to restrict certain documents from disclosure for no more than 12 years. Id. That said, an incumbent or former President may assert claims of con- stitutionally based privileges to bar public disclosure, even after the expiration of the 12-year period. See generally id. § 2208. Claims of executive privilege must “be made personally by a former President or incumbent President.” Id. § 2208(b)(1). The asserting President must then notify the Archivist and certain congressional committees of the assertion of privilege “on the same day that the claim is asserted.” Id. § 2208(b)(2). If a former President asserts a claim of privilege, the incumbent and

14 Constitutionality of the Presidential Records Act

Archivist must consult “to determine whether the incumbent President will uphold the claim asserted by the former President.” Id. § 2208(c)(1). If the incumbent President upholds the claim, the Archivist may not publicly disclose the record unless the incumbent withdraws the decision upholding the claim or a final court order directs the disclosure of the material. See id. § 2208(c)(2)(B); see also id. § 2208(d). If the incumbent does not uphold the claim of privilege, the former President may initiate an action in the U.S. District Court for the District of Columbia to “as- sert[] that a determination made by the Archivist violates the former President’s rights or privileges.” Id. § 2204(e). The former President then bears the burden of demonstrating that the relevant material is privileged. “[U]nless otherwise directed by a court order,” however, the Archivist “shall release the Presidential record” 90 days after first receiving notice of the claim of privilege. Id. § 2208(c)(2)(C). The PRA is “a significant departure from historical practice.” Mazars, 140 S. Ct. at 2031; see also Nixon, 978 F.2d at 1276; Congressional Oversight at *45–46. It stands far afield from the “tradition of negotiation and compromise” that characterized Congress’s access to the President’s papers for the first two centuries of our Nation’s history. Mazars, 140 S. Ct. at 2031. Never before or since has Congress enacted a prospective law compelling the creation, retention, and disclosure of presidential records. Rather than requesting specific information from an incumbent President in response to particular events or potential legislation, the PRA obliges the President—and all future Presidents—to create and disclose wide swaths of information, regardless of whether that information serves contemplated legislation. This capacious scheme runs headlong into the axiom that “[n]o inquiry is an end in itself; it must be related to, and in furtherance of, a legitimate task of the Congress.” Watkins v. United States, 354 U.S. 178, 187 (1957). More still, the Act allows Congress to “walk away from the bargaining table and compel compliance” rather than “negotiating over information requests” in particular cases. Mazars, 140 S. Ct. at 2034. Although the PRMPA was the first statute to obviate the give-and-take of the political process, the PRA is the first and only statute to do so on a sweeping and prospective basis. “Such a lack of historical precedent is generally a telling indication of a severe constitutional problem with the asserted power.” Trump v. Ander- son, 144 S. Ct. 662, 669 (2024) (citation and internal quotation marks

15 50 Op. O.L.C. __ (Apr. 1, 2026)

omitted). That is because it reflects the constitutional interpretations of “both the Executive Branch and Congress,” which have recognized “their respective constitutional obligations to seek accommodation through good faith negotiations over their respective interests.” Congressional Over- sight at *38. This longstanding practice “is a consideration of great weight in cases concerning the allocation of power between the two elected branches of Government.” Mazars, 140 S. Ct. at 2031 (cleaned up). As the Supreme Court has put it, “a page of history is worth a volume of logic.” New York Tr. Co. v. Eisner, 256 U.S. 345, 349 (1921) (Holmes, J., for the Court); accord Eldred v. Ashcroft, 537 U.S. 186, 200 (2003); see also Mazars, 140 S. Ct. at 2026 (explaining that the “rivalry and reciprocity” that characterizes the history of requests for presidential information “necessarily informs [the] analysis of the question”). Here, history and doctrinal logic align to confirm that Congress has never had the Article I power to regulate the President’s records in the manner contemplated by the PRA.

II.

“Every law enacted by Congress must be based on one or more of its powers enumerated in the Constitution,” Morrison, 529 U.S. at 607, or “such implied powers as are necessary and proper to carry into effect the enumerated powers,” Carter, 298 U.S. at 291; see also Congressional Authority to Adopt Legislation Establishing a National Lottery, 10 Op. O.L.C. 40, 41 (1986) (“An act of Congress therefore is invalid unless it is affirmatively authorized under the Constitution.”). The Legislature “can claim no powers which are not granted to it by the constitution, and the powers actually granted, must be such as are expressly given, or given by necessary implication.” Martin v. Hunter’s Lessee, 14 U.S. (1 Wheat.) 304, 326 (1816); see also Marbury v. Madison, 5 U.S. (1 Cranch) 137, 176 (1803). When it comes to legislation directly targeting the independ- ence and autonomy of the President, moreover, Congress bears a height- ened burden to demonstrate the basis for its action. See, e.g., Congres- sional Oversight at *2. The PRA exceeds the powers granted to Congress by the Constitution and threatens legislative aggrandizement at the expense of the Executive Branch. See Separation of Powers, 20 Op. O.L.C. at 131 (“[The Framers’] primary fears were directed toward congressional self-aggrandizement,

16 Constitutionality of the Presidential Records Act

and the Supreme Court’s decisions call for careful scrutiny of legislation that has the purpose or effect of extending Congress’s authority beyond the legislative process.” (footnote omitted)). It is not a lawful exercise of Congress’s oversight power because it is unsupported by any valid legis- lative need. For the same reason, it cannot be justified as part of Con- gress’s inherent preservation power, which is incidental to the oversight power. Nor is it supported by Congress’s implicit authority to regulate government agencies and offices created by statute, because it applies to an office created by the Constitution, not an office created by Congress. Neither is it defensible under the Appropriations Clause, which confers no independent legislative authority, nor under the Spending Clause, which does not allow Congress to regulate a coordinate branch of government directly. And it is not a valid exercise of Congress’s power to enact neces- sary and proper legislation to assist the President in his performance of the executive power because it encroaches upon rather than facilitates the exercise of his duties.

The Supreme Court has made it “plain that Congress’s formal authority is limited to the enactment of legislation and activities in aid of the legis- lative process such as investigation and oversight,” so we start there. Separation of Powers, 20 Op. O.L.C. at 171. The PRA is not a valid exercise of Congress’s “authority to investigate in furtherance of its power to legislate.” Congressional Oversight at *10. Although “Congress has no enumerated constitutional power to conduct investigations,” it “has power ‘to secure needed information’ in order to legislate.” Mazars, 140 S. Ct. at 2031 (quoting McGrain, 273 U.S. at 161); see Eastland v. U.S. Servicemen’s Fund, 421 U.S. 491, 504 (1975) (“[T]he power to investi- gate is inherent in the power to make laws.”). This power has come to be known as oversight authority, and it forms the principal mechanism by which Congress may seek to obtain records from the President. Congres- sional Oversight at *10; see also Ways and Means Committee’s Request for the Former President’s Tax Returns and Related Tax Information Pursuant to 26 U.S.C. § 6103(f)(1), 45 Op. O.L.C. __, at *19–20 (July 30, 2021). But the oversight power is “subject to several limitations.” Mazars, 140 S. Ct. at 2031. And it is particularly limited when Congress directs the oversight power against the President, because unenumerated powers

17 50 Op. O.L.C. __ (Apr. 1, 2026)

to regulate the President cannot be lightly inferred consistent with the separation of powers. See id. at 2033–35. Although the PRA might appear to exercise the oversight power by taking records that would otherwise have been subject to the President’s sole discretion and making them available to Congress, see, e.g., 44 U.S.C. §§ 2203(e)(1), 2205(2)(C), it exceeds the separation of powers limitations articulated by the Supreme Court in Trump v. Mazars.2 In Mazars, three committees of the House of Representatives issued subpoenas seeking information from third parties about President Trump’s finances. See Mazars, 140 S. Ct. at 2026. The President challenged the subpoenas as being beyond Congress’s authority, id., without asserting executive privilege over the requested records, id. at 2028. The D.C. and Second Circuits held that the subpoenas served valid legislative purposes and thus were within Congress’s investigatory power. See id. at 2028–29. The Supreme Court vacated the judgments below. See id. at 2036. In the Court’s view, “[c]ongressional subpoenas for information from the President,” even for information not shielded by executive privilege, “implicate special concerns regarding the separation of powers.” Id. Because the lower courts “did not take adequate account of those con- cerns,” the Court remanded the cases for further proceedings, consistent with separation of powers principles. Id. To start, the Court emphasized that the history of “rivalry and reci- procity” between Congress and the President with respect to presidential records “necessarily informs” whether the Legislature has exceeded the limits of its inherent investigative authority. Id. at 2026. Relying in part on testimony given to Congress by then-Assistant Attorney General

2 Although Mazars concerned the President’s personal papers, our Office has previous-

ly concluded that the separation of powers principles articulated in that opinion “guide the appropriate approach to congressional oversight requests directed at the White House” involving the President’s official records. Congressional Oversight at *15. We explained, “[a]lthough Mazars addressed a subpoena that sought the President’s personal financial information, there is no reason to think that a lesser standard would apply to oversight requests directed at the White House and its staff—requests that bear even more closely upon interests of confidentiality and the autonomy of the Executive Branch.” Id. And even when congressional requests for information do not formally implicate executive privilege, both our Office and the Supreme Court have recognized that the separation of powers principles underlying Mazars must factor into an analysis of Congress’s Article I authority. See id. at *22–23; Mazars, 140 S. Ct. at 2033–34.

18 Constitutionality of the Presidential Records Act

Antonin Scalia, the Court noted that “[h]istorically, disputes over con- gressional demands for presidential documents have not ended up in court.” Id. at 2029. “Instead, they have been hashed out in the ‘hurly- burly, the give-and-take of the political process between the legislative and the executive.’” Id. (citation omitted); see also, e.g., id. at 2029–31 (collecting historical examples). The Court feared that giving Congress “a limitless subpoena power would transform the ‘established practice’ of the political branches” and that “[i]nstead of negotiating over information requests, Congress could simply walk away from the bargaining table.” Id. at 2034 (citation omitted). The Court then made plain that Congress faces a heightened burden when it invokes its inherent investigative power to seek the President’s papers in particular. The Court rejected the House’s proffered standard for evaluating the validity of its subpoenas: that they “relate[d] to a valid legislative purpose” or “concern[ed] a subject on which legislation could be had.” Id. at 2033 (citations omitted). Although these formulations apply in requests for information that do not seek the President’s papers, the House’s test “fail[ed] to take adequate account of the significant separation of powers issues” that attend a request “for the President’s information.” Id. Nearly all presidential papers “could potentially ‘relate to’ a conceivable subject of legislation,” said the Court, so without more stringent limits on its investigative power, “Congress could ‘exert an imperious controul’ over the Executive Branch and aggrandize itself at the President’s expense.” Id. at 2034 (quoting The Federalist No. 71, at 484 (Alexander Hamilton) (Jacob E. Cooke ed., 1961)). That concern remains even when the Legislature seeks information that is not protected by executive privilege. See id. at 2034–35. Indeed, “there is not always a clear line between [the President’s] personal and official affairs,” and “congressional demands for the President’s papers can implicate the relationship between the branches regardless [of] whether those papers are personal or official.” Id. at 2034. At the same time, Congress need not show that the requested information is “‘demonstrably critical’ to its legislative purpose” when it seeks information not protected by executive privilege. Id. at 2032. Applying such a “demanding standard” would frustrate Congress in fulfilling its valid legislative responsibilities. Id. Thus, to assess whether a request for the President’s papers is “related to, and in furtherance of, a legitimate task of the Congress,” we “must

19 50 Op. O.L.C. __ (Apr. 1, 2026)

perform a careful analysis” that balances the Legislature’s interests against “the unique position of the President.” Id. at 2035 (citations and internal quotation marks omitted). Four factors guide this analysis, which we consider and apply in a sequence that best illuminates the PRA’s constitutional infirmity: (1) whether there is a valid legislative purpose; (2) whether that legislative purpose can be adequately served by collect- ing information from sources other than the President’s papers; (3) whether the Act is appropriately tailored to that legislative purpose; and (4) whether the Act imposes an improper burden on the Executive. First, and perhaps most importantly, there is scant evidence of any valid legislative purpose for targeting the records subject to the PRA. See Mazars, 140 S. Ct. at 2036 (“The more detailed and substantial the evi- dence of Congress’s legislative purpose, the better.”). Indeed, when it enacted the PRA, Congress did not identify any contemplated legislation that it could better evaluate with access to the President’s papers. And we are aware of none today that would rescue the Act from invalidation. Cf. Shelby County v. Holder, 570 U.S. 529, 550–51 (2013) (concluding that “a statute’s ‘current burdens’ must be justified by ‘current needs’” and therefore that the Voting Rights Act’s coverage formula was no longer supported by a valid legislative purpose (citation omitted)). The PRA’s requirements are temporally unfettered, always binding the President regardless of whether Congress is actively considering legislation in- formed by the records subject to the Act. Indeed, Congress’s choice of vehicle—a forward-looking law rather than a contemporaneous sub- poena—illustrates the disconnect between the statute and a valid legisla- tive purpose. Should the President’s papers shed light on legislation under consideration, Congress can employ a more surgical subpoena or a nar- rowly tailored statute rather than the blunt instrument of the PRA. See Mazars, 140 S. Ct. at 2036 (“The specificity of the subpoena’s request ‘serves as an important safeguard against unnecessary intrusion into the operation of the Office of the President.’” (citation omitted)). Because any possible legislation that directly involves presidential ac- tivities is likely to raise “sensitive constitutional issues,” id., Congress’s “duty . . . to justify its requests” is at its apex, Congressional Requests, 13 Op. O.L.C. at 159; see also Congressional Oversight at *13 (“Alt- hough the Executive Branch should seek to accommodate legitimate requests for information concerning the departments and agencies, this

20 Constitutionality of the Presidential Records Act

Office has advised that such accommodation may not be required where congressional committees’ requests appear to fall outside their delegated legislative jurisdiction or lack a legitimate legislative purpose.”). The PRA mandates that the President create and preserve records of his most sensitive deliberations—including decisions concerning his constitutional and statutory duties, communications with advisors, and internal Execu- tive Branch discussions. See 44 U.S.C. § 2203(a). But “Congress will seldom have any legitimate legislative interest in knowing the precise predecisional positions and statements of particular executive branch officials” embodied in these materials. Congressional Requests, 13 Op. O.L.C. at 159. If Congress invokes the PRA to access such materials in furtherance of legislation, that legislation would necessarily “concern[] the Presidency” and might even seek to regulate the Presidency outright. Mazars, 140 S. Ct. at 2036. For that reason, our Office and the Supreme Court have both empha- sized that when Congress seeks records from the President, “it must explain its need carefully and convincingly.” Congressional Requests, 13 Op. O.L.C. at 159; see Mazars, 140 S. Ct. at 2036 (noting that when Congress legislates vis-à-vis the President, “it is ‘impossible’ to conclude that a [request] is designed to advance a valid legislative purpose unless Congress adequately identifies its aims and explains why the President’s information will advance its consideration of the possible legislation” (citation omitted)); Congressional Oversight at *15 (“Congress may be expected to clearly articulate its legislative purpose, and the Executive Branch may independently review the proffered purpose.”). Congress has not done so here. “Without such an explanation,” it is “impossible to assess the needs of one branch and relate them to those of the other.” Congressional Requests, 13 Op. O.L.C. at 159. If Congress “has a reason for needing to obtain” the President’s papers, “it should be able to express it.” Id. The Act’s legislative history only underscores that the PRA is untethered from any valid legislative purpose. The PRA—which Congress understood “would terminate the tradition of private ownership of Presi- dential papers”—was enacted to ensure “the preservation and public availability of these records at the end of a Presidential administration.” H.R. Rep. No. 95-1487, at 2 (1978). But as the Court has reiterated time and again, “there is no congressional power to expose for the sake of exposure.” Watkins, 354 U.S. at 200; accord Mazars, 140 S. Ct. at 2032.

21 50 Op. O.L.C. __ (Apr. 1, 2026)

The public’s interest in the President’s papers does not confer upon Con- gress a power that the Constitution does not. And as we detail below, other congressional powers—such as an inherent preservation power or the power to establish Executive Branch agencies and offices—cannot save the PRA. Second, even if the PRA advanced some valid legislative purpose, Congress cannot carry its burden of showing that “other sources” are inadequate to “reasonably provide [it] the information it needs.” Mazars, 140 S. Ct. at 2035–36. “Unlike in criminal proceedings, where the very integrity of the judicial system would be undermined without full disclo- sure of all the facts, efforts to craft legislation . . . are not hampered in quite the same way when every scrap of potentially relevant evidence is not available.” Id. at 2036 (cleaned up). We have explained that, “because Congress may not conduct oversight of the President’s exclusive constitu- tional functions, legitimate congressional oversight inquiries will almost always pertain to executive branch implementation of statutory pro- grams.” Congressional Oversight at *26–27. “But the departments and agencies, not the White House, principally administer such programs, and thus it is generally unnecessary for congressional committees to request information directly from the White House unless they are unable to obtain the information from agencies.” Id. at *27. Here, the PRA requires us to assume that even if Congress someday undertook an investigation with a valid legislative purpose, presumptive access to the President’s papers would be the only reasonable means it would have of obtaining information material to that investigation. This layered speculation cannot “warrant[] the significant step of involving the President and his papers.” Mazars, 140 S. Ct. at 2035. Stated differently, the PRA relieves Congress of its obligation to demonstrate that the President’s papers would offer necessary insight into the subject of contemplated legislation that is unavailable from other sources. Third, the statute is far “broader than reasonably necessary.” Id. at 2036; cf. Nixon v. Administrator, 433 U.S. at 454 (considering whether the PRMPA was “overbroad”). The Act is indiscriminate, collecting and publicly disclosing all “materials relating to the political activities of the President or members of the President’s staff” that pertain to “the carrying out of constitutional, statutory, or other official or ceremonial duties of the President,” subject to certain limited exceptions. 44 U.S.C.

22 Constitutionality of the Presidential Records Act

§ 2201(2)(A). Far from targeting “precisely identified” and “specifically enumerated” documents, the PRA “ask[s] for everything under the sky.” Cheney, 542 U.S. at 387 (cleaned up). The PRA thus inverts the constitu- tional order of operations. Rather than identifying a valid legislative need and seeking relevant information with specificity, Congress created a capacious dragnet for all presidential records, some of which might someday prove relevant to a valid legislative purpose. The statute is overbroad for yet another reason. “Because Congress may not legislate with respect to the President’s discharge of his exclusive constitutional functions, it similarly may not seek information from White House staff concerning the decision-making process in connection with the President’s performance of those functions in particular matters.” Congressional Oversight at *16; see, e.g., Assertion of Executive Privi- lege with Respect to Clemency Decision, 23 Op. O.L.C. 1, 3–4 (1999) (Reno, Att’y Gen.) (“[I]t appears that Congress’ oversight authority does not extend to the process employed in connection with a particular clem- ency decision, to the materials generated or the discussions that took place as part of that process, or to the advice or views the President received in connection with a clemency decision.”). The President wields the exclu- sive power to issue pardons, to appoint and remove officers of the United States, to recognize foreign countries, to sign or veto legislation, to con- duct foreign diplomacy and national defense, 3 and so on. See Congres- sional Oversight at *18–19; Trump, 144 S. Ct. at 2327–28. But the PRA requires the President to create and preserve records of all “activities, deliberations, decisions, and policies that reflect the performance of the President’s constitutional, statutory, or other official or ceremonial du- ties”—including those that pertain to the President’s exclusive functions. 44 U.S.C. § 2203(a). Because “Congress may not conduct oversight of the President’s discharge of his exclusive constitutional authority,” Congres- sional Oversight at *16, the PRA sweeps well beyond any valid legisla- tive purpose Congress could claim. And even if some provisions of the Act were appropriately tailored to serve a valid legislative purpose, the public-disclosure provisions do not

3 Although the “President’s exclusive authorities also include his powers in the area of

diplomacy and national defense,” we acknowledge that “in many cases those powers closely abut areas in which Congress may legislate.” Congressional Oversight at *19.

23 50 Op. O.L.C. __ (Apr. 1, 2026)

further Congress’s investigatory power. The PRA requires the Archivist to make “records available to the public as rapidly and completely as possi- ble consistent with the provisions of [the rest of the Act].” 44 U.S.C. § 2203(g)(1). Subject to a handful of exceptions, those records must be made public within twelve years of the President leaving office. See id. § 2204(a). These provisions have nothing at all to do with Congress’s investigative powers, because Congress could obtain specific evidence from the President in the course of lawmaking without making that infor- mation public. The Act’s impermissibly broad sweep is thrown into even sharper relief by comparison to the constitutional accommodation process. Once Con- gress has identified an appropriate subject on which legislation may be had, it can request information from the President, who will evaluate that request in good faith. “An important feature of the accommodation pro- cess is the dialogue that takes place between the committee and the White House to ensure that information requests are not ‘unnecessarily broad.’” Congressional Oversight at *42 (quoting Cheney, 542 U.S. at 390). The “specificity” of a request is “an important safeguard against unnecessary intrusion” into the Presidency. Cheney, 542 U.S. at 387; accord Mazars, 140 S. Ct. at 2036. And that process is better suited to separation of pow- ers review, because Congress can supply “detailed and substantial” evi- dence of legislative need. Mazars, 140 S. Ct. at 2036. And if Congress employs a congressional subpoena against a President, it is subject to all of the limitations placed on the invocation of that authority by Mazars. The accommodation process is a scalpel; the PRA is a sledgehammer. Fourth, we consider “the burdens imposed on the President” by the Act. Id. We understand that White House Counsel’s Office spends a considerable amount of time ensuring the President’s compliance with the PRA, which detracts from its capacity to advise the President on sensitive questions of law and policy. This illustrates the risk that, although the Act may not place a severe and direct burden on the President personally, it may still force the President’s advisers “to divert time and attention from their duties to the President,” threatening “significant congressional encroachment on, and interference with, the President’s prerogatives and his ability to discharge his duties with the advice and assistance of his closest advisers.” Testimonial Immunity, 43 Op. O.L.C. at 112 (citation and internal quotation marks omitted); see Assertion of Executive Privi-

24 Constitutionality of the Presidential Records Act

lege Regarding White House Counsel’s Office Documents, 20 Op. O.L.C. 2, 3 (1996) (“Impairing the ability of the Counsel’s Office to perform its important functions for the President would in turn impair the ability of you and future Presidents to carry out your constitutional responsibili- ties.”). Perhaps more importantly, the mere possibility that the “activities, deliberations, decisions, and policies” pertaining to the President’s official duties may become public, 44 U.S.C. § 2203(a), threatens to “chill presi- dential advisers from providing unpopular advice,” impairing the sound functioning of the Presidency, Testimonial Immunity, 43 Op. O.L.C. at 114 (citation omitted). “This is true whether or not the President might ultimately assert executive privilege” over the documents. Id. We discuss this point in greater detail below, see infra Part III.B, but we need not dwell on it here, since the first three Mazars factors support our conclu- sion independently. We acknowledge that some aspects of the PRA might mitigate the se- verity of Congress’s encroachment. For one, the PRA’s requirement to create presidential records where they otherwise would not exist gives the President considerable discretion in determining what steps are “neces- sary” to assure that his performance of official duties is “adequately documented.” 44 U.S.C. § 2203(a). And the PRA allows an incumbent or former President to assert executive privilege and therefore shield the most sensitive communications from disclosure. See id. § 2208. But these facts do not substitute for a valid legislative purpose, which is necessary to protect the Executive Branch against congressional self- aggrandizement of any degree. “Executive branch lawyers . . . have a constitutional obligation” to resist “inadvertent or intentional congres- sional intrusion,” no matter how seemingly benign. Separation of Powers, 20 Op. O.L.C. at 126. Even “minor (but unconstitutional) aggrandize- ments” can incrementally alter the balance of power between the depart- ments of government. Id. at 132. Encroachments operate as a one-way ratchet—what the Executive tolerates today becomes a baseline for fur- ther incursions tomorrow, slowly but inexorably altering the structure the Constitution establishes. See Congressional Oversight at *22 (“If Con- gress could freely demand the President’s information, it would exert an imperious controul over the Executive Branch and aggrandize itself at the President’s expense, just as the Framers feared.” (citations and inter- nal quotation marks omitted)). Put simply, “[t]o acquiesce in legislation

25 50 Op. O.L.C. __ (Apr. 1, 2026)

encroaching upon the executive authority results in the establishment of dangerous precedents.” Authority of Congressional Committees to Disap- prove Action of Executive Branch, 41 Op. Att’y Gen. 230, 233 (1955). The PRA itself illustrates this dynamic. What was once a novel asser- tion of congressional authority—one that stood in stark contrast to nearly two hundred years of history—has come to seem ordinary and uncontro- versial, making it harder to recognize the constitutional anomaly for what it is. The seeming acquiescence of the Executive Branch since the PRA’s enactment may be motivated by nothing more than discretionary choices to avoid an interbranch conflict. But, over time, it has subtly reshaped how the political branches and the public perceive the separation of pow- ers. Cf. The Declaration of Independence para. 2 (U.S. 1776) (“[A]ll experience hath shewn, that mankind are more disposed to suffer, while evils are sufferable, than to right themselves by abolishing the forms to which they are accustomed.”). This shift has already gone so far that attempts have been made to subject a former President to criminal liability for his handling of presidential records that, for most of this Nation’s history, would have been subject to his complete discretion. See, e.g., United States v. Trump, 739 F. Supp. 3d 1131, 1137 (S.D. Fla. 2024). Despite these shifting expectations, executive forbearance cannot vali- date what the Constitution prohibits, nor can it estop future Presidents from vindicating structural principles that belong not to any one office- holder but to the office itself. See Letter for Robert G. Damus, General Counsel, Office of Management and Budget, from Walter Dellinger, Assistant Attorney General, Office of Legal Counsel, at 1 (Sept. 16, 1994) (“Because of the structural nature of this [separation of powers] principle, even Presidential acquiescence cannot legitimize such an otherwise- unconstitutional assignment of executive functions to a legislative agent.”); Metro. Wash. Airports Auth. v. Citizens for the Abatement of Aircraft Noise Inc., 501 U.S. 252, 276–77 (1991) (declining to uphold a statute that provided “a blueprint for extensive expansion of the legisla- tive power beyond its constitutionally confined role” despite a “practical accommodation between the Legislature and the Executive”).

We have also considered the argument that Congress has an inherent preservation power incidental to its oversight power. Cf. H.R. Rep.

26 Constitutionality of the Presidential Records Act

No. 95-1487, at 4 (“[T]he materials would be accessible to an incumbent President or the Congress when not otherwise available and necessary to conduct the ongoing business of Government; and would also be accessi- ble under demand of subpoena or other judicial process.”). Like the power to conduct investigations, the power to preserve documents is not enu- merated in the Constitution. See Mazars, 140 S. Ct. at 2031 (“Congress has no enumerated constitutional power to conduct investigations or issue subpoenas, but we have held that each House has power ‘to secure needed information’ in order to legislate.” (citation omitted)). Thus, Congress may exercise a preservation power only as an auxiliary to its legislative functions; Congress has no power to preserve documents simply for the sake of historical preservation rather than in furtherance of a valid legisla- tive purpose. Cf. Congressional Oversight at *12 n.6 (“The Supreme Court has made clear that Congress may only investigate into those areas in which it may potentially legislate or appropriate, and transmitting information to inform the public is not a part of the legislative function.” (cleaned up)). When Congress properly invokes its oversight power (such as by issuing a subpoena under appropriate circumstances), it may per- haps temporarily foreclose the destruction of the evidence sought. But this order of operations is constitutionally significant. Congress may not fish through the President’s papers in the hope that such an expedition will produce a justification it presently lacks. Any such preservation power is therefore quite limited: Because it is incidental to the oversight power, which itself is incidental to the legislative power, it “cannot be [a] great substantive and independent power[].” Id. at *26 (citation and internal quotation marks omitted). It follows that Congress does not possess a roving preservation power simply because it might someday find some preserved information rele- vant to its lawmaking. Adopting such a theory would require us to “pile inference upon inference” to sustain the PRA. United States v. Lopez, 514 U.S. 549, 567 (1995). We would have to assume that Congress will some- day undertake an investigation that will serve a valid legislative purpose, that presidential records will be relevant to a proposed law’s considera- tion, that presidential records will be material to the bill’s consideration, that the President would otherwise destroy those records in the interim, and that Congress could not act expediently to enjoin the President from doing so. Even if it is theoretically possible that this scenario will come to

27 50 Op. O.L.C. __ (Apr. 1, 2026)

pass, the link between the PRA and Congress’s preservation interest is simply too attenuated. See id. at 567–68; cf. United States v. Comstock, 560 U.S. 126, 146 (2010) (suggesting that a statute may be unconstitu- tional if the “links between” its provisions “and an enumerated Article I power are . . . too attenuated”). More fundamentally, such a preservation theory also vitiates the pre- sumption of regularity and legitimacy that attends presidential action. See Martin v. Mott, 25 U.S. (12 Wheat.) 19, 33 (1827) (“Every public officer is presumed to act in obedience to his duty, until the contrary is shown; and, a fortiori, this presumption ought to be favourably applied to the chief magistrate of the Union.”); see also Nat’l Archives & Recs. Admin. v. Favish, 541 U.S. 157, 174–75 (2004) (applying “a presumption of legitimacy” in a Freedom of Information Act case about document disclo- sure). The presumption prevents one branch of government from “attrib- uting bad faith to an officer of a coordinate branch” based on little more than insinuation or speculation. Dep’t of Com. v. New York, 139 S. Ct. 2551, 2583 (2019) (Thomas, J., concurring in part and dissenting in part). It “is driven by separation-of-powers concerns, which increase” as Con- gress or the Judiciary “venture closer to core executive activity.” Conley v. United States, 5 F.4th 781, 791 (7th Cir. 2021); see also United States v. Armstrong, 517 U.S. 456, 465–67 (1996). Where “any doubt” may exist about whether the President will act in bad faith, “the presumption of regularity resolves it.” Dep’t of Com., 139 S. Ct. at 2582–83 (Thomas, J., concurring in part and dissenting in part). The presumption of regularity has particular purchase in the context of congressional requests for information. The Constitution requires the Executive Branch and Congress “to seek accommodation” of document requests “through good faith negotiations.” Congressional Oversight at *38. Consistent with the President’s oath to support and defend the Con- stitution, it has been “long-standing executive branch policy that upon receipt of an authorized oversight request that is in furtherance of a legit- imate legislative purpose,” the White House will enter “good faith negoti- ations” with Congress to attempt to accommodate the request. Id. at *37 (citation omitted); see, e.g., Testimonial Immunity, 43 Op. O.L.C. at 119– 20. The PRA inverts the presumption of regularity by presuming instead that Presidents might destroy presidential records to obstruct Congress’s legitimate legislative ends and thwart the accommodation process. And

28 Constitutionality of the Presidential Records Act

from the warped assumption that a future President cannot be trusted to act in good faith, the PRA requires that all presidential records—those created today and in perpetuity—must be preserved. To sustain the PRA on this basis would “display[] gross disrespect to the President”—and to the Presidency itself. Utah v. Evans, 536 U.S. 452, 511 (2002) (Scalia, J., dissenting). Here, too, history is relevant. For the better part of two centuries, Pres- idents exerted total control over their papers, both while holding office and after leaving it. They enjoyed unfettered discretion to dispose of official and personal documents as they saw fit, though Presidents did not exercise that discretion for the purpose of frustrating Congress; rather, they uniformly engaged with Congress in good-faith negotiations over access to presidential documents. At the same time, an inescapable cost of this system was that “there [wa]s lost to public record some of the most interesting documents of governmental origin bearing on the history of an administration.” Taft, supra, at 34. This loss was not because Presidents acted in bad faith to thwart Congress’s legislative efforts, but for any number of innocuous reasons, such as a good-faith assertion of executive privilege, the absence of a valid legislative purpose, a determination that the requested materials could be obtained from other sources, the dissipa- tion of the need for such documents, and so on. Yet until the PRMPA, Congress did not require the preservation of presidential documents for even a limited period of time, nor did it require the President to provide notice of his intent to destroy such documents. That Congress simply accepted this loss for nearly two hundred years—without claiming any power to preserve, access, or even receive notice about presidential rec- ords—suggests that the Founders left such control to the Executive alone. And even if Congress could exercise a preservation power with respect to presidential documents before identifying a valid and relevant legisla- tive purpose, that power would justify only a narrow sliver of the PRA. The Act contains a requirement that the Archivist notify Congress before the President destroys documents that “may be of special interest to the Congress.” 44 U.S.C. § 2203(e)(1).4 The President may nevertheless

4 While in office, “the President may dispose of those Presidential records of such

President that no longer have administrative, historical, informational, or evidentiary value . . . .” 44 U.S.C. § 2203(c). Before doing so, the President must obtain the written views of the Archivist on the proposed disposal, see id. § 2203(c)(1), and the Archivist

29 50 Op. O.L.C. __ (Apr. 1, 2026)

dispose of those documents after providing a disposal schedule to Con- gress. See id. § 2203(d). Although these provisions give “neither the Archivist nor the Congress the authority to veto the President’s decision to destroy the records,” Armstrong v. Bush, 924 F.2d 282, 286 (D.C. Cir. 1991), Congress could theoretically subpoena—and therefore prevent the destruction of—documents that are necessary to the consideration of possible legislation. As we explain below, however, even if this narrow provision of the Act were a valid exercise of the preservation power, it is inseverable from the unconstitutional provisions of the PRA.

The PRA also exceeds Congress’s authority to regulate agencies and offices that it has created. Cf. Buckley v. Valeo, 424 U.S. 1, 138 (1976) (per curiam) (“Congress may undoubtedly under the Necessary and Prop- er Clause create ‘offices’ in the generic sense and provide such method of appointment to those ‘offices’ as it chooses.”). Congress’s greater power to create and abolish statutory agencies and offices might potentially include the lesser power to regulate their handling of records. Cf. Nixon v. Administrator, 433 U.S. at 445 (describing congressional regulations of records belonging to Executive Branch agencies as “statutory precedent” for regulation of “documents in the possession of the Executive Branch”).5 But see City of Lakewood v. Plain Dealer Publ’g Co., 486 U.S. 750, 762–63 (1988) (“The key to the dissent’s analysis is its ‘greater- includes-the-lesser’ syllogism. But that syllogism is blind to the radically

must certify that he or she “does not intend to take any action under subsection (e)” of section 2203, id. § 2203(c)(2). Subsection (e) allows the Archivist to “request the advice” of certain congressional committees “with respect to any proposed disposal of Presidential records” if the Archivist believes that the “records may be of special interest to the Congress,” or if “consultation with the Congress regarding the disposal of these particular records is in the public interest.” Id. § 2203(e). But neither Congress nor the Archivist may block the disposal of documents designated for destruction by the President except, perhaps, by a subpoena. If the Archivist invokes subsection (e)’s notification provisions, the President still may dispose of the documents “if copies of the disposal schedule are submitted to the appropriate Congressional Committees” by the statutory deadline. Id. § 2203(d). 5 Chief Justice Burger’s dissent distinguished this prior legislation, noting that Con-

gress’s authority to “provide for access to records of the Executive Departments which Congress itself created” was not at issue in Nixon v. Administrator. 433 U.S. at 513.

30 Constitutionality of the Presidential Records Act

different constitutional harms inherent in the ‘greater’ and ‘lesser’ re- strictions.”). But any such power would not extend to constitutional offices that Congress did not create and cannot abolish. The Presidency is a constitutional office that does not owe its existence to Congress and therefore is not subject to the same congressional authori- ty as a statutory office. See U.S. Const. art. II, § 1, cl. 1; see also Trump, 144 S. Ct. at 2327. Thus, the first Congress drew a distinction between regulating presidential papers and regulating the papers of executive departments when it passed the Records Act of 1789. See Act of July 27, 1789, ch. 14, § 7, 1 Stat. 68, 69 (“Records Act of 1789”). The law gov- erned the retention and transmission of “books, records and papers” that “appertain[ed] to the Treasury department, or War department,” subject to the President’s direction. Id. That “housekeeping statute” “spell[ed] out the authority for executive officials to set up offices and file Government documents.” Chrysler Corp. v. Brown, 441 U.S. 281, 309 & n.39 (1979) (citation omitted). Yet this framework for maintaining Executive Branch records tellingly did not regulate the President’s papers; it regulated only the records and papers of executive agencies, and even then, the disposi- tion of those documents was subject to the direction of the President. Records Act of 1789 § 7, 1 Stat. at 69 (“[The] records and papers, as may appertain to the Treasury department, or War department, shall be deliv- ered over to the principal officers in the said departments respectively, as the President of the United States shall direct.”). Later amendments to this statute confirmed that this exclusion was a feature rather than a bug when it continued to omit the President’s papers. See, e.g., 4 Rev. Stat. § 161 (2d ed. 1878); see also, e.g., 5 U.S.C. § 22 (1958); 5 U.S.C. § 301 (1970). The same reasoning applies to the President’s “immediate advisors,” who “are an extension of the President” himself and act as “the Presi- dent’s alter ego.” Testimonial Immunity, 43 Op. O.L.C. at 112 (citation omitted). The President’s advisers are “constitutionally distinct from the heads of executive departments and agencies, whose offices are created by acts of Congress, whose appointments require the Senate’s advice and consent, and whose responsibilities entail the administration of federal statutes.” Id. at 111. Presidential advisers “exercise no statutory authority and instead act solely to advise and assist the President.” Id. But the PRA reaches all of them. It applies to documents “created or received by the President, the President’s immediate staff, or a unit or individual of the

31 50 Op. O.L.C. __ (Apr. 1, 2026)

Executive Office of the President whose function is to advise or assist the President.” 44 U.S.C. § 2201(2). These positions exist by operation of the Constitution, not by the grace of Congress. Congress did, of course, create the Executive Office of the President (“EOP”). See Congressional Oversight at *4–7. But the EOP’s creation simply formalized the mechanics of the President receiving “advice and assistance from individuals other than department and agency heads,” a practice that existed “[l]ong before the EOP’s establishment” and that was necessary for the President to carry out his constitutional duties. Id. at *5; see also Ass’n of Am. Physicians & Surgeons, Inc. v. Clinton, 997 F.2d 898, 909 (D.C. Cir. 1993) (“Article II not only gives the President the ability to consult with his advisers confidentially, but also, as a corollary, it gives him the flexibility to organize his advisers and seek advice from them as he wishes.”); Testimonial Immunity, 43 Op. O.L.C. at 111–12 (confirming that the President’s immediate advisers are not dependent on statutory authorization). Thus, as “the White House developed as an organization, all three branches of government recognized that it should be viewed differently from the departments and agencies of the Executive Branch.” Congressional Oversight at *7. For example, even though the Freedom of Information Act does not exempt certain EOP components from its coverage, “the Supreme Court held that Congress did not include the President’s immediate personal staff or units in the Executive Office whose sole function is to advise and assist the President.” Id. at *8 (inter- nal quotation marks omitted). That Congress created the EOP therefore does not rescue the PRA. The Act covers the President and his advisers, none of whose positions depend on the EOP for their existence.

Neither is the PRA necessary and proper to Congress’s funding of the Presidency, even though presidential records are created using funds appropriated by Congress. Some might defend the PRA as an exercise of Congress’s power of the purse. Congress paid for the pens and papers—or more recently, keyboards, monitors, and servers—used to create presiden- tial records, and so it is necessary and proper for Congress to assure itself, by dictating the preservation of records it paid for, that the President and his staff are not misusing funds. Just as Congress can criminally proscribe bribes “to safeguard the integrity of the state, local, and tribal recipients

32 Constitutionality of the Presidential Records Act

of federal dollars,” Sabri v. United States, 541 U.S. 600, 605 (2004), the argument goes, so too can it require the preservation of presidential rec- ords to safeguard the integrity of the presidential use of federal dollars. This argument fails because the exercise of Congress’s appropriations power must be necessary and proper to effectuate Congress’s other enu- merated powers; appropriations to the other branches are not themselves a source of congressional power. The Spending Clause fares no better, as it is not a source of power to regulate the other branches. If the PRA is justified either as a congressional restriction on appropriations or as a Spending Clause law, such justification would swallow the limits on Congress’s enumerated powers and subordinate the coordinate branches to the Legislature. Start with the Appropriations Clause. U.S. Const. art. I, § 9, cl. 7. That provision does not supply any independent source of congressional power. “To be sure, the Appropriations Clause presupposes Congress’ powers over the purse.” CFPB v. Cmty. Fin. Servs. Ass’n of Am., Ltd., 144 S. Ct. 1474, 1488 (2024). But the Clause’s “phrasing and location in the Consti- tution make clear that it is not itself the source of those powers.” Id.; see also Cincinnati Soap Co. v. United States, 301 U.S. 308, 321 (1937) (“It means simply that no money can be paid out of the Treasury unless it has been appropriated by an act of Congress.”). After all, it is “phrased as a limitation,” not a grant of authority, and “it is placed within a section of other such limitations.” CFPB, 144 S. Ct. at 1488; see also Kate Stith, Congress’ Power of the Purse, 97 Yale L.J. 1343, 1349–50 (1988) (“[The Appropriations Clause] is not a grant of affirmative power, or an expan- sion of other congressional powers, but is, rather, a condition or limitation on the exercise of legislative power.” (footnotes omitted)). Congress may provide appropriations, therefore, only to the extent necessary and proper to ensure the exercise of another enumerated power. And “Congress has no more authority to control the executive branch through appropriations than it would under other provisions of the Constitution.” Edwin Meese III, Panel IV: The Appropriations Power and the Necessary and Proper Clause, 68 Wash. U. L. Q. 623, 640 (1990) (statement of Geoffrey Miller). Likewise, “[t]he Constitution has no ‘Spending Clause,’ strictly speak- ing.” Medina v. Planned Parenthood S. Atl., 145 S. Ct. 2219, 2230 (2025). The so-called “Spending Clause” more precisely refers to Arti-

33 50 Op. O.L.C. __ (Apr. 1, 2026)

cle I, section 8, clause 1, which grants Congress the “Power To lay and collect Taxes, Duties, Imposts and Excises, to pay the Debts and provide for the common Defence and general Welfare of the United States[.]” Unlike other powers enumerated in that same section, this clause “does not expressly endow Congress with the power to regulate conduct.” Medi- na, 145 S. Ct. at 2230. When Congress attaches conditions to its spend- ing—for instance, through grants to states or private entities—these conditions are permissible because “spending-power legislation is ‘in the nature of a contract.’” Id. at 2234 (citation omitted). “For that reason, the legitimacy of Congress’ power to enact Spending Clause legislation rests not on its sovereign authority to enact binding laws, but on whether the recipient voluntarily and knowingly accepts the terms of that contract.” Cummings v. Premier Rehab Keller, P.L.L.C., 142 S. Ct. 1562, 1570 (2022) (cleaned up). To be sure, once Congress has validly invoked its spending power and a recipient has accepted federal funds, the Necessary and Proper Clause allows Congress “to see to it that taxpayer dollars appropriated under that power are in fact spent for the general welfare,” including perhaps by regulating third parties. Sabri, 541 U.S. at 605. But see Medina, 145 S. Ct. at 2243 (Thomas, J., concurring) (“Congress’s spending power is the power to spend only and does not carry with it any independent regulatory authority.” (citation and internal quotation marks omitted)). “But nothing in Article I, section eight, clause one endows Congress with a power to regulate” in the first instance, “for if it did, the enumeration of specific powers elsewhere in Article I would be rendered largely pointless.” Medina, 145 S. Ct. at 2231 (citation and internal quota- tion marks omitted). Congressional appropriations for the operations of a coordinate branch, whether the Executive or the Judiciary, are not an exercise of such spend- ing power; they are instead simply necessary and proper to carry out activities authorized by Article I. See Stith, supra, at 1348 (“Congress’ power to appropriate originates in article I, section 8. The [Necessary and Proper Clause] . . . includes the power to spend public funds on author- ized federal activities.”). 6 Spending Clause legislation usually effectuates Congress’s power to induce voluntary behavior through conditional grants—not direct regulation—and is almost always directed at states,

6 Nor does Congress’s authority to take actions necessary and proper in facilitating the

powers of a coordinate branch support the PRA, as explained in Part II.E below.

34 Constitutionality of the Presidential Records Act

tribes, and private parties. Appropriations, by contrast, allow the President and Judiciary to draw money from the Treasury for funding the operation of government and are not contractual relationships between Congress and the other branches. See Gillian E. Metzger, Taking Appropriations Seri- ously, 121 Colum. L. Rev. 1075, 1085 (2021) (“[S]pending is the term generally applied to grants of funds outside of the federal government, especially to state and local governments or private actors, whereas appropriations is used to refer to funding the federal government.”); compare, e.g., CFPB, 144 S. Ct. at 1480–86 (describing appropriations), with Medina, 145 S. Ct. at 2230–34 (describing congressional spending).7 Because the execution of federal law depends on appropriations, treating appropriations as an exercise of the spending power would nullify the limits on Congress’s enumerated powers. If appropriations constituted Spending Clause authority, every regulation promulgated by an agency would be traceable to an exercise of the spending power and thereby slip the bounds of limited government. Congress could create, for example, a Department of Intrastate Non-Economic Regulation for the General Welfare, bypassing the limits on the Commerce Clause simply by direct- ing the new agency to promulgate binding regulations using appropriated funds. Under such logic, “the Nation would trade a limited federal gov- ernment for an unlimited one.” Medina, 145 S. Ct. at 2231 (citation and internal quotation marks omitted). The PRA is not an exercise of the spending power, and thus it cannot be necessary and proper to Congress’s allocation of funds. The PRA is a direct regulation of the President and his closest advisers. It sweeps in all “documentary materials” created or received by the President and his immediate staff that pertain to the “constitutional, statutory, or other

7 Spending Clause legislation and appropriations may each contain conditions. In the

Spending Clause context, for example, Congress will often condition a state’s receipt of funds upon that state taking a particular action, such as adopting a minimum drinking age in exchange for highway funding. See, e.g., South Dakota v. Dole, 483 U.S. 203 (1987). Typical Spending Clause legislation will also limit use of federal funds to particular purposes. See, e.g., Medina, 145 S. Ct. at 2226 (“In return for federal [Medicaid] funds, States agree to spend them in accordance with congressionally imposed conditions.” (citation and internal quotation marks omitted)). Appropriations may also contain intrinsic restrictions on the use of funds—such as by providing that funds may be used to purchase construction materials of only a certain quality—but they do not attempt to induce behavior from a coordinate branch through quid pro quo.

35 50 Op. O.L.C. __ (Apr. 1, 2026)

official or ceremonial duties of the President.” 44 U.S.C. § 2201(2). And it ostensibly requires the creation of such documents where they might not otherwise exist. Id. § 2203(a). Nothing about the PRA is conditional, nor does it afford the President the opportunity to decline its strictures. Compare id. § 2201 et seq., with Armstrong v. Exceptional Child Ctr., Inc., 575 U.S. 320, 323 (2015) (“Like other Spending Clause legislation, Medicaid offers the States a bargain: Congress provides federal funds in exchange for the States’ agreement to spend them in accordance with congressionally imposed conditions.”). Moreover, the appropriations that fund such recordkeeping activity do not make receipt or use of those monies conditional on compliance with the PRA. See generally Executive Office of the President Appropriations Act, 2024, Pub. L. No. 118-47, div. B, tit. II, 138 Stat. 531; see also Medina, 145 S. Ct. at 2232 n.4 (“Congress must clearly and unambiguously alert [recipients] to condi- tions associated with federal funding . . . .”). Nor could Congress, under the Spending Clause, condition the Presi- dent’s receipt of funds on compliance with the PRA. See Biden v. Nebras- ka, 143 S. Ct. 2355, 2375 (2023) (“It would be odd to think that separa- tion of powers concerns evaporate simply because the Government is providing monetary benefits rather than imposing obligations.”). Because appropriations and Spending Clause legislation are analytically distinct, the Supreme Court has never considered whether appropriations would be subject to the anti-coercion principle that applies in spending cases—but there is every reason to think that the vertical limits governing federal- state relationships would apply horizontally with even greater force. Just as Congress may not “indirectly coerce[]” states into accepting condition- al federal funds, NFIB v. Sebelius, 567 U.S. 519, 578 (2012) (opinion of Roberts, C.J.), the separation of powers and Article II independence prevent Congress from wielding the spending power to interfere with executive functions. The threat of coercion is far more acute between coordinate branches than in the federal-state context. When Congress attaches unacceptable conditions to federal grants, states retain a mean- ingful choice: They can decline the funds and turn to alternative revenue sources, including their own taxing power. But the Executive and Judici- ary possess no such escape hatch. The Appropriations Clause forbids them from drawing money from the Treasury except “in Consequence of Ap- propriations made by Law.” U.S. Const. art. I, § 9, cl. 7. Just as “Congress is obliged to provide public funds for constitutionally mandated activi-

36 Constitutionality of the Presidential Records Act

ties,” including the “constitutional activities of the President,” Stith, supra, at 1350–51, the President has “no real option but to acquiesce,” Sebelius, 567 U.S. at 582 (opinion of Roberts, C.J.), in accepting appro- priations integral to “the power to execute the laws and manage the execu- tive branch,” Meese III, supra, at 631 (statement of William Barr). Ac- cepting that appropriations are an exercise of the Spending Clause, without horizontal limits, would allow Congress to raise “a gun to the head” of the President, empowering it to superintend and subordinate the functions of a coordinate branch of government. Sebelius, 567 U.S. at 581 (opinion of Roberts, C.J.). Congress may not authorize appropriations—whatever the source of their power—in contravention of these principles. “A general appropria- tions authority to control the executive branch would circumvent, under- mine, indeed vitiate . . . th[e] laundry list of powers given to Congress.” Meese III, supra, at 642 (statement of Geoffrey Miller). As then-Assistant Attorney General William Barr once asked rhetorically, “[d]id the framers really believe that the appropriations clause transformed the relationship between Congress and the other coordinate branches into a relationship of master-servant, that the congressional master directs the activities of the Presidential or judicial servant simply because the money passes from hand-to-hand?” Id. at 631 (statement of William Barr). Although Con- gress may limit how a department uses appropriated funds, we have “long adhered to the view that Congress cannot use the appropriations power to control a Presidential power that is beyond its direct control.” Statutory Restrictions on the PLO’s Washington Office, 42 Op. O.L.C. 108, 131 n.16 (2018) (citation omitted); see also Constitutionality of Committee Approval Provision in Department of Housing and Urban Development Appropriations Act, 6 Op. O.L.C. 591, 592 (1982) (“HUD Appropria- tions”) (“Congress may not, by conditions attached to appropriations, provide for a discharge of the functions of Government in a manner not authorized by the Constitution.” (emphasis and citation omitted)). “If such a practice were permissible, Congress could subvert the Constitution. It might make appropriations on condition that the executive department abrogate its functions.” HUD Appropriations, 6 Op. O.L.C. at 592 (quot- ing Constitutionality of Proposed Legislation Affecting Tax Refunds, 37 Op. Att’y Gen. 56, 61 (1933)). Although Congress retains the power of the purse, the Executive Branch has emphasized that Congress may not “deprive the President” of his constitutional powers, such as his responsi-

37 50 Op. O.L.C. __ (Apr. 1, 2026)

bility to faithfully execute the laws, “by purporting to deny him the mini- mum obligational authority sufficient to carry this power into effect.” Authority for the Continuance of Government Functions During a Tempo- rary Lapse in Appropriations, 5 Op. O.L.C. 1, 5–6 (1981). We have consistently rejected the fallacy that Congress’s power to ap- propriate funds for executive operations authorizes regulation of how the President exercises his constitutional duties. Our repudiation has been no less forceful in cases involving the President’s Take Care Clause obliga- tion than in cases involving conclusive and preclusive presidential pow- ers. The salaries of senior presidential advisers, for instance, are paid through appropriations, yet we have been unequivocal that Congress may not compel their testimony, through statute or otherwise, irrespective of the assertion of executive privilege. See Testimonial Immunity, 43 Op. O.L.C. at 111; id. at 114 (“This is true whether or not the President might ultimately assert executive privilege . . . .”). Such compulsion “would interfere directly with the President’s ability to faithfully discharge his responsibilities.” Id. at 112. Similarly, although the Department of Justice is funded by congressional appropriations, “Congress, through exercise of its appropriations power or otherwise,” may not “prevent the Executive from advancing before the courts a particular view of the constitutionality of an Executive agency action or policy.” The Effect of an Appropriations Rider on the Authority of the Justice Department to File a Supreme Court Amicus Brief, 14 Op. O.L.C. 13, 19 (1990). Like receiving advice from presidential advisers, filing such briefs “is integral to the discharge of [the President’s] constitutional duty to see that the laws are faithfully execut- ed,” and the separation of powers prevents Congress from regulating that function through appropriations. Id. So, too, for the PRA, which mi- cromanages the documentation of “the performance of the President’s constitutional, statutory, or other official or ceremonial duties.” 44 U.S.C. § 2203(a); see also Separation of Powers, 20 Op. O.L.C. at 135 (“[L]egislation that has the purpose or would have the effect of ‘mi- cromanaging’ executive action” may “threaten the structural values pro- tected by the general separation of powers.”). We have applied this very logic to the related context of congressional oversight. “[T]he fact that the President or the federal courts may rely upon appropriated funds to carry out their activities does not mean that everything they do falls within the scope of the oversight authority.” Congressional Oversight at *10 n.5. “Therefore, the limits placed on 38 Constitutionality of the Presidential Records Act

Congress when conducting oversight pursuant to its general legislative power also apply to oversight conducted pursuant to its appropriations authority.” Id. at *11 n.5. Although Congress may review “non- substantive data” about the President’s use of appropriated funds, “Con- gress lacks the authority to inquire into the Executive’s substantive deci- sion-making in . . . areas [of exclusive executive authority].” Id. Moreo- ver, even though the President uses appropriated funds to perform statutory duties in addition to his constitutional obligations, “the depart- ments and agencies, not the White House, principally administer such programs,” and Congress ought first to seek information from those departments and agencies before seeking to regulate the President’s rec- ords. Id. at *27. If it were otherwise, then Congress would have the same power to enact a companion to the PRA in a JRA—a Judicial Records Act—that required the Chief Justice and Associate Justices of the United States Supreme Court to preserve or create “judicial records.” Congress’s power over the use of federal money is awesome, but neither the Appropriations Clause nor the Spending Clause could go that far. Congress may set the funding for federal courts, address their jurisdiction, and impeach judges who transgress “good Behaviour,” U.S. Const. art. III, § 1, but it may not invade the Supreme Court’s independence and deliberative secrecy on the mere ground that it funds their buildings, staff, and writings. 8 Just as Congress could not gild the gavel and then purport to guide it, it could not expropriate judicial records as necessary and proper to safeguard the integrity of its appropriations to the Judiciary. 9 Like the President’s inde-

8 Inherent features of Article III would also protect the Judiciary against some congres-

sional encroachments, much as executive privilege protects Article II prerogatives. But the Constitution protects the separation of powers first by limiting congressional powers to those enumerated and those necessary and proper to effectuating them. No enumerated congressional power authorizes or necessitates such a regulation of judicial or presidential records. See Congressional Oversight at *2 (“A congressional information request is valid only if it is related to, and in furtherance of, a legitimate task of the Congress. Conse- quently, the Executive Branch must scrutinize the asserted legislative purpose underlying a congressional request by examining the objective fit between that purpose and the information sought.” (cleaned up)). 9 Congress has never attempted to regulate the records of sitting Supreme Court Justic-

es. In fact, only once has it ever attempted to mandate the disclosure of information from a sitting Supreme Court Justice, subpoenaing Justice Thomas C. Clark to testify in front of the House Un-American Activities Committee about his actions as President Truman’s

39 50 Op. O.L.C. __ (Apr. 1, 2026)

pendence protected by Article II, there is an “imperative need for total and absolute independence of judges in deciding cases or in any phase of the decisional function,” even though congressional appropriations make the decisional function possible. Chandler v. Jud. Council of the Tenth Cir., 398 U.S. 74, 84 (1970). Congress has no enumerated power to regulate judicial records, and it cannot conjure such a power into existence through its power of the purse. So, too, with presidential records. Finally, historical practice reinforces this conclusion. For nearly two centuries, Presidents often declined to furnish all documents requested by Congress, instead choosing to supply only those requested documents that were, in the President’s judgment, appropriate. Yet Congress never at- tempted to use appropriations or invoke the Spending Clause to force the disclosure of all documents it thought relevant. Political negotiation, not Spending Clause legislation, carried the day.

E.

Finally, the PRA is not a valid exercise of Congress’s power to enact legislation to assist in the execution of the President’s powers. In 2020, our Office considered whether the Logan Act—which criminalizes the conduct of diplomacy by private citizens absent presidential authorization, see 18 U.S.C. § 953—was a constitutional exercise of Congress’s enu- merated powers. See generally The Logan Act, 44 Op. O.L.C. 258 (2020). We concluded that it was. Even though the authority to conduct diplomacy is a quintessential exclusive executive power, we explained that the Necessary and Proper Clause authorizes Congress to enact legisla- tion that assists the coordinate branches in the execution of their powers. See id. at 294. The PRA—which lacks the 226-year pedigree of the Logan Act, see id. at 258—cannot be sustained on a similar basis. Where the Logan Act responded to an external threat to presidential authority (that of unauthor- ized private diplomacy), the PRA posits a threat endogenous to the Presi- dency itself—that the President would threaten the very office he holds.

Attorney General. See Todd Garvey, Cong. Rsch. Serv., LSB10962, Enlisting Assistance or Intruding on Judicial Independence? Compelling Testimony by Supreme Court Justic- es, at 2 (May 9, 2023). Justice Clark refused, “asserting that his appearance would threaten the ‘complete independence of the judiciary.’” Id.

40 Constitutionality of the Presidential Records Act

This premise is at odds with both the presumption of regularity and the great weight of historical practice. Congress may not superintend the Presidency on the theory that Congress has its coordinate branch’s inter- ests more at heart than does the President himself. Moreover, the Logan Act vested prosecutorial discretion in the Executive Branch, ensuring that the statute would never interfere with executive prerogatives. The PRA, by contrast, operates as a mandatory regulatory scheme that constrains the President’s day-to-day operations.10 Far from aiding the President in executing his constitutional duties, the Act threatens to impede his per- formance in ways we have already explained. A statute that burdens rather than facilitates the exercise of executive power is not “necessary and proper” to carry that power into execution.

III.

Nixon v. Administrator does not save the PRA. It is distinguishable because it addressed a materially narrower statute under extraordinary circumstances, balancing the asserted interests of Congress and a former President with little attention to Article I. And its separation of powers analysis is wrong because it elides the distinction between the President and congressionally created agencies and fails to recognize that congres- sional regulation of presidential records implicates the foundational constitutional principle of executive independence in addition to confi- dentiality.

Congress enacted the PRMPA in the wreckage of Watergate, in the face of the very real and imminent probability that former President Nixon would destroy documents proving misconduct. See supra Part I.C (dis- cussing the facts of Nixon v. Administrator). Shortly after his resignation, Nixon entered into an agreement with the Administrator of General Ser-

10 Recasting the PRA as a criminal prohibition would only undermine rather than save

its constitutionality. See Trump, 144 S. Ct. at 2330–31 (“Criminally prosecuting a Presi- dent for official conduct undoubtedly poses a far greater threat of intrusion on the authori- ty and functions of the Executive Branch than simply seeking evidence in his posses- sion . . . . The danger is . . . that the President would be chilled from taking the ‘bold and unhesitating action’ required of an independent Executive.” (citation omitted)).

41 50 Op. O.L.C. __ (Apr. 1, 2026)

vices that would have, among other things, permitted him to direct the destruction of tape recordings directly relevant to ongoing Watergate investigations. See Nixon v. Administrator, 433 U.S. at 432.11 Congress responded by enacting the PRMPA, which targeted presidential docu- ments and recordings made between Nixon’s 1969 inauguration and his 1974 resignation. See id. at 433–34. The statute directed the Administrator to take custody of Nixon’s materials, screen them for relevance to ongo- ing investigations and judicial proceedings, and preserve them subject to valid claims of executive privilege. Id. at 433–35. The PRMPA and the PRA could not be more different. The PRMPA sought a discrete set of identified materials under extraordinary circum- stances—a President’s resignation amidst constitutional crisis, ongoing criminal investigations, and the very imminent possibility of the destruc- tion of recordings directly relevant to the Watergate scandal. See id. at 452–54; see also id. at 432–33. As the Court saw it, Congress had a legitimate interest in “preserving the materials as a source for facilitating a full airing of the events leading to [Nixon]’s resignation, and Congress’ need to understand how those political processes had in fact operated in order to gauge the necessity for remedial legislation.” Id. at 453. Thus, the Court held that the PRMPA “may be thought to aid the legislative process and thus to be within the scope of Congress’ broad investigative power.” Id. The PRA, by contrast, applies to all Presidents in perpetuity, regard- less of scandal or specific investigatory need. Where the PRMPA ad- dressed completed events with a statute that would naturally sunset once those events were investigated, the PRA establishes a permanent regime governing all presidential records, including those that have not yet been created, and without a showing that such records might aid in legislation presently under consideration. 12

11 The agreement also would have allowed Nixon, after three years, to withdraw “any

or all of the Materials” and “retain them for any purpose determined by him.” Nixon v. Administrator, 433 U.S. at 432 (cleaned up). 12 As we have explained in the context of executive privilege:

When the Supreme Court held that the need for presidential communications in the criminal trial of President Nixon’s close aides outweighed the constitutional privi- lege, an important premise of its decision was that it did not believe that advisers will be moved to temper the candor of their remarks by the infrequent occasions of disclosure because of the possibility that such conversations will be called for in the

42 Constitutionality of the Presidential Records Act

True, Nixon v. Administrator employed broad language that might seem to support Congress’s authority to enact the PRA. The Court suggested, for example, that “Congress can legitimately act” to ensure that an incum- bent President is not “dependent on happenstance or the whim of a prior President when he seeks access to records of past decisions that define or channel current governmental obligations.” Id. at 452–53. And the Court alluded to “the American people’s ability to reconstruct and come to terms with their history.” Id. But these statements do not change our conclusion. Context matters: The Court was addressing Nixon’s materials specifically—records of Watergate, an aborted Presidency, and ongoing criminal and congressional investigations. Ensuring the incumbent’s access to those records and the public’s ability to reconstruct that history served concrete needs that the Court believed were sufficient to justify what it viewed as a limited intrusion on a former President. Id. at 451–55 (balancing asserted interests of Congress against asserted burden imposed on the Executive Branch by “the mere screening” of presidential records); cf. id. at 481 (noting that Nixon had voluntarily placed his records in the custody of the General Services Administration before the PRMPA’s enactment). Extrapolating from that holding a general congressional authority to regulate all presidential records for all time makes hash of the Court’s situated reasoning.13

context of a criminal prosecution. By contrast, congressional requests for executive branch deliberative information are anything but infrequent. Moreover, compared to a criminal prosecution, a congressional investigation is usually sweeping; its is- sues are seldom narrowly defined, and the inquiry is not restricted by the rules of evidence. Finally, when Congress is investigating, it is by its own account often in an adversarial position to the executive branch and initiating action to override judgments made by the executive branch. This increases the likelihood that candid advice from executive branch advisers will be taken out of context or misconstrued. Congressional Oversight at *29 n.11 (quoting Congressional Requests, 13 Op. O.L.C. at 156–57). 13 Before the PRA’s enactment, this Office relied on some of the broad statements in

Nixon v. Administrator while testifying to Congress about the constitutionality of the Presidential Papers Act of 1978. See Presidential Records Act of 1978: Hearings on H.R. 10998 and Related Bills Before a Subcomm. of the H. Comm on Gov’t Operations, 95th Cong. 87–133 (1978) (statement of Lawrence A. Hammond, Deputy Assistant Attorney General, Office of Legal Counsel). At the time, we understood Nixon v. Admin- istrator to indicate that “it is within the appropriate ambit of Congress’ power to legislate with respect to the preservation of the historically valuable papers of the Chief Execu-

43 50 Op. O.L.C. __ (Apr. 1, 2026)

We are particularly hesitant to treat the Court’s broader statements in Nixon v. Administrator as controlling because President Nixon did not raise—and the Court therefore did not address—any challenge to Con- gress’s Article I power to enact the PRMPA. See Brief for Appellant, Nixon v. Administrator, 433 U.S. 425 (No. 75-1605), 1977 WL 189790 (raising separation of powers, Bill of Attainder Clause, privacy, and First Amendment claims). Nixon’s brief mentioned Congress’s interests in regulating presidential documents only to argue that “it was not Congress’ province to decide” that those interests “outweigh[ed] the President’s interest in confidentiality.” Id. at *112. That is why the Court discussed the interests underlying the PRMPA in the context of balancing those interests against the Executive Branch’s need for confidentiality—not as part of a holding about whether Congress possessed the Article I power to enact the PRMPA in the first place. See, e.g., 433 U.S. at 452 (“[A]dequate justifications are shown for this limited intrusion into execu- tive confidentiality . . . .”).14

tive.” Id. at 112 (citing 433 U.S. at 477–78 (discussing Nixon’s bill of attainder argu- ment)). We made this observation in passing, not to validate a congressional power to preserve presidential records but because we thought “it follow[ed] that, at least insofar as declaring the President’s official papers to be public property [instead of private property of the President] is concerned, Congress’ action is not subject to serious challenge.” Id. We nevertheless qualified that advice, explaining that we reviewed the bill “with an eye toward identifying those matters that deserve clarification in order to avoid what might be serious constitutional problems in the ultimate application of the statute . . . [in] particular cases.” Id. at 109 (emphases added). That advice has not withstood the test of time. See, e.g., Nixon, 978 F.2d at 1284 (“The government has, pursuant to PRMPA, taken complete possession and control of the Nixon papers. Although a great public interest may justify a taking, it does not convert the taking into mere regulation.”). The preservation of historically important documents must relate to a valid legislative purpose, as we explain in greater detail above. See supra Part II.B. Thus, “intervening developments in the law appear to cast doubt” upon the testimony given before the PRA’s enactment. Reconsidering Whether the Wire Act Applies to Non- Sports Gambling, 45 Op. O.L.C. 158, 178 (2018). Aside from that testimony, our Office has never opined on the facial constitutionality of the PRA. 14 In a section of Nixon v. Administrator that is even farther afield, the Court addressed

the former President’s argument that the PRMPA constituted a bill of attainder. See 433 U.S. at 468–84. In this context, the Court noted “Congress’ interest in and expansive authority to act in preservation of monuments and records of historical value to our national heritage” to underscore the PRMPA’s nonpunitive purpose. Id. at 477–78. Although this bill-of-attainder analysis has little relevance here, we note that the PRA is

44 Constitutionality of the Presidential Records Act

Mazars also confirms that we should read these statements narrowly. When Congress invokes its oversight authority, it must provide “detailed and substantial” evidence of a valid legislative aim rather than “vague and loosely worded evidence.” 140 S. Ct. at 2036 (citation and internal quota- tion marks omitted). And when it contemplates “legislation concerning the Presidency,” it must “adequately identif[y] its aims” and explain how “the President’s information will advance its consideration of the possible legislation.” Id.

Nixon v. Administrator is not only distinguishable. It was also wrong in concluding that the PRMPA’s “regulation of the disposition of Presi- dential materials within the Executive Branch” was not “a violation of the principle of separation of powers.” 433 U.S. at 441. Otherwise-valid exercises of congressional power must “take adequate account of the significant separation of powers issues raised” when directed at the Presi- dent. Mazars, 140 S. Ct. at 2033. For this reason, the Supreme Court has warned that an investigation that targets the President’s papers cannot be supported by “[i]nvoking our precedents concerning investigations that did not target the President’s papers.” Id. But to support the PRMPA’s assertion of power over presidential records, Nixon v. Administrator invoked “statutory precedent” showing that Congress had previously regulated the records of agencies created by statute. 433 U.S. at 445. This reflects the “ancien regime” of the Court’s “mid-twentieth century” ap- proach to separation of powers, not the more thoughtful approach appro-

plainly not an attempt to exercise the power of eminent domain: Unlike the PRMPA, it contains no just-compensation provision. And wielding the eminent-domain power against the President would raise serious constitutional questions not addressed in Nixon v. Administrator’s argument about the nonpunitive purpose of the PRMPA. Be- cause Congress has no enumerated power of eminent domain, it may exercise this implied power only “so far as is necessary to the enjoyment of the powers conferred upon it by the Constitution.” Kohl v. United States, 91 U.S. 367, 372 (1875); see also Chappell v. United States, 160 U.S. 499, 509–10 (1896); PennEast Pipeline Co., LLC v. New Jersey, 141 S. Ct. 2244, 2266 (2021) (Barrett, J., dissenting) (“Any taking of property provided for by Congress is thus an exercise of another constitutional power . . . augmented by the Necessary and Proper Clause.”). But Nixon v. Administrator did not identify any relevant Article I power or address how separation of powers concerns might limit any implied power to exercise eminent domain against the President.

45 50 Op. O.L.C. __ (Apr. 1, 2026)

priately required by subsequent developments in Supreme Court doctrine. See Ziglar v. Abbasi, 582 U.S. 120, 131 (2017). Nixon v. Administrator was also mistaken in reasoning that the PRMPA was not “unduly disruptive of the Executive Branch” because “[t]he Executive Branch remain[ed] in full control of the Presidential materials, and the Act facially [was] designed to ensure that the materials [could] be released only when release [was] not barred by some applicable privilege inherent in that branch.” 433 U.S. at 444–45. This analysis was incom- plete. Congress’s pursuit of information from the President has proven to raise additional separation of powers concerns beyond those considered in Nixon v. Administrator, and those concerns are not resolved by executive privilege and custody. “Article II not only gives the President the ability to consult with his advisers confidentially, but also, as a corollary, it gives him the flexibility to organize his advisers and seek advice from them as he wishes.” Con- gressional Oversight at *8 (quoting Ass’n of Am. Physicians & Surgeons, Inc., 997 F.2d at 909). Such flexibility protects the President’s constitu- tional “interests in autonomy and independence,” id. at *3, “as well as the confidentiality of his communications,” id. at *22. These interests are distinct. See Testimonial Immunity, 43 Op. O.L.C. at 125 (“The immunity from compelled congressional testimony implicates fundamental separa- tion of powers principles that are separate from the confidentiality of specific information.”). And both must be protected to ensure the Presi- dent’s ability to receive “candid advice” from his advisers, which is “essential to presidential decision-making.” Id. at 111. Nixon v. Administrator was wrong to suggest that the executive privi- lege provisions of the PRMPA avoided separation of powers concerns, and those concerns apply even more strongly to the PRA. In the context of congressional subpoenas for the testimony of the President’s senior advis- ers, we have recognized a “testimonial immunity” that is “distinct from, and broader than, executive privilege.” Id. In particular, we have ex- plained why “[executive] privilege is insufficient to ameliorate several threats that compelled testimony poses to the independence and candor of executive councils.” Id. at 113. First, “even the prospect of compelled interrogation by a potentially hostile congressional committee about confidential communications with the President . . . could chill presiden- tial advisers from providing unpopular advice or from fully examining an

46 Constitutionality of the Presidential Records Act

issue with the President or others.” Id. at 114 (cleaned up). Second, “compelled congressional testimony creates an inherent and substantial risk of inadvertent or coerced disclosure of confidential information.” Id. at 113 (cleaned up). And third, “preparing for such examinations would force [the President’s senior advisers] to divert time and attention from their duties to the President at the whim of congressional committees.” Id. at 112. Although not a perfect analogue to the compelled live testimony of presidential advisers, the PRA engenders many of the same risks, notwith- standing the Act’s provisions for the assertion of executive privilege. Contra Nixon v. Administrator, 433 U.S. at 444 (rejecting Nixon’s separa- tion of powers challenge because “[t]he Executive Branch remains in full control of the Presidential materials, and the Act facially is designed to ensure that the materials can be released only when release is not barred by some applicable privilege inherent in that branch”). First, the PRA may chill the President’s advisers from offering candid or unpopular advice, because there is uncertainty as to whether the Presi- dent will invoke the privilege on that specific topic. See Testimonial Immunity, 43 Op. O.L.C. at 111–14. Invoking executive privilege incurs “political costs,” and an adviser may self-censor given the “uncertainty over whether a particular matter” will ultimately be subject to future disclosure. Id. at 114; see also Congressional Oversight at *28. The PRA’s mechanisms for the assertion of privilege also might not adequate- ly secure the candor the President requires from his advisers. To start, an incumbent President might choose “not to uphold [a] claim of privilege asserted by the former President,” requiring the latter to assert the claim in court. 44 U.S.C. § 2208(c)(2)(C). And because such a claim would thus be subject to litigation, advisers might reasonably fear that a good-faith claim of privilege will nevertheless fail in court. Moreover, assertions of privilege “must be made personally by a former President or the incum- bent President” to shield documents from otherwise compelled disclosure. Id. § 2208(b)(1). Once a former President passes away, only an incumbent President may assert privilege with respect to the deceased President’s documents, providing further reason for advisers to worry that their ad- vice will become public, notwithstanding the PRA’s privilege provisions. And an incumbent President might be unwilling to assert privilege over the documents of one of his predecessors. But see Testimonial Immunity,

47 50 Op. O.L.C. __ (Apr. 1, 2026)

43 Op. O.L.C. at 123 (“[I]f the doctrine of separation of powers and the independence of the Presidency is to have any validity at all, it must be equally applicable to a President after his term of office has expired when he is sought to be examined with respect to any acts occurring while he is President.” (citation omitted)). Second, and relatedly, the PRA poses a risk of inadvertent disclosure of sensitive or privileged information. See id. at 113. In considering the permissibility of live testimony before a congressional committee, we emphasized that senior presidential advisers might be faced with “a wide range of unanticipated and hostile questions,” and that “[i]n the heat of the moment, without the opportunity for careful reflection,” advisers might inadvertently disclose privileged information. Id. (citation omitted). Here, the risk of inadvertent disclosure arises from the sheer volume of records that must be created, preserved, and reviewed and from the PRA’s default toward disclosure if grounds for special treatment are not identi- fied and affirmatively asserted. Cf. Congressional Oversight at *52–53, *58 (discussing the dangers of overbroad congressional subpoenas for White House records). Third, notwithstanding executive privilege, the mandates imposed by the PRA might “burden White House personnel to a degree that prevents them from effectively advising and assisting the President in the perfor- mance of his constitutional duties.” Id. at *52–53; cf. Testimonial Immuni- ty, 43 Op. O.L.C. at 112 (discussing the “time and attention” required to prepare for congressional testimony). Given the capacious sweep of the PRA, “the White House’s autonomy may be compromised” by the stat- ute’s potential to “distract personnel and drain critical resources.” Con- gressional Oversight at *28; see also id. (“Intrusive congressional over- sight of the White House’s interaction with departments and agencies may cause White House staff members to conform their information-gathering and policy-formulation processes to the demands of Congress instead of the needs of the President.”). In comparison to executive departments, “White House components have small staffs who are primarily devoted to advising and assisting the President.” Id. at *58. The PRA requires the expenditure of valuable White House resources to ensure compliance, impeding the President’s access to information and advice that he might otherwise receive. These resource constraints also reify our point about the risk of inadvertent disclosure. White House personnel must either devote considerable time ensuring the protection of privileged information 48 Constitutionality of the Presidential Records Act

(diverting attention away from advising the President) or prioritize advis- ing the President (heightening the risk of inadvertent disclosure). Nor is it an answer, as Nixon v. Administrator suggested, that the “Ex- ecutive Branch remains in full control of the Presidential materials.” 433 U.S. at 444. The PRA contains several mechanisms by which records might leave the control of the Executive Branch. We have also advised in other contexts that presidential control is not a panacea for the separation of powers. For example, we have opined that an Inspector General for the EOP would be unconstitutional, even if the Inspector General “would be subject to the authority, direction, and control of the President” with respect to certain types of information access and even absent a disclosure of protected information. Memorandum for Andrew Fois, Assistant Attorney General, Office of Legislative Affairs, from Randolph D. Moss, Deputy Assistant Attorney General, Office of Legal Counsel, Re: Inspector General for the Executive Office of the President at 2–3 (July 24, 1996). Nixon v. Administrator’s failure to appreciate the Article II conse- quences of permitting Congress to regulate presidential records is, in a sense, understandable. Its analysis concerned the records of a scandal- ridden President who had already left office; it did not address the bur- dens imposed by a forward-looking scheme like the PRA on incumbent Presidents. That Nixon and the Administrator had entered into an agree- ment allowing for the destruction of materials needed by the Watergate Special Prosecutor, see 433 U.S. at 431–32, colored the Court’s analysis, as the Court suggested that Congress’s power to regulate such documents was “augmented” by the PRMPA’s “important interests,” id. at 445–46. And because the PRMPA was the first attempt to regulate the disposition of presidential records by statute, Nixon’s advisers would have had little reason to think—until the Court’s opinion in the resulting case—that such records would have ever been within the reach of Congress outside of the accommodation process. Thus, applying Nixon v. Administrator’s crabbed separation of powers analysis to the PRA would vitiate the Article II interests in executive autonomy and independence.

IV.

We turn finally to severability. We must assess whether the PRA “con- tains unobjectionable provisions separable from those found to be uncon-

49 50 Op. O.L.C. __ (Apr. 1, 2026)

stitutional.” El Paso & N.E. Ry. Co. v. Gutierrez, 215 U.S. 87, 96 (1909). We presumptively give “full effect” to those provisions that “are not repugnant to the [C]onstitution.” Bank of Hamilton v. Dudley’s Lessee, 27 U.S. (2 Pet.) 492, 526 (1829) (Marshall, C.J., for the Court); see Con- stitutionality of Race-Based Department of Education Programs, 49 Op. O.L.C. __, at *14–15 (Dec. 2, 2025) (“Race-Based Education Programs”) (explaining the presumption of severability). This presumption is over- come when the remaining portions of the statute either cannot function independently without the unconstitutional provisions or “would not function in a manner consistent” with Congress’s intent. Race-Based Education Programs at *14 (citations, internal quotation marks, and emphasis omitted); see also id. at *15 n.9 (“[O]ur own precedents on severability . . . have long followed the familiar, two-part function-and- intent inquiry.”); Alaska Airlines, Inc. v. Brock, 480 U.S. 678, 685 (1987) (holding that provisions are not severable where the statute would not “function in a manner consistent with the intent of Congress” (emphasis in original)). Applying these standards, we conclude that the PRA is invalid in its entirety. This finding follows directly from the nature of the PRA’s prin- cipal constitutional defect: Congress lacks Article I authority to regulate or access the President’s records absent a valid legislative purpose, and no such purpose exists for the PRA. Because this infirmity pervades the entire statute, the statute cannot survive. In other words, the PRA general- ly contains no “unobjectionable provisions,” much less ones that are “separable from those found to be unconstitutional.” El Paso & N.E. Ry. Co., 215 U.S. at 96. Although a handful of narrow custodial provisions might seem to offer permissible assistance to the President in performing his constitutional and statutory duties, these provisions are inseverable from the unconstitu- tional provisions of the Act and cannot stand on their own. Sec- tion 2203(f), for example, authorizes the Archivist to “maintain and preserve Presidential records on behalf of the President” during the Presi- dent’s term in office. 44 U.S.C. § 2203(f). Assuming arguendo that Con- gress has authority to assign such custodial duties to the Archivist, this provision serves a function divorced from the statute’s core purpose— ensuring public access to presidential records at the end of an administra- tion. See H.R. Rep. No. 95-1487, at 2. Untethered from the PRA’s broader

50 Constitutionality of the Presidential Records Act

regulatory framework, this isolated provision would not alone “function in a manner consistent with the intent of Congress.” Alaska Airlines, 480 U.S. at 685 (emphasis in original). Moreover, should the President wish to employ an Officer in maintaining and preserving presidential records, it is unlikely the President requires a bespoke statute to do so. Other provisions might appear salvageable because they could apply only when Congress has a valid legislative interest. For instance, when the President decides to dispose of presidential records, the Archivist must request advice from certain congressional committees if the “records may be of special interest to the Congress.” 44 U.S.C. § 2203(e)(1). But even assuming for the sake of argument that this provision applies only when Congress has a valid legislative purpose—which is far from clear—it would not survive rigorous severability analysis. The congressional con- sultation provision bears little relation to the PRA’s central aim, so it would not function consistently with congressional intent without the statute’s more robust (but unconstitutional) regulatory provisions. See Alaska Airlines, 480 U.S. at 685. Severability might be a closer question if Congress or the Archivist could require preservation upon certifying that the records serve a valid legislative purpose. But as we have ex- plained, neither Congress nor the Archivist can “veto the President’s decision to destroy the records.” Armstrong, 924 F.2d at 286; Citizens for Resp. & Ethics in Wash. v. Trump, 302 F. Supp. 3d 127, 131 (D.D.C. 2018). Congress likely did not intend this standalone notification provi- sion to survive absent the statute’s broader framework. Ultimately, the PRA falls entirely. The constitutional defects lie at the statute’s core: Congress’s attempt to regulate presidential records without a valid legislative purpose, and the burdens on the Presidency that such regulation effects. Even if scattered auxiliary provisions might be valid standing alone, they cannot survive without the central regulatory frame- work Congress designed. Severing that foundational framework while allowing these provisions to remain would leave a statute that bears no resemblance to what Congress intended to enact.

* * * * *

The PRA is not a valid exercise of Congress’s Article I authority and unconstitutionally intrudes on the independence and autonomy of the President guaranteed by Article II. The Act establishes a permanent and 51 50 Op. O.L.C. __ (Apr. 1, 2026)

burdensome regime of congressional regulation of the Presidency unteth- ered from any valid and identifiable legislative purpose. For these rea- sons, the PRA is unconstitutional, and the President need not further comply with its dictates. See Issues Raised by Provisions Directing Issu- ance of Official or Diplomatic Passports, 16 Op. O.L.C. 18, 32 (1992) (concluding that the President may decline to enforce statutes he views as unconstitutional).

T. ELLIOT GAISER Assistant Attorney General Office of Legal Counsel