Paper Coup: How a DOJ Opinion Could Erase Evidence of Election Suppression and Solidify Presidential Authoritarianism
On April Fools’ Day, 2026, the Justice Department’s Office of Legal Counsel issued a 52-page opinion declaring the Presidential Records Act of 1978 unconstitutional in its entirety. The opinion, authored by Assistant Attorney General T. Elliot Gaiser, was not issued as a joke.
Gaiser’s opinion asserts that the president “need not further comply” with the Act’s requirements and that the records generated by his administration are his personal property, not the government’s.
Most coverage treated this as a legal curiosity—an aggressive but abstract assertion of executive power. It is not abstract.
With the 2026 midterm elections seven months away, the Gaiser opinion could serve as the legal keystone of an operational architecture that would help make federal voter suppression operations significantly harder to detect or challenge.
To understand why, you have to grasp what the opinion actually does—not just to presidential recordkeeping, but to the entire accountability infrastructure surrounding the use of federal law enforcement and military assets during an election cycle.
What the Gaiser Opinion Actually Does
The Presidential Records Act was enacted in the aftermath of Watergate to ensure that the deliberations, decisions, and communications of the president and his immediate staff became the property of the American people via the United States government, not the private property of the person who happened to hold the office. Under the Act, presidential records must be preserved, transferred to the National Archives at the end of an administration, and eventually made available to the public.
The Gaiser opinion declares this entire framework unconstitutional.
It concludes that Congress had no enumerated or implied power to enact the PRA, that the statute aggrandizes the Legislative Branch at the expense of executive independence, and that the president’s papers—including his deliberations on policy, his communications with advisers, and his operational directives—belong to him personally and may be destroyed at his discretion.
The opinion is legally aggressive but not entirely frivolous.
It draws on real doctrine, particularly the Supreme Court’s 2020 decision in Trump v. Mazars USA, which imposed heightened constitutional scrutiny on congressional subpoenas targeting presidential papers. But the opinion dramatically overreads Mazars, which addressed specific subpoenas—not prospective records legislation. And it ignores fifty years of case law—the DC Circuit’s Armstrong line of decisions, the district court’s ruling in American Historical Association v. Peterson—that carefully balanced executive prerogative against institutional preservation interests.
The legal merits, however, are largely beside the point.
OLC opinions bind the executive branch as a matter of institutional practice. The moment this opinion issued, every White House, NSC, and EOP advisory component was freed as a matter of institutional policy to create, preserve, transfer, or destroy records of the president’s deliberations. The documentary trail of executive branch decision-making—the planning documents, targeting analyses, deployment orders, and legal justifications that would ordinarily become the evidentiary record of how power was exercised—can now be destroyed under the legal rationale of the new OLC opinion—and those who destroy them will claim, with at least colorable authority, that they had every right to do so. Indeed, prior federal case law likely makes Gaiser’s gambit more plausible.
Under the due process doctrine of entrapment by estoppel—rooted in the Supreme Court’s decisions in Raley v. Ohio, 360 US 423 (1959), and Cox v. Louisiana, 379 US 559 (1965)—a defendant who reasonably relies on an official government representation that conduct is lawful has a constitutional due process defense to prosecution. Circuit courts have elaborated the doctrine to require that an authorized government official, aware of all relevant facts, affirmatively tell the defendant the proscribed conduct was permissible, and that the defendant’s reliance was reasonable. An OLC opinion is about as authoritative an official representation as exists in the federal system. It is the president’s chief legal interpreter and advisor, telling executive branch actors what the law requires.
The Operational Architecture
The threat to the 2026 midterms is not hypothetical. The administration has assembled an enforcement infrastructure that could readily be repurposed for federal election interference. What it previously lacked was a legal framework to make that infrastructure accountable to no one. The Gaiser opinion provides it.
Consider what is already in place. DHS and ICE have, as documented in myriad press stories and multiple federal court proceedings, developed geospatial targeting tools that identify “target-rich enforcement environments” using demographic data. Those tools can be repurposed—with simple changes to query parameters—to identify polling locations in competitive congressional districts with high concentrations of Latino and Asian American voters who trend Democratic. The data infrastructure for that operation already exists. It was built for immigration enforcement. Adapting it for voter suppression operations would be trivial.
In February 2026, DHS Deputy Assistant Secretary Heather Honey told state election officials that ICE would not be deployed “at” polling places. She said nothing about ICE operating in the vicinity of polling places.
The legal distinction matters: most state statutes prohibiting armed personnel near polls define the protected zone as within 100 to 150 feet of the entrance. Immigration enforcement checkpoints positioned on access roads and sidewalks beyond those boundaries do not violate the letter of those statutes while achieving the same chilling effect. A voter who passes a line of ICE agents on the way to cast a ballot—regardless of citizenship status—faces a calculated deterrent.
The statutory framework prohibiting such operations is clear. 18 U.S.C. § 592 bars armed federal personnel at election sites. 52 U.S.C. § 10307(b) prohibits voter intimidation. The administration’s likely argument—that INA enforcement authority under 8 U.S.C. § 1357 provides independent authorization—fails under the canon requiring specific statutes to control over general ones.
Election protection statutes are specifically designed to limit federal law enforcement presence near polling places; INA grants general immigration enforcement authority. The specific controls…or at least it would if the local US Attorney were genuinely politically neutral and dedicated to upholding the law.
As we’ve seen repeatedly with the Trump Justice Department, apparent fealty to Trump’s agenda, not the law or the Constitution, is what matters. It seems entirely plausible that those US Attorneys may be instructed to prioritize immigration enforcement operations in connection with the midterms and prosecutions of protesters who “impeded” federal officers performing said tasks.
Under the OLC opinion, orders to that effect that flow from the White House itself would not be subject to retention or disclosure. Legal arguments challenging such a voter suppression scheme require evidence to litigate. And evidence requires, among other things, records. That is where the new opinion becomes operationally decisive.
The Records Destruction Architecture
In the hopefully unlikely event such a scheme were deployed, a sophisticated election suppression operation would not be executed through a single directive. It would be layered—each layer at a different classification and distribution level, each performing a distinct function, and each exploiting the Gaiser opinion’s elimination of statutory record preservation requirements.
The public-facing layer would be a facially neutral executive order invoking “election security” authority. The operational layer would be a Controlled Unclassified Information directive—marked Law Enforcement Sensitive—directing ICE and CBP field offices to conduct “enhanced immigration enforcement operations” in designated geographic areas during the election period. That designation, combined with the DHS geospatial targeting infrastructure, would place those operations in the specific precincts that matter.
The directive could specify that operational logs and deployment records are to be submitted to a White House Counsel-designated Records Control Officer rather than entered into standard ICE case management systems—characterizing them as generated in support of a presidential constitutional function and therefore, under Gaiser, as personal presidential property subject to no statutory preservation requirement.
The military component adds another layer.
The administration’s prior domestic military deployments—in Los Angeles, Washington, DC, Chicago, and Portland—demonstrate a consistent pattern: each deployment was predicated on a national security characterization of what were in fact civilian protests; each was in a Democratic-led city; and each was used to test and expand the boundaries of Posse Comitatus Act exceptions. A classified Presidential Directive pre-positioning military assets for rapid deployment in response to post-election protests—justified by a classified intelligence assessment characterizing anticipated protests as a foreign-influenced national security threat—would generate the same type of NSC deliberative records that the Gaiser opinion now characterizes as presidential personal property.
The assessment can be destroyed. The NSC meeting minutes can be destroyed. The deployment authorization can be destroyed. What remains is only the classified directive itself, shielded by standard national security classification authority rather than the PRA’s eventual disclosure clock.
This is not mere speculation about a future threat. The JAG officers who raised legal concerns about proposed or actual administration policies have been fired. Their replacements have not, so far as we know, actively challenged orders that outside observers and legal scholars deem dubious at best. The Supreme Court’s 1827 ruling in Martin v. Mott–whose broad unreviewability language the administration has already invoked in Chicago and Los Angeles, and which Ninth Circuit judges have argued in recent en banc proceedings is narrower than the administration claims, applying only to intra-military chain-of-command deference rather than foreclosing judicial review of presidential deployment determinations altogether—has supplied the claimed legal architecture for those deployments. The OLC opinion adds the records destruction authorization that makes the pre-deployment planning unrecoverable.
The Closure of Accountability Mechanisms
Three accountability mechanisms would ordinarily constrain what I have described. The Gaiser opinion systematically disables each of them.
Congressional oversight is at its best when it obtains access to executive branch records. A congressional subpoena for NSC planning documents related to election-period enforcement operations now encounters an administration that will argue those records, if they exist, are the president’s personal property. Under the heightened standard the administration derives from Mazars, Congress must identify a specific, valid legislative purpose before demanding them. The administration will refuse, litigate, and the midterms will be over before a court orders production.
FOIA litigation—the mechanism I have used in my own ongoing cases against the FBI, NSA, and NORTHCOM—is neutralized for EOP records because those records were never required to be created or preserved under Gaiser’s opinion. EOP records were already largely exempt from FOIA before Gaiser—that exemption is not new. What Gaiser adds is the elimination of the preservation obligation itself: records that previously had to be created, retained, and eventually transferred to the Archives (even if FOIA-exempt during that period) can now simply not exist. A FOIA request for White House communications directing election-period enforcement operations will accordingly receive a response asserting that no responsive federal records exist—which, under Gaiser’s framework, is technically accurate. The records that do exist in agency systems will be withheld under law enforcement privilege for ongoing operations and law enforcement techniques and procedures.
Criminal accountability is the most completely foreclosed. 18 U.S.C. § 2071, which criminalizes the willful destruction of federal records, applies to federal records within the meaning of the Federal Records Act. Under Gaiser, presidential records are not federal records. The deliberate routing of operational logs through the White House Counsel or National Security Council channel—rather than standard agency case management systems—characterizes the destruction of those records as the president exercising dominion over personal property, not as obstruction of justice.
This is precisely the lesson the administration appears to have drawn from the Mar-a-Lago prosecution. Judge Cannon dismissed that case. The Gaiser opinion retroactively supplies the constitutional argument that would have made that outcome more secure: the records were never the government’s property to begin with.
The Federal Records Act Problem
The Gaiser opinion creates a constitutional paradox that its author chose not to acknowledge. The Federal Records Act—which has governed agency records management since 1950 and whose constitutional validity has never been the subject of a sustained constitutional challenge—remains in full force. Cabinet departments and their agencies remain subject to its preservation and disclosure requirements. Gaiser’s OLC opinion acknowledges this.
But the FRA applies to agencies created by Congress. The presidency is a constitutional office Congress did not create. The NSC, the White House Office, and other EOP components whose sole function is to advise and assist the president—these are, under the Gaiser framework, constitutionally distinct from the statutory agencies Congress established and can regulate.
The result is a records preservation regime with a deliberate hole at its center. DHS’s operational logs are federal records. The ICE field office’s deployment orders are federal records. But the White House communications directing DHS to conduct those operations, the NSC deliberations authorizing the deployment, the OLC legal opinions generated at the president’s direction as part of his advisory function—as distinct from DOJ’s ordinary agency records, which remain subject to the FRA—justifying the targeting methodology—those are now presidential personal property, subject to destruction at will.
If the Gaiser opinion’s logic were applied consistently, it would also threaten the FRA as applied to EOP components that straddle the agency/advisory boundary—the NSC, the Office of Science and Technology Policy, and OMB divisions that both advise the president and administer statutory programs. The opinion is silent on this problem—perhaps strategically, to avoid exposing the full reach of its logic, or because its authors regard extension to those EOP components as an acceptable consequence rather than an embarrassment. Either way, acknowledging it would expose the administration’s own agencies to the accountability gap it is trying to create only at the presidential level.
The Paper Coup
Justice Holmes wrote that a page of history is worth a volume of logic. The history the Gaiser opinion invokes—two centuries of presidents treating their papers as personal property—is real. So is the history it elides: that the same two centuries produced a Federal Records Act that has never been seriously challenged, a body of case law carefully balancing executive autonomy against institutional accountability, and a democratic tradition in which the peaceful transfer of power depended on the integrity of elections that the president has no constitutional authority to supervise.
The records of what the administration planned, directed, and executed in connection with those elections—the targeting analyses, the deployment orders, the legal justifications, the NSC deliberations—are now characterized as the personal property of one man. They can be destroyed. And destroying them, under the Gaiser framework, is not obstruction of justice.
You cannot easily prosecute what you cannot document. You cannot document what was never required to exist. You cannot hold power accountable when power has declared itself exempt from the only law that required it to leave a record.
That is what an April Fools’ Day OLC legal opinion looks like when it is not a joke.
Eliminating the Presidential Records Black Hole…and More
Thwarting the Gaiser gambit will require congressional action.
First, Congress should enact new legislation—or, at a minimum, a joint resolution—expressly re-asserting that the PRA governs records generated by the president and EOP components, and providing that any presidential directive characterizing such records as personal property under the Gaiser framework is void as against federal records accountability and law enforcement statutes. A mere declaration of constitutional validity carries limited legal weight; a new statutory enactment, grounded in Congress’s unambiguous Article I powers to legislate on preservation of government records and protection of elections, creates a fresh legal basis for court enforcement that does not depend on vindicating the original PRA.
The legislation should explicitly provide that records generated by any federal agency in connection with operations conducted during an election period—regardless of whether those operations are characterized as immigration enforcement, national security operations, or any other mission—are federal records subject to FRA preservation requirements and subject to congressional subpoena without the heightened Mazars standard.
Second, Congress should enact a statutory prohibition on federal law enforcement operations—including immigration enforcement operations—within a defined perimeter of polling places during any federal election, with an explicit private right of action for affected voters and state election officials.
The existing criminal prohibitions in 18 U.S.C. §§ 592 and 594, and the civil remedies in 52 U.S.C. § 10307, have proven insufficient when the executive branch controls prosecutorial discretion. The private right of action provision must also vitiate qualified immunity in order to ensure those responsible for violating the statutes cannot escape accountability for their misconduct.
Third, the Gaiser gambit highlights the continued dangers of allowing the Justice Department to remain within a political branch of government. In the Winter 2026 edition of Free Society, I made the case (in brief) for taking DoJ out of the Executive branch entirely.
Whether that structural separation can be achieved via statutory changes alone—as I believe, by reconstituting DoJ as an independent agency lodged within the federal judiciary, with Senate-confirmed leadership insulated from at-will presidential removal—or whether it would require a constitutional amendment to avoid Appointments Clause and Vesting Clause objections is a genuinely contested structural question. At present, I hold the former view; other thoughtful scholars hold the latter. Either way, the case for removing DoJ from day-to-day control by political partisans has never been stronger.
To be clear, the Gaiser opinion will not, by itself, suppress a single vote. It issues no orders, deploys no agents, and targets no precinct. What it does is remove a layer of accountability infrastructure at the moment when that infrastructure matters most—before an election, when enforcement decisions are being made and documented, or not documented, at the discretion of those making them.
The cumulative effect is worth stating plainly: federal enforcement agencies already possess the operational tools to conduct election-period operations with measurable deterrent effects on voter turnout in targeted communities. Existing statutory prohibitions on voter intimidation and armed federal presence near polling places depend, for their enforcement, on prosecutorial discretion that this administration has shown little inclination to exercise against its own priorities. And federal litigation challenging such operations depends on the existence of records documenting what was planned, authorized, and directed. The Gaiser opinion systematically narrows that last avenue of accountability by characterizing the most consequential records—those generated at the White House and NSC level—as personal presidential property subject to no preservation requirement.
None of this means a voter suppression scheme is imminent or inevitable. But accountability mechanisms are most valuable precisely because they deter conduct that might otherwise occur. A legal architecture that makes misconduct harder to document, harder to litigate, and harder to prosecute does not merely respond to abuses after the fact—it incrementally raises the ceiling on what can be attempted without consequence. The Gaiser opinion raises that ceiling. Congress should lower it.