A second, independent reason for the result. Even setting aside whether the voter file is covered at all, Title III's § 20703 lets the Attorney General compel records only by a written demand that contains “a statement of the basis and purpose” for the request. The court reads that to require both a basis and a purpose in a single demand, and finds none of DOJ's three letters supplied both: the earlier letters stated a basis but no Title III purpose, while the August 14 letter stated a purpose but no basis. Because the requirement is mandatory, the court holds Michigan did not violate Title III by refusing to produce the file. (The dissent reads the letters together as composing one adequate demand.) This procedural holding could recur in DOJ's other pending voter-roll suits.

The Justice Department's Office of Legal Counsel (OLC) is the unit that issues authoritative legal opinions binding within the executive branch. Here, DOJ submitted a newly issued OLC opinion to the court — by the court's account, the day before oral argument and about 66 years after Title III was enacted — concluding that Title III reaches statewide voter-registration lists. Under Loper Bright, an executive interpretation carries the most persuasive weight when it is roughly contemporaneous with the statute and has stayed consistent over time. The court found this one was neither, and rejected it on the merits as well.

This recent Sixth Circuit case is closely related background. In Public Interest Legal Foundation v. Benson, a private organization sued the same Michigan Secretary of State, claiming Michigan failed to adequately maintain its voter rolls under the NVRA. The court rejected that claim, holding that the NVRA requires only a “reasonable effort” — “a serious attempt that is rational and sensible,” not a “perfect, or even optimal” one — and that Michigan's program satisfied it. The majority cites the case here for the factual point that the qualified voter file draws on many data sources unrelated to “acts requisite to voting”; the dissent separately invokes it on the question of DOJ's stated basis for its demand.

Two interpretive canons drive the majority's holding. The harmonious-reading canon directs courts to read overlapping statutes so they work together rather than conflict: reading Title III to “freeze” the voter file (no alterations for 22 months) would collide with the NVRA and HAVA duties to keep updating it, so the court rejects that reading. The surplusage canon, discussed on the next page, holds that statutory words should not be rendered meaningless; the court reasons that reading Title III to reach every record in an official's possession would make the limiting phrase “come into [her] possession” do no work. Together, the two canons support the narrower construction the court adopts.

Loper Bright (2024) is the Supreme Court decision that overruled Chevron v. NRDC and ended “Chevron deference” — the doctrine under which courts deferred to an agency's reasonable interpretation of an ambiguous statute. After Loper Bright, courts must independently determine a statute's “single, best meaning” using the ordinary tools of construction. That framework resurfaces late in this opinion: it is why the court gives no controlling weight to the DOJ Office of Legal Counsel's reading of Title III and instead asks only whether that reading is persuasive — concluding it is not.

A note on posture. The district court dismissed the case on a motion to dismiss — a ruling that the complaint failed as a matter of law, with no factual trial. On appeal the government pressed only its Title III claim, abandoning the NVRA and HAVA theories it had also pleaded below. “De novo” review (stated in the next sentence) means the court of appeals decides the legal question afresh, giving no deference to the district court's reasoning. So this opinion resolves a pure question of statutory interpretation: whether Title III reaches Michigan's voter file at all.

Michigan's “qualified voter file” (QVF) is its centralized statewide voter-registration database. Beyond names and addresses, it contains each registered voter's date of birth, driver's-license or state-ID number, and partial Social Security number, plus voting history and (historically) digitized signatures — the full statutory contents are reproduced in the opinion's footnote 1. Michigan law and the privacy laws of most states restrict disclosure of that sensitive personal information, which is why Benson released only a redacted “public version” of the list rather than the full file. The distinction between the unredacted QVF and the public list is the practical heart of the dispute.

This March 2025 executive order, “Preserving and Protecting the Integrity of American Elections,” directed a range of federal election-related actions, including instructing the Attorney General to prioritize enforcement of laws restricting noncitizen voting and to act against states deemed noncompliant with the NVRA and HAVA's list-maintenance requirements. The order set off DOJ's nationwide demands for state voter data, including the demands to Michigan at issue here. Several of the order's provisions have been challenged in separate lawsuits, and courts have blocked portions of it; those challenges are distinct from this case, which turns only on the scope of Title III.

The Help America Vote Act of 2002 (codified at 52 U.S.C. § 21083) was Congress's response to the disputed 2000 presidential election and the Florida recount that ended in Bush v. Gore. Among other reforms, HAVA required every state to build a “single, uniform, official, centralized” computerized statewide voter-registration list. That mandate is why Michigan's “qualified voter file” exists in its current consolidated form — the very database DOJ sought here. In effect, the record at the center of this case exists largely because a different federal statute required states to create it.

The National Voter Registration Act of 1993 (codified at 52 U.S.C. § 20501 et seq.) is widely known as the “Motor Voter” law because it requires states to offer voter registration at motor-vehicle agencies and other public assistance offices. Alongside expanding registration, the NVRA imposes the voter-list-maintenance duties — removing registrants who have died or moved — that the 2025 executive order and DOJ's demand letters relied on as their justification for seeking the rolls. Those same maintenance duties become central to the majority's reasoning: because the NVRA requires Michigan to keep updating its file, the court concludes Title III cannot also require the state to freeze it.

Title III of the Civil Rights Act of 1960 is codified today at 52 U.S.C. §§ 20701–20706. Enacted five years before the Voting Rights Act of 1965, it was designed to give federal investigators access to local voting records at a time when those records were often the only way to prove racial discrimination in registration. Its central command — the requirement in § 20701 that election officers retain voting records for 22 months after a federal election — still governs federal election-record retention. But the statute was written for a paper-records era, and this case concerns whether it can be used to compel production of a modern, centralized statewide database.

This sentence situates the case in a nationwide effort. Beginning in 2025, and following an executive order on election integrity, the Department of Justice sent letters to election officials in nearly every state and the District of Columbia seeking their complete, unredacted statewide voter-registration lists. Roughly 30 states and D.C. declined, citing state and federal privacy laws, and provided only public versions of their rolls; DOJ then sued more than two dozen of them. By the date of this decision, at least nine federal district courts had dismissed those suits, and none had ordered production. This is the first federal court of appeals to rule, and because Kentucky and Tennessee also sit in the Sixth Circuit, the opinion is binding precedent on DOJ's pending suits there.

A roadmap of the order, in its own terms. Granted: the court preliminarily stays, under 5 U.S.C. § 705, “part (i)” of the Department’s “professional degree” definition and the preamble’s “supervised by a licensed professional … cannot be performed independently” requirement. Denied: (1) the PA plaintiffs’ request to also stay “part (ii)” of the definition; (2) the AANP plaintiffs’ broader request to enjoin enforcement of the statutory loan caps in 20 U.S.C. § 1087e(a)(4) until a new rule issues; and (3) the PA plaintiffs’ request to have PA students treated as “professional students.” The court also directed the parties to file a joint status report and proposed schedule by July 2, 2026, and the order stands until further order of the court. The reasoning is in the separately filed Memorandum Opinion this order accompanies.

“Physician associate” and “physician assistant” name the same credential. In 2021 the American Academy of Physician Associates (AAPA) voted to adopt “physician associate” as the profession’s official title; the change is being recognized state by state and does not alter scope of practice. PAs are licensed clinicians who diagnose, treat, and prescribe. Under the RISE rule, PA programs fall outside the “professional degree” definition and so are subject to the lower graduate borrowing caps. The PA plaintiffs’ specific request here — to have current and incoming PA students treated as “professional students” for loan-limit purposes — is the request this order denies.

Section 1087e sets the terms and conditions of loans under the William D. Ford Federal Direct Loan Program. The order — quoting the plaintiffs’ proposed order — points to subsection (a)(4) as the statutory source of the borrowing caps for graduate and professional students (the $20,500 / $50,000 annual limits added by Public Law 119-21). The citation underlies a question the court pressed at the June 23 hearing: because the dollar caps themselves are fixed by statute, the plaintiffs are not attacking the caps, but the Department’s rule defining who counts as a “professional” student eligible for the higher limit. Consistent with that line, this order denies the AANP plaintiffs’ separate request to enjoin enforcement of the statutory caps.

This requirement appears in the rule’s preamble (91 Fed. Reg. 23787) rather than in the codified text, and it is central to the dispute: nurse practitioner and physician associate practice is, in many states, structured around collaboration with or supervision by a physician. Scope-of-practice rules vary considerably by state, however — some grant nurse practitioners full, independent practice authority, while others require a collaborative or supervisory relationship; physician associate practice likewise ranges from required physician supervision to expanded-autonomy (“PA modernization”) models. The plaintiffs contend this “cannot be performed independently” criterion has no footing in the statute; the Department defends it as a permissible interpretation. It is among the requirements this order preliminarily set aside.

The challenged definition is codified here. Under the rule, a “professional degree” is one awarded in a core list of fields — medicine, osteopathic medicine, podiatry, chiropractic, optometry, pharmacy, dentistry, veterinary medicine, law, theology, and clinical psychology (eleven in all: the ten long enumerated in 34 C.F.R. § 668.2, plus clinical psychology, added during the rulemaking) — or a degree that satisfies a multi-part test. As published, that test requires, among other things, that the degree be “generally at the doctoral level,” require at least six academic years of postsecondary coursework, generally require professional licensure, and carry a four-digit CIP code in the same intermediate group as a listed field. (CIP — Classification of Instructional Programs — is the standardized program taxonomy maintained by the National Center for Education Statistics.) The order acts on “part (i)” and “part (ii)” of this definition; the exact text of each prong, and the effect of the court’s ruling on each, are set out in the accompanying Memorandum Opinion.

The RISE rule implements the student-loan provisions of Public Law 119-21 — the 2025 budget reconciliation law styled in the rule as the “Working Families Tax Cuts Act” and commonly called the “One Big Beautiful Bill Act” (OBBBA). For the first time, that law set separate federal borrowing limits for “graduate” versus “professional” students: professional students may borrow up to $50,000 per year ($200,000 aggregate), while other graduate students are capped at $20,500 per year ($100,000 aggregate). The law also ended the Grad PLUS program for new borrowers, which had let graduate and professional students borrow up to the full cost of attendance. Because the statute ties the higher “professional” limits to the meaning of “professional degree,” the Department’s regulatory definition of that term — the subject of this suit — controls which students reach the higher caps. The rule takes effect July 1, 2026.

The American Association of Nurse Practitioners (AANP) is the largest national professional organization for nurse practitioners. It and co-plaintiffs brought the first of the two consolidated suits (No. 26-cv-1780). The second (No. 26-cv-1941) was filed by the PA Education Association (PAEA) and others; PAEA is the national association of physician assistant/associate educational programs. Both sets of plaintiffs challenge the same Department of Education rule, and the cases were consolidated before Judge Howell. Separately — and not before the court in this order — a coalition of states and a coalition of nursing organizations led by the American Nurses Association have filed their own challenges to the rule.

Section 705 of the Administrative Procedure Act lets a reviewing court, pending judicial review, postpone the effective date of an agency action (or otherwise preserve status and rights) when justice so requires. It is the mechanism the AANP plaintiffs invoked alongside a request for a preliminary injunction. A § 705 stay and a preliminary injunction are related but distinct: a stay suspends the challenged agency action itself, while an injunction is a court order directing a party’s conduct. Courts in this circuit generally evaluate a § 705 stay under the same four factors used for preliminary injunctions — likelihood of success, irreparable harm, the balance of equities, and the public interest.

CoreCivic, Inc. v. Governor of New Jersey, 145 F.4th 315 (3d Cir. 2025), struck down New Jersey's AB 5207 — a law barring contracts for civil immigration detention — but only "as applied to CoreCivic," a private detention contractor, on intergovernmental-immunity grounds (2-1, Judge Bibas writing; Judge Ambro dissenting). The Third Circuit's reasoning distinguished laws that bar private parties from contracting with the federal government in a market where the government is the only buyer (impermissible, per GEO Group, Inc. v. Newsom, 50 F.4th 745 (9th Cir. 2022) (en banc)) from laws that merely restrict a state's own entities while leaving the federal government free to use private contractors (permissible, per McHenry County v. Kwame Raoul, 44 F.4th 581 (7th Cir. 2022)). The court here reads CoreCivic narrowly: it does not invalidate AB 5207 as applied to state and local governments, so AB 5207 remains an independent bar on municipal immigration-detention contracts — a further obstacle to the government's detention-agreement theory of injury.

Redressability is the third element of Article III standing (Lujan v. Defenders of Wildlife, 504 U.S. 555 (1992)): it must be likely, not merely speculative, that a favorable decision will remedy the plaintiff's injury. The principle stated here — that relief is not redressable when an independent, unchallenged source would keep causing the same injury — is the core of the ruling. The court draws it from Fischer v. Governor of N.J., 842 F. App'x 741 (3d Cir. 2021), and the Supreme Court's McConnell v. FEC, 540 U.S. 93 (2003). Applied here: because the ITD independently forbids the same conduct, enjoining only the city policies would leave municipal officers bound by the ITD, so it would produce no practical change.

A Rule 12(b)(1) motion challenging subject-matter jurisdiction can be either a "facial" attack — accepting the complaint's allegations as true and arguing they are jurisdictionally insufficient on their face — or a "factual" attack, which disputes the actual facts underlying jurisdiction and permits the court to weigh evidence outside the pleadings (Const. Party of Pa. v. Aichele, 757 F.3d 347 (3d Cir. 2014)). Because the cities filed pre-answer motions challenging standing, the court treats them as facial attacks and takes the complaint's well-pleaded facts as true. It may still consider matters subject to judicial notice — such as the ITD, the relevant statutes, and the prior court decisions — without converting the motion to one for summary judgment.

Federal Rule of Civil Procedure 25(d) provides that when a public official who is sued in their official capacity leaves office, their successor is automatically substituted as a party — no motion required. That is why the caption and Defendant list reflect current officeholders (Mayors Jabbour, Solomon; council presidents Ramos, Ridley, Mimms) rather than the officials who were in office when the policies issued or when suit was filed. The substitution is purely procedural and says nothing about the merits; it reflects that the suit targets the offices, not the individuals personally.

All three concrete examples of non-cooperation in the complaint involve the Essex County Correctional Facility, which releases detainees despite ICE detainers. But under New Jersey law, jails are run by counties, not municipalities (N.J. Stat. Ann. §§ 30:8-1, 30:8-19), and Essex County is not one of the four defendant cities. Essex County is bound by the ITD and by none of the challenged city policies. The court treats this as significant for standing: the only specific injuries the government identified are traceable to a non-party county facility governed by the unchallenged ITD, not to any defendant's policy — underscoring the traceability and redressability gap.

Intergovernmental immunity is a Supremacy Clause doctrine, traceable to McCulloch v. Maryland, 17 U.S. 316 (1819), that bars states from (1) directly regulating the federal government or (2) discriminating against the federal government or those it deals with. It is distinct from preemption: preemption asks whether a federal statute displaces state law, while intergovernmental immunity asks whether a state law impermissibly targets or controls federal operations. The modern formulation the opinion quotes comes from United States v. Washington, 596 U.S. 832 (2022), and North Dakota v. United States, 495 U.S. 423 (1990). Two of the government's three counts rest on this doctrine.

In United States v. New Jersey, No. 20-1364, 2021 WL 252270 (D.N.J. Jan. 26, 2021), the federal government directly challenged the ITD on the same preemption and intergovernmental-immunity theories it advances against the cities here. Judge Wolfson rejected those arguments, holding (among other things) that the ITD regulates only state and local law enforcement and does not directly regulate or discriminate against the federal government. The government did not seek reconsideration or appeal. The court notes that any future attempt by the United States to relitigate the ITD's validity would have to confront ordinary preclusion principles (Montana v. United States, 440 U.S. 147 (1979)).

This is the Third Circuit decision affirming that the ITD is not preempted by federal law. The opinion was written by Judge Thomas Hardiman. Because it is a precedential Third Circuit decision, it binds the district court: under Third Circuit practice, only the court of appeals sitting en banc or the Supreme Court can overrule it (United States v. Garner, 961 F.3d 264 (3d Cir. 2020)). That binding status is why the court rejects the government's argument that a favorable decision here could eventually unsettle the ITD — a single district judge cannot disturb circuit precedent.

The two prior challenges are central to why this suit cannot indirectly reach the ITD. First, in County of Ocean v. Grewal, 475 F. Supp. 3d 355 (D.N.J. 2020), two counties and others argued the ITD was preempted by 8 U.S.C. §§ 1373 and 1644 (federal information-sharing statutes); Judge Wolfson rejected the challenge, and the Third Circuit affirmed in Ocean County Bd. of Comm'rs v. Att'y Gen. of N.J., 8 F.4th 176 (3d Cir. 2021), on the ground that those statutes regulate only government entities, not private actors, and so cannot preempt. Second, in 2020 the federal government itself directly sued over the ITD on the same preemption and intergovernmental-immunity theories — and lost, without appealing.

Section 287(g) authorizes the Department of Homeland Security to enter written agreements deputizing state and local officers to perform specified federal immigration-enforcement functions under federal supervision. Participation is voluntary — a locality must opt in. The ITD and the challenged city policies both bar New Jersey law enforcement from entering or exercising authority under such agreements (with narrow exceptions). The provision recurs throughout the opinion because the agreements are one of the specific forms of cooperation the policies decline to provide.

The ITD is Attorney General Law Enforcement Directive No. 2018-6, issued by then-Attorney General Gurbir Grewal in November 2018 and amended in September 2019 to add the bar on new Section 287(g) agreements. In New Jersey, the Attorney General is the state's chief law enforcement officer and can issue directives that bind all police departments statewide with the force of law (N. Jersey Media Grp. v. Twp. of Lyndhurst, 229 N.J. 541 (2017)). That statewide binding force is why the directive reaches the defendant cities' own police regardless of the cities' separate policies. (Grewal later became Director of Enforcement at the SEC.)

The Immigrant Trust Directive (ITD) is the linchpin of the entire ruling. It is a statewide law-enforcement directive that independently binds every state, county, and municipal law enforcement agency in New Jersey — including the agencies in the four defendant cities — and imposes substantially the same limits on voluntary immigration-enforcement cooperation as the cities' own policies. Critically, the federal government did not sue over the ITD in this case and New Jersey is not a defendant. The court's logic follows from that gap: because the ITD would keep binding municipal officers even if the city policies were struck down, an injunction against the city policies alone would change nothing on the ground.

"Sanctuary city" is a political label, not a legal term of art, and policies grouped under it vary widely. What they generally share is a refusal to devote local resources to voluntary assistance with federal civil immigration enforcement — they do not, and constitutionally could not, bar federal agents from enforcing federal law themselves. The anti-commandeering principle (Printz v. United States, 521 U.S. 898 (1997); Murphy v. NCAA, 584 U.S. 453 (2018)) holds that the federal government cannot compel states or localities to administer a federal regulatory program, which is the doctrinal backdrop for why a locality may decline to assist. The opinion uses the term descriptively and then sets it aside to focus on standing.

The opinion notes that only one European country — Ukraine — held a TPS designation when the current term began, and that designation had not yet come up for review (it was extended through October 19, 2026; see 90 Fed. Reg. 5936). The only other European states ever designated for TPS were Kosovo and Bosnia-Herzegovina, both during the 1990s Balkans conflicts and both since terminated. This factual point is offered in response to the plaintiffs' argument that no predominantly white nation has had its TPS terminated under the current policy.

The opinion refers to the Chasseurs-Volontaires de Saint-Domingue, a regiment of more than 500 free men of color from the French colony of Saint-Domingue (present-day Haiti) who fought on the American/French side at the 1779 Siege of Savannah. It was the largest unit of soldiers of African descent to serve in the Revolutionary War. Among them, by tradition, was a young drummer named Henri Christophe, who later became a leader of the Haitian Revolution and ultimately King of Haiti. A monument to the regiment was dedicated in Franklin Square, Savannah, in 2007–2009.

The plurality concludes that the statements the plaintiffs cite were "not overtly racial" and could rest on race-neutral justifications, and it declines to reproduce them. The dissent quotes them at length and characterizes them as racially coded, including statements attributed to the President describing Haiti in derogatory terms and associating Haitian immigrants with disease and crime. The factual record on which both opinions draw comes from the district court in the Haiti case (Miot v. Noem), where Judge Ana Reyes found the existing record "strongly suggests" the decision was motivated at least in part by racial animus. Whether those statements establish a discriminatory motive under Arlington Heights is the precise point on which the plurality and dissent divide; it remains unresolved on the merits because this ruling addresses only interim relief.

Arlington Heights (1977) sets the framework for proving that a facially neutral government action was motivated by a racially discriminatory purpose. A challenger need not show race was the sole or even the primary motive — only that it was "a motivating factor." Courts conduct a "sensitive inquiry" into circumstantial and direct evidence, including the historical background, the sequence of events, and contemporary statements by decisionmakers. The plurality and the dissent agree this is the governing test but disagree on its application: the dissent emphasizes that under Arlington Heights race need only be one factor among several, while the plurality concludes the cited statements are insufficient to show race was a motivating factor at all.

Trump v. Hawaii (2018) upheld the third version of the travel ban restricting entry from several countries, most of them majority-Muslim. The Court there applied a highly deferential rational-basis review, asking only whether the policy was "plausibly related" to a legitimate government objective. The Government argued that same deferential standard should govern here; the plaintiffs argued Hawaii applies only to the exclusion of aliens seeking to enter, not to those already lawfully present. The plurality sidesteps the dispute by assuming (without deciding) that the more demanding Arlington Heights standard applies and finding the claim fails even under it.

This is the clear-statement rule from Webster v. Doe, 486 U.S. 592 (1988): a statute will not be read to bar judicial review of constitutional claims unless Congress's intent to do so is clear. The plurality declines to decide whether the TPS bar meets that standard, resolving the equal protection claim on the merits instead. (Justice Thomas, concurring separately, would hold that the bar does clearly preclude even constitutional claims — but his is a one-Justice view and is not part of the Court's holding.)

In McNary (1991), the Court held that a statutory bar on review of "a determination respecting an application" under a farmworker amnesty program did not block broad challenges to the procedures used to administer the program, because "determination" there described a single act — a ruling on an individual application. Both the majority and the dissent cite McNary: the dissent reads it as authority that "determination" means a discrete decision, not the surrounding process; the majority distinguishes it as turning on the specific narrower wording of that different statute. The case recurs throughout this opinion as the central precedent the two sides read oppositely.

The majority leans on Patel v. Garland, 596 U.S. 328 (2022), for the principle that the phrase "with respect to" broadens a statute's coverage to matters relating to its subject, not just the subject itself. Patel was itself an immigration case construing a different jurisdiction-stripping provision (8 U.S.C. § 1252(a)(2)(B)) and reading it broadly to bar review. The dissent's footnote 1 contests this move, arguing that "with respect to" modifies "determination" in a way that does the majority "no good" if a determination is only the final decision.

In 2025, the Court twice stayed district court orders that had blocked the termination of Venezuela's TPS designation — referred to here as NTPSA I and NTPSA II. Those were emergency-docket ("shadow docket") stay orders issued without full briefing or a signed majority opinion, so they set no binding precedent on the merits. The Government argued the lower courts in these Syria and Haiti cases should have followed NTPSA; the plaintiffs argued unreasoned stay orders do not control. This decision is the Court's first full-opinion ruling on the TPS judicial-review bar.

Executive Order 14159, issued January 20, 2025 (90 Fed. Reg. 8443), directed Cabinet officers to ensure TPS designations are "appropriately limited in scope" and last "only so long as may be necessary" under the statute. The order is the policy origin of the across-the-board review that followed: under it, the Secretary terminated every TPS designation that came up for renewal — 13 of the 17 then-existing designations. The order's framing of immigration as an "invasion" is itself cited by the Miot plaintiffs as context for their equal protection claim.

Somalia was designated for TPS in 1991 and the designation has been continuously extended since. Other long-running designations the opinion cites include Nicaragua and Honduras (1999) and El Salvador (2001), each over 25 years old. The recurring extensions reflect that the statute requires re-designation or extension whenever country conditions still meet the criteria; critics across administrations have noted that "temporary" protected status has in practice often become indefinite.

This is 8 U.S.C. § 1254a(b)(5)(A), the provision the entire case turns on. In full, it states: "There is no judicial review of any determination of the [Secretary of Homeland Security] with respect to the designation, or termination or extension of a designation, of a foreign state under this subsection." The dispute is over the scope of "determination" — whether it reaches only the Secretary's ultimate decision about country conditions (the plaintiffs' and dissent's reading) or also the procedural steps leading up to it, including the consultation requirement (the majority's reading).

The statute (8 U.S.C. §§ 1254a(b)(2)–(3)) requires the Secretary to review each TPS designation at least 60 days before the end of its current period, which can run from 6 to 18 months. At each review the Secretary must consult appropriate agencies, evaluate whether conditions still meet the statutory criteria, and either extend or terminate. If the Secretary takes no action, the designation automatically extends for six months. This mandatory periodic-review structure is the procedural framework the plaintiffs say was not followed — specifically the consultation step.

Before TPS existed, the Executive Branch granted similar humanitarian relief on a purely discretionary basis through "Extended Voluntary Departure," beginning around 1960. Under EVD there were no statutory standards governing when relief was granted or withdrawn, and the D.C. Circuit had concluded the decisions were unreviewable exercises of prosecutorial discretion (Hotel & Restaurant Employees Union, Local 25 v. Smith, 846 F.2d 1499 (CADC 1988)). The opinion recounts this history because Congress created TPS specifically to add standards to that previously standardless regime — a point both the majority and dissent draw on for opposite conclusions about reviewability.

Congress created TPS in the Immigration Act of 1990 (Pub. L. 101–649), codified at 8 U.S.C. § 1254a. It allows the Secretary of Homeland Security (originally the Attorney General) to designate a country whose nationals already in the United States cannot safely return because of armed conflict, environmental disaster, or other extraordinary conditions. Designated nationals receive protection from removal and work authorization, but TPS confers no path to permanent residence or citizenship on its own. As of the start of the current administration, 17 countries held TPS designations.

The cases are captioned "Mullin v. Doe" because Markwayne Mullin is the current Secretary of Homeland Security and is automatically substituted as the named party under the rule for suits against officials in their official capacity (Fed. Rule Civ. Proc. 25(d)). The actual termination decisions at issue, however, were made by his predecessor, Kristi Noem, who served as Secretary when the Syria and Haiti TPS designations were terminated in 2025. The opinion refers throughout to Noem because she signed the determinations being challenged.

“Non-refoulement” is the core protection of international refugee law: the principle that a state may not return a refugee to a territory where their life or freedom would be threatened on account of a protected ground. The French term “refouler” — meaning to repel, drive back, or turn away — appears in the authoritative text of Article 33.1 alongside “expel or return.” Its precise scope was the crux of Sale: the majority in that case read “return (‘refouler’)” narrowly, as referring to the exclusion of those already at or within a nation's borders rather than action taken extraterritorially. How far the non-refoulement duty reaches — and whether physically blocking someone at the threshold of a port of entry implicates it — is the unresolved question the dissent presses here.

The 1951 Convention Relating to the Status of Refugees is the foundational international treaty defining who qualifies as a refugee and the protections owed to them. It was drafted in the aftermath of World War II and initially limited to European refugees displaced before 1951; the 1967 Protocol Relating to the Status of Refugees removed those geographic and temporal limits and extended the Convention's core protections globally. The United States is not a party to the 1951 Convention directly but acceded to the 1967 Protocol in 1968, thereby binding itself to Articles 2 through 34 of the Convention — including the Article 33 bar on refoulement at issue here. Congress enacted the Refugee Act of 1980 to bring U.S. statutory law into conformity with these obligations.

In Sale (1993), the Supreme Court held 8–1 (Stevens, J.; Blackmun, J., dissenting) that neither the INA's non-return provision nor Article 33 of the Refugee Convention applied to the Coast Guard's interdiction and forced repatriation of Haitian asylum seekers intercepted on the high seas. The Court reasoned that Article 33's bar on “refoulement” does not govern a nation's conduct toward people outside its own territory. The majority here treats that holding as foreclosing respondents' treaty-based argument: because noncitizens standing in Mexico are outside U.S. territory, declining to admit them does not violate Article 33. The dissent distinguishes Sale on the ground that it concerned conduct on the high seas, not U.S. officers physically stationed on U.S. soil blocking entry — a fact pattern Sale did not address.

The presumption against extraterritoriality is a canon under which courts assume that federal statutes apply only within U.S. territory unless Congress clearly indicates otherwise. The Court applies a two-step framework (from Morrison and RJR Nabisco): first ask whether the text unmistakably reaches foreign conduct; if not, ask whether the statute's “focus” is on domestic or foreign conduct. The majority reasons that reading the asylum and inspection duties to attach while a person is still in Mexico would give the provisions extraterritorial reach, so the presumption counsels against that reading. The dissent's rejoinder is that all the relevant conduct — U.S. officers, standing on U.S. soil, allegedly withholding inspection — is domestic, so the presumption does no work here.

The canon against surplusage is the interpretive presumption that every word in a statute should be given effect, so that no clause is rendered redundant or meaningless. Respondents argued that if “arrives in the United States” requires physical presence, it adds nothing to the neighboring phrase “physically present in the United States,” making it surplusage — so “arrives in” must reach people not yet inside. The majority responds that the canon is “not an iron rule,” that Congress sometimes enacts overlapping language, and that any redundancy here is partial and explainable (it links the asylum provision to the newly created expedited-removal trigger). The dissent counters that the canon is at its strongest when one reading would render another part of the same sentence superfluous.

IIRIRA was a sweeping 1996 overhaul of federal immigration law. Among other changes, it abolished the prior distinction between “exclusion” proceedings (for those seeking entry) and “deportation” proceedings (for those already inside), merging them into a single “removal” process; it created expedited removal; and it made “admission” — defined as lawful entry after inspection — the pivotal concept in the statute. Both provisions at issue in this case took their current form in IIRIRA, which replaced the Refugee Act's phrase “physically present in the United States or at a land border or port of entry” with the current “arrives in the United States.” Whether that wording change narrowed the provision's reach is the central interpretive dispute between the majority and the dissent.

The majority cites Keene for a well-known interpretive canon usually traced to Russello v. United States, 464 U.S. 16 (1983): when Congress includes particular language in one section of a statute but omits it from another, the disparate inclusion or exclusion is presumed to be intentional. Here the Court applies it to the observation that other INA provisions expressly reach “attempts” to enter (e.g., the criminal entry statutes), while §§ 1158(a)(1) and 1225(a)(1) refer only to those who “arrive” — supporting the inference that Congress did not mean to cover attempted entry in the asylum and inspection provisions. The dissent's response is that Congress's use of different words elsewhere does not foreclose using different language to reach the same result.

Federal appeals are normally decided by three-judge panels. “En banc” rehearing is a rare procedure in which a larger group of the circuit's judges reconsiders a panel decision — in the Ninth Circuit, an 11-judge en banc court rather than the full bench, because the court is so large. A party can petition for en banc review, but it is granted only when a majority of active judges vote for it, typically for questions of exceptional importance or to resolve intra-circuit conflicts. Here the court declined to rehear the case, but 12 judges publicly recorded their disagreement with that denial — an unusually large dissent that often signals a circuit split or an issue ripe for Supreme Court review.

This is the Ninth Circuit decision under review (decided 2025). A divided three-judge panel affirmed the District Court, holding that a noncitizen “arrives in the United States” — triggering the duties to inspect and to allow an asylum application — when, standing on the Mexican side of the border, the person encounters a U.S. official. Judge Ryan Nelson dissented. The full Ninth Circuit then declined to rehear the case en banc, over the recorded dissent of 12 judges (Judge Bress, joined by 11 active judges, plus Judge Bea and two senior judges). “138 F. 4th 1102” is the citation to volume 138 of the Federal Reporter, Fourth Series, page 1102.

A “putative” class action is one filed on behalf of a proposed class that has not yet been certified by the court. Certification under Federal Rule of Civil Procedure 23 is a separate step requiring the plaintiffs to show, among other things, that the class is sufficiently numerous, shares common questions, and is adequately represented. Until the court certifies, the case is “putative.” Here the District Court did later certify a class — defined as noncitizens who sought or will seek asylum at certain U.S.-Mexico ports and were or will be denied access — which is what made the declaratory relief class-wide in effect.

“Metering” (also called “queue management”) refers to the practice, formalized at the southwest border beginning in 2016, of limiting how many asylum seekers CBP would process at ports of entry on a given day. Officers physically stationed at or near the border line turned back arrivals once a daily threshold was reached. The policy operated under two presidential administrations before a third rescinded it in 2021. According to the document, the Government's stated rationale was port capacity and safety; the dissent points to an inspector-general finding that metering was often invoked regardless of actual capacity. The named defendant, Markwayne Mullin, is sued in his official capacity as Secretary of Homeland Security, so the case caption reflects the office rather than any individual conduct.

These are the five “protected grounds” that define a refugee under 8 U.S.C. § 1101(a)(42)(A): race, religion, nationality, membership in a particular social group, and political opinion. The list is drawn directly from the 1951 Refugee Convention and was incorporated into U.S. law by the Refugee Act of 1980. To qualify for asylum, an applicant must show persecution (or a well-founded fear of it) “on account of” one of these five grounds — meaning generalized hardship, crime, or poverty, without a nexus to a protected ground, does not suffice. “Particular social group” is the most litigated of the five, because its boundaries are not defined by statute.

A “credible fear” interview is the first screening step for an arriving noncitizen who seeks asylum while in expedited removal, conducted by a trained asylum officer rather than an immigration judge. The standard is whether there is a “significant possibility” that the person could establish eligibility for asylum (8 U.S.C. § 1225(b)(1)(B)(v)) — a deliberately low threshold meant to screen out only clearly meritless claims and to funnel plausible ones into full proceedings. A negative finding can be reviewed by an immigration judge on a “prompt” and limited basis; a positive finding results in detention for further consideration of the asylum application.

Expedited removal is a streamlined deportation process Congress created in IIRIRA (1996), codified at 8 U.S.C. § 1225(b)(1). It allows immigration officers to order certain arriving noncitizens removed without a hearing before an immigration judge and without the procedural protections of ordinary removal proceedings. It typically applies to people who lack valid entry documents or who are found to have committed fraud or misrepresentation. The major exception, central to this case, is that an officer may not proceed with expedited removal if the noncitizen expresses an intention to apply for asylum or a fear of persecution — at which point the “credible fear” screening process begins.

After Bruen named six jurisdictions with "proper cause"-style carry regimes (Hawaii, California, the District of Columbia, Maryland, Massachusetts, and New Jersey), several responded with the same "default flip" Hawaii adopted — barring firearms on private property open to the public absent the owner's express consent. The opinion cites parallel statutes in California, Maryland, New Jersey, and New York. Because those laws share Hawaii's structure, this decision's holding has direct implications beyond Hawaii: the same reasoning would bear on the comparable provisions in those States. The Ninth Circuit had in fact reviewed Hawaii's and California's versions together below.

This figure describes Hawaii's licensing regime before Bruen, when state law allowed carry licenses only in an "exceptional case." The "four licenses" statistic comes from Young v. Hawaii, Ninth Circuit litigation over that prior regime. That "exceptional case" standard is the same type of "may-issue" discretionary scheme that Bruen held unconstitutional in 2022 — which is why Hawaii had to replace it. Understanding how restrictive the old system was clarifies the opinion's framing that Hawaii responded to Bruen by "replacing its old law on carry permits with new laws that achieved a similar result." Note: the practical scarcity of permits under the old regime is the State's point too — it is the backdrop for Hawaii's argument that residents never expected armed carry.

The plaintiffs sued under 42 U.S.C. § 1983, the federal statute that lets individuals sue state officials for violating constitutional rights, naming Hawaii's Attorney General. Procedurally, this case reaches the Court at the preliminary-injunction stage — a request to block enforcement of the law while the litigation proceeds, not a final judgment after trial. The District Court enjoined the rule as applied to private property open to the public; a Ninth Circuit panel reversed that injunction; and the full Ninth Circuit denied rehearing en banc ("en banc" meaning reheard by an expanded bench of the court rather than the usual three-judge panel), over a dissent by Judge VanDyke joined by five other judges. The Supreme Court granted review limited to the private-property-open-to-the-public question.

Caetano v. Massachusetts (2016) was a Second Amendment case decided by an unsigned ("per curiam") opinion in which the Court unanimously vacated the conviction of Jaime Caetano, a woman who carried a stun gun to protect herself from an abusive ex-boyfriend and was prosecuted under a Massachusetts law banning stun guns. The Court held that the Second Amendment extends to "all instruments that constitute bearable arms," including those not in existence at the founding. Justice Alito — the author of the present opinion — wrote a separate concurrence in Caetano (joined by Justice Thomas) detailing Caetano's circumstances; the opinion here draws its extended hypothetical about a "young woman" running daily errands directly from that concurrence.

This 1865 Louisiana statute is the analogue the majority calls Hawaii's "most remarkable." It was enacted as part of Louisiana's Black Code — the body of post-Civil War laws Southern States passed to control the newly freed Black population. The majority gives two reasons for rejecting it: first, it was neither widespread nor widely accepted (and so carries little weight under Bruen regardless of its content); and second, as a tool designed to disarm Black citizens, it cannot, in the majority's view, illuminate the original understanding of the right to keep and bear arms. Justice Jackson's dissent devotes a full section to this statute, arguing that the majority cannot both make history the test and then categorically exclude an uncomfortable part of the historical record. "Black Codes" were the racially discriminatory state laws enacted in 1865–66 across the defeated Confederacy.

This 1721 Pennsylvania statute is the lead example in a cluster of colonial-era laws (others cited are a 1722 New Jersey act, a 1728 Maryland act, a 1763 New York act, and a 1771 New Jersey act) that restricted carrying guns onto another person's "improved or inclosed" land without permission. The legal fight is over what principle these laws stand for. The majority reads them narrowly, as anti-poaching measures aimed at the specific harms of unauthorized hunting (theft of game, stray gunfire, property damage), and concludes they are too different from Hawaii's general rule. The dissents read them at a higher level of generality, as laws protecting a landowner's right to exclude armed entrants — a principle they say Hawaii's law shares. Both sides agree these are the strongest analogues Hawaii offers.

A "historical analogue" is an older law — typically from the founding era (around 1791, when the Second Amendment was ratified) or Reconstruction (around 1868, when the Fourteenth Amendment was ratified) — that a court compares to a modern regulation to decide whether the modern law fits within an accepted tradition. Under Bruen, the government evaluates analogues along three lines: how many jurisdictions adopted them, how well-accepted they were, and whether they are "relevantly similar" to the modern law in terms of "how" and "why" they burdened the right. The bulk of this opinion is an argument over whether Hawaii's proffered analogues — colonial anti-poaching statutes and a Reconstruction-era Louisiana law — are close enough to its modern rule to count.

This is the analytical structure at the center of the case. Step one asks whether the Second Amendment's "plain text" covers the conduct the law restricts (does it burden "the people" in "keeping" or "bearing" "Arms"?). If so, the law is "presumptively unconstitutional." Step two shifts the burden to the government to show the law is "consistent with the Nation's historical tradition of firearm regulation" by identifying historical analogues. Much of the disagreement among the opinions in this case is about which step does the work: the majority resolves step one easily and decides the case at step two, while the principal dissent argues the challenge should fail at step one because there is no right to enter private property without consent in the first place.

United States v. Rahimi (2024) was the Court's most recent Second Amendment decision before this one. By an 8–1 vote, with Chief Justice Roberts writing and Justice Thomas the lone dissenter, the Court upheld 18 U.S.C. § 922(g)(8), the federal statute barring a person subject to a domestic-violence restraining order from possessing a firearm. Rahimi is significant to this opinion because it clarified how the Bruen historical test operates when a modern law addresses a problem the founders did not face: the government need not produce a "historical twin," only a "relevantly similar" analogue consistent with "the principles that underpin our regulatory tradition." Both the majority and the dissents here invoke Rahimi's "principles, not a mold" approach — they disagree over how to apply it.

New York State Rifle & Pistol Assn. v. Bruen (2022) held, 6–3, that the Second Amendment protects a right to carry a handgun in public for self-defense, striking down a New York law that required applicants to show "proper cause" to obtain a carry license. Justice Thomas wrote for the majority. Bruen is the case the present opinion applies: it set out the two-step "text-and-history" framework — first ask whether the conduct falls within the Second Amendment's plain text, then ask whether the government can point to a historical tradition of analogous regulation. Hawaii enacted the law challenged here in direct response to Bruen, which had specifically named Hawaii as one of six jurisdictions with a "proper cause"-type regime.

McDonald v. Chicago (2010) "incorporated" the Second Amendment against the States — that is, it held the right applies to state and local governments, not just the federal government, by operation of the Fourteenth Amendment. The vote was 5–4, but the majority split on reasoning: a four-Justice plurality (Alito, writing, joined by Roberts, Scalia, and Kennedy) located the right in the Fourteenth Amendment's Due Process Clause, while Justice Thomas concurred only in the judgment, providing the fifth vote on the theory that the right is secured instead by the Amendment's Privileges or Immunities Clause. "Incorporation" is the doctrine by which Bill of Rights protections, which originally restrained only the federal government, are applied to the States.

District of Columbia v. Heller was the Supreme Court's first decision to hold that the Second Amendment protects an individual right to keep and bear arms, independent of service in a militia. By a 5–4 vote, with Justice Scalia writing for the majority, the Court struck down a District of Columbia law that effectively banned handgun possession in the home. The case was brought by Dick Heller, a D.C. special police officer who carried a handgun on duty but was denied a permit to keep one at home. Heller is the foundation for every later Second Amendment case discussed in this opinion; its instruction to interpret the right "by looking to history" became the basis for the test the Court applies here.

Loper Bright Enterprises v. Raimondo (2024) is the decision in which the Supreme Court overruled Chevron deference, ending the doctrine that courts defer to an agency's reasonable interpretation of an ambiguous statute it administers. Durnell invokes it to argue that EPA's pesticide-labeling regulations exceed the agency's actual statutory authority under FIFRA and so should not receive preemptive effect. The majority rejects the argument, holding that FIFRA's text itself — not deference to EPA — directs the agency to make registration and labeling determinations, so Loper Bright does not undermine those regulations.

This is one of EPA's ongoing oversight tools. FIFRA requires EPA to formally re-evaluate every registered pesticide at least once every 15 years through "registration review," reassessing whether the product still meets the statutory standard in light of current science. It operates alongside EPA's other continuing powers — requiring new data, mandating label changes, and cancelling or suspending registrations. The majority catalogs these tools to rebut the argument that preemption leaves the public unprotected against new safety information: EPA, not state juries, is the designated mechanism for updating pesticide labels over time.

Monsanto, the original developer of Roundup, was acquired by the German pharmaceutical and life-sciences company Bayer AG in 2018. Bayer inherited the glyphosate litigation along with the acquisition and has since faced tens of thousands of Roundup cancer claims and several large jury verdicts. The corporate relationship surfaces again in the dissent, which notes that a Bayer subsidiary is among the manufacturers that have previously added state-specific cancer warnings to pesticide labels as "minor modifications" — a fact the dissent uses to dispute the majority's claim that such warnings require EPA's prior approval.

This is the circuit split the Court took the case to resolve. The federal and state courts had divided over whether FIFRA preempts a Roundup failure-to-warn claim. The Third Circuit in Schaffner (2024) held such claims preempted; the Eleventh and Ninth Circuits (Carson, Hardeman) and several state appellate courts held they were not. As the dissent emphasizes, the great majority of courts to reach the question had rejected preemption — making the Third Circuit's Schaffner decision the outlier that the majority now effectively vindicates.

This citation flags an important wrinkle: EPA's 2020 interim glyphosate decision — one of the determinations the majority relies on — was vacated by the Ninth Circuit in 2022. In Natural Resources Defense Council v. EPA, the court found EPA had not adequately supported its conclusion that glyphosate poses no unreasonable risk to human health and sent the human-health portion back to the agency for further analysis. The opinion cites this vacatur in a string citation; it indicates that EPA's no-cancer-warning position, while longstanding, has not gone unchallenged in the federal courts.

The IARC classification is the scientific event that triggered the Roundup litigation wave. In March 2015, the International Agency for Research on Cancer — the World Health Organization's cancer research arm — classified glyphosate as "probably carcinogenic to humans" (Group 2A), citing limited human evidence and sufficient animal evidence, and noting an observed association with non-Hodgkin's lymphoma specifically. That conclusion diverged from EPA's longstanding position and from other national regulators. IARC's classification is a hazard assessment (whether a substance can cause cancer under some conditions), which differs methodologically from the risk assessment EPA conducts. The opinion notes EPA re-examined glyphosate in 2017 and 2019 after IARC but adhered to its prior view.

The 1972 Federal Environmental Pesticide Control Act is what transformed FIFRA from a labeling statute into the comprehensive regulatory regime at issue here. It moved pesticide authority to the newly created EPA (established 1970), required EPA to affirmatively find that a pesticide "will not generally cause unreasonable adverse effects" before registering it, and strengthened EPA's enforcement powers. The majority leans on this transformation to argue that modern EPA registration reflects a substantive safety judgment — unlike the near-automatic registration under the 1947 Act.

FIFRA was first enacted in 1947, but in a much weaker form than exists today. The original Act required pesticides to be registered with the Secretary of Agriculture, but gave the Secretary an essentially passive role — registration was largely mandatory and did not signal that a product met any safety standard. The majority and dissent both draw on this history: the dissent notes that § 136a(f)(2) (registration is only "prima facie" evidence of compliance) is a vestige of this earlier era, while the majority argues that the 1972 overhaul fundamentally changed the regulatory picture.

This is the statutory core of the case. Under FIFRA, a pesticide is "misbranded" — and therefore illegal to sell — if its label is false or misleading, or if it omits a warning "necessary and adequate to protect health and the environment." 7 U. S. C. §§ 136(q)(1)(A), (G). The misbranding prohibition is the requirement the dissent says Durnell's failure-to-warn claim simply parallels (and therefore should survive preemption), while the majority locates the operative federal requirement instead in EPA's specific approval of Roundup's label. Nearly every argument in both opinions traces back to what "misbranded" means and who decides when a label meets the standard.

The Medical Device Amendments of 1976 (MDA) amended the Food, Drug, and Cosmetic Act to give the FDA authority over medical devices, including a premarket approval process for high-risk (Class III) devices. Its express preemption clause, 21 U. S. C. § 360k(a), bars states from imposing device requirements "different from, or in addition to" federal requirements — wording that closely tracks FIFRA's § 136v(b). Because the two clauses are textually similar, the majority treats Riegel's interpretation of the MDA clause as controlling here. The dissent's key counterpoint is that the MDA contains no provision analogous to FIFRA's § 136a(f)(2).

Riegel v. Medtronic, Inc. (2008) is the precedent the majority calls dispositive. It interpreted the preemption clause of the Medical Device Amendments of 1976 — language the Court describes as nearly identical to FIFRA's — and held that the FDA's premarket approval of a medical device imposes federal "requirements" that preempt state tort claims demanding additional or different safety measures. The case involved a balloon catheter that ruptured during a coronary angioplasty. The majority reasons by analogy that EPA's pesticide registration works the same way; the dissent argues the two statutes differ in a way that breaks the analogy.

Bates v. Dow Agrosciences LLC (2005) is the Supreme Court's foundational interpretation of FIFRA's preemption clause and the case both the majority and dissent treat as the central precedent. The Court there held that FIFRA preempts state labeling requirements that are "in addition to or different from" federal ones, but that state-law claims "equivalent to" FIFRA's own misbranding requirements survive. Bates arose from Texas peanut farmers who alleged a Dow herbicide damaged their crops; the Court remanded for the lower courts to decide whether the specific failure-to-warn claims were equivalent to FIFRA's standards. The majority and dissent in this case divide sharply over what Bates requires.

This references Trump v. CASA, Inc., 606 U.S. 831 (2025), the Supreme Court decision that sharply limited the ability of federal district courts to issue "universal" (nationwide) injunctions. After CASA, injunctive relief generally must be limited to what is necessary to give the actual plaintiffs "complete relief," rather than barring the government's conduct against everyone everywhere. That rule shapes the remedy section here: the court tailors the § 7(a), § 7(b), and § 4(a) injunctions to the plaintiff States only. For §§ 2(a) and 3(d), however, the court still enjoins the provisions nationwide — reasoning that because they govern a single national Federal Form and Federal Post Card Application, the plaintiff States cannot obtain complete relief unless the form-related provisions are blocked in full. The First Circuit's post-CASA decision in Doe v. Trump supports that a broader injunction remains permissible where a narrower one would leave the plaintiffs' harms unremedied.

This is the most significant "what to watch" footnote in the opinion. In Republican National Committee v. Wetzel, 120 F.4th 200 (5th Cir. 2024), the Fifth Circuit held that the federal Election Day statutes (2 U.S.C. § 7; 3 U.S.C. § 1) require ballots to be received by Election Day, not merely cast by then — meaning state laws counting late-arriving but timely-postmarked ballots conflict with federal law. That reading is the legal theory underlying §§ 7(a) and 7(b) of the Executive Order. This court, like the D.C. and W.D. Washington courts, declines to follow Wetzel. But the Supreme Court has granted certiorari (sub nom. Watson v. Republican National Committee), so the question whether states may count ballots received after Election Day is now pending before the Justices — and a ruling there could directly affect the § 7 holdings in this case.

Purcell v. Gonzalez is the source of the "Purcell principle" — the idea that courts should be cautious about changing election rules in the period close to an election, because late changes risk voter confusion and administrative disruption. It is most often invoked to counsel against judicial intervention shortly before an election. Here, the court cites it for a related point in the opposite direction: the imminence of the upcoming midterm cycle weighs against staying the case, because leaving the rules unsettled would itself create the confusion Purcell warns about. The court pairs it with Justice Kavanaugh's observation that running a statewide election is a "massive coordinated effort" requiring clear and settled rules.

This phrase invokes the most influential framework in separation-of-powers law: Justice Robert Jackson's concurrence in Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579 (1952) — the "Steel Seizure" case. Jackson described three tiers of presidential power. Presidential authority is at its maximum when the President acts with congressional authorization; in a "zone of twilight" when Congress is silent; and "at its lowest ebb" when the President acts against the expressed or implied will of Congress. In that third category, the President may rely only on his own constitutional powers minus any constitutional powers of Congress over the matter. By placing § 2(a) at the "lowest ebb," the court signals that the EO conflicts with Congress's will (as expressed in the NVRA) and can survive only if grounded in some exclusive presidential power — which the court finds absent.

"Nonstatutory review" refers to a federal court's inherent equitable power to review and enjoin unlawful government action even when no statute expressly provides a cause of action. Its roots trace to Ex parte Young (1908) and the line of cases recognizing implied equitable actions "directly under the Constitution." The jurisdictional anchor is the general federal-question statute, 28 U.S.C. § 1331. As the court explains, equitable ultra vires claims based purely on statutory violations "rarely succeed," but courts afford broader latitude where, as here, the plaintiff alleges a constitutional violation. This distinction — statutory versus constitutional — is why the court spends several pages characterizing the States' claims as constitutional separation-of-powers challenges rather than mere statutory disputes.

"Ultra vires" is Latin for "beyond the powers." In administrative and constitutional law, an ultra vires claim asserts that a government official acted outside the bounds of any authority granted to them — either by the Constitution or by statute. The doctrine matters here because the States are not suing under the Administrative Procedure Act (the usual vehicle for challenging agency action). The APA does not reach the President directly, because the President is not an "agency" and an executive order is not "final agency action." The States therefore rely on a "nonstatutory" equitable cause of action — a residual judicial power to enjoin unlawful executive conduct — which the court holds is available because their claims are constitutional in nature rather than merely statutory.

The "shall accept and use" command (52 U.S.C. § 20505(a)(1)) is the statutory hook for much of this case. In Arizona v. Inter Tribal Council of Arizona (ITCA), 570 U.S. 1 (2013), the Supreme Court explained that while states may design their own registration forms, the Federal Form serves as a "backstop": no matter what hurdles a state's own form imposes, the Federal Form guarantees a simple national means of registering for federal elections. Because the form is mandatory and national, the court later reasons, an unlawful change to it cannot be cured by a partial, state-by-state injunction — every covered state is bound to use whatever the single form says.

The Help America Vote Act was enacted in 2002 in the wake of the disputed 2000 presidential election. It created the Election Assistance Commission (EAC) — a four-member, bipartisan independent agency — and transferred to it the responsibility for maintaining the Federal Form (previously held by the Federal Election Commission). HAVA also established the federal funding program for election administration that the EAC distributes to the states. The EAC's bipartisan, independent structure is central to the court's reasoning: the opinion treats a presidential command dictating the outcome of EAC rulemaking as an intrusion on an entity Congress deliberately insulated from direct executive control. The current commissioners named as defendants are Donald Palmer, Thomas Hicks, Christy McCormick, and Benjamin Hovland.

The National Voter Registration Act of 1993 (52 U.S.C. §§ 20501 et seq.), often called the "Motor Voter" law, established baseline voter-registration procedures every covered state must offer — registration by mail, alongside driver's-license applications, and at designated agencies such as public-assistance offices. It also created the national mail voter registration form (the "Federal Form"), which states subject to the NVRA must "accept and use." Critically, the statute provides that the Federal Form "may require only such identifying information . . . as is necessary" to assess eligibility, and sets a specific procedure — consultation with state election officials plus notice-and-comment rulemaking — for changing its contents. Sections 2(a) and 4(a) of the Executive Order are evaluated against these requirements.

The Uniformed and Overseas Citizens Absentee Voting Act (52 U.S.C. §§ 20301 et seq.) governs absentee voting for military service members, their families, and U.S. citizens living abroad. It created the Federal Post Card Application (FPCA) — the single national form used to both register and request an absentee ballot — and requires states to accept it. UOCAVA verifies citizenship through attestation rather than documentary proof. By statute, the Secretary of Defense (here, defendant Pete Hegseth) is the "presidential designee" responsible for promulgating the FPCA. Section 3(d) of the challenged Executive Order, which would require documentary proof of citizenship on the FPCA, is measured against this statute.

This quotation comes from Trump v. United States, 603 U.S. 593 (2024) — the Supreme Court's presidential-immunity decision. The cited language appears in the portion of that opinion describing the limits of presidential authority over elections: the Constitution vests "executive Power" in the President and directs him to "take Care that the Laws be faithfully executed," but it grants him no specific powers over the administration of elections. The court here invokes this passage to establish a baseline — that whatever authority the Executive Order claims, it cannot derive from any direct constitutional role of the President in election administration.

The Elections Clause (U.S. Const. art. I, § 4, cl. 1) is the provision at the center of this entire dispute. It assigns the states the power to set the "Times, Places and Manner" of congressional elections, while giving Congress the power to "make or alter" those state rules. The Supreme Court has described the Clause as functioning like a "default provision": states run the mechanics of federal elections unless and until Congress steps in to preempt them. Notably, the President appears nowhere in the Clause — a structural omission the court returns to throughout this opinion. For presidential elections, a parallel provision, the Electors Clause (art. II, § 1, cl. 2), gives states the primary authority to decide how electors are chosen.

The brief quotes this D.C. Circuit decision for the proposition that the power over the purse was among the most important authorities the Constitution assigned to Congress. The opinion was authored by then-Judge Brett Kavanaugh, who joined the Supreme Court in 2018; the case held that, under federal appropriations law, a Navy facility generally could not spend appropriated funds on bottled water when safe tap water was available. The brief’s parenthetical “(Kavanaugh, J.)” signals the authorship deliberately — citing a now-sitting Justice’s appropriations-law reasoning lends weight to the amici’s Spending Clause argument. The quoted language itself derives from Federalist No. 51 (James Madison).

The “power of the purse” refers to Congress’s control over federal spending, grounded in Article I — the Vesting Clause (§ 1), the Spending and related powers (§ 8), and the Appropriations Clause (§ 9, cl. 7: “No Money shall be drawn from the Treasury, but in Consequence of Appropriations made by Law”). The settled principle is that money cannot leave the Treasury unless Congress has appropriated it. The brief’s separation-of-powers argument is that using a litigation settlement to move funds from the Treasury, without a congressional appropriation, encroaches on this legislative power. The opposing position is that the Judgment Fund is itself a standing appropriation Congress enacted for exactly such payments. The reach of the appropriations power over settlement payments is a contested constitutional question.

“Fraud on the court” is a narrow and demanding legal standard, distinct from ordinary fraud between parties. It refers to misconduct that corrupts the judicial process itself — classic examples include bribery of a judge or fabrication of evidence by officers of the court — and it must typically be shown by clear and convincing evidence. Rozier v. Ford Motor Co. is a Fifth Circuit decision the brief cites for the proposition that egregious misconduct can justify relief from a judgment. (Fifth Circuit decisions issued before October 1, 1981 are binding precedent in the Eleventh Circuit, where this case sits, under Bonner v. City of Prichard.) The brief argues the parties’ conduct meets this high bar; that is the precise question the court would have to decide, and it is contested.

The underlying lawsuit arose from the unauthorized disclosure of tax returns. 26 U.S.C. § 7431 creates a civil damages action against the United States for unlawful disclosure of tax return information, and § 7431(c) caps recovery at actual or specified statutory damages plus, in some cases, costs and attorney’s fees. The Privacy Act, 5 U.S.C. § 552a, separately governs federal handling of personal records, with its own civil-remedy and damages provisions in § 552a(g). The factual backdrop (per court records and reporting) is the 2019–2020 leak of many taxpayers’ returns by an IRS contractor, Charles Littlejohn, who was criminally convicted. The brief’s point here is that these statutes cap relief at money damages to the plaintiffs — so, the amici argue, a billion-dollar fund for third parties and a grant of immunity are not remedies these statutes authorize. The brief also asserts the government declined to raise several available defenses (including a statute-of-limitations defense and arguments that the leaker was a contractor, not an employee); whether those defenses would have succeeded is not adjudicated.

“Consideration” is the bargained-for exchange that makes a promise enforceable as a contract — each side must give something of legal value. A long-standing wrinkle, reflected in the Restatement provision the brief cites, is that giving up a legal claim counts as consideration only if the claim is at least doubtful or honestly believed valid; surrendering a worthless claim gives up nothing. The brief’s argument proceeds in the alternative: first, that the released claims were meritless and so supplied no consideration; and second, that even if they had value, the benefits conferred (the Fund and the immunity addendum) were disproportionate to that value. The brief notes the Settlement Agreement itself stated the Fund’s capitalization “does not represent the value of any current claim by Plaintiffs.” Whether the released claims had value, and whether the exchange was adequate, are disputed.

28 U.S.C. § 2414 authorizes the Attorney General to settle claims against the United States and provides for their payment. The brief reads § 2414 — together with Government Accountability Office guidance and a line of Attorney General opinions dating to 1900 — to require a “bona fide dispute” over liability or amount before the government may compromise a claim, and to limit settlements to relief of the kind a court could award on the underlying claim. The brief pairs this with a contract-law principle (Restatement (Second) of Contracts § 74) that surrendering a claim is valid “consideration” only if the claim is doubtful or believed to be valid. The amici argue these limits were exceeded here; the opposing parties contend the settlement was a lawful exercise of § 2414 authority. The scope of § 2414 settlement authority is the central legal question the court must resolve.

The Judgment Fund (31 U.S.C. § 1304) is a permanent, indefinite appropriation Congress created so the government can pay final judgments and certain settlements against the United States without a case-by-case appropriation. Because it is “permanent and indefinite,” it does not require Congress to vote new money for each payment. Its use is statutorily limited — for settlements, payment generally must be authorized by the Attorney General or made under an agency’s own authority, and the underlying claim must be one the government is actually liable for. According to the Justice Department’s announcement and contemporaneous reporting, the $1.776 billion for the Anti-Weaponization Fund was to be drawn from the Judgment Fund. The brief’s argument is that routing money this way, for these purposes, exceeds those statutory limits and bypasses Congress’s appropriations role — a contention the opposing parties dispute.

Article III of the Constitution limits federal courts to deciding actual “Cases” and “Controversies.” A foundational requirement is genuine adversity — the parties must have truly opposing interests. Courts have long refused to decide “collusive” or “feigned” suits in which the nominal opponents actually share the same goal, because such cases lack a real dispute for the court to resolve. The brief notes that the district court itself raised this concern and ordered the parties to brief whether a case or controversy existed, given that the plaintiffs (the President and his family/company) had sued executive-branch agencies whose leadership the President oversees. According to news reporting, the presiding judge, Kathleen Williams, had questioned whether the parties were “sufficiently adverse.”

Rule 41(a)(1)(A)(i) of the Federal Rules of Civil Procedure lets a plaintiff dismiss its own case unilaterally — by filing a notice, without a court order and without the defendant’s consent — but only before the opposing party serves an answer or a summary-judgment motion. Because the dismissal takes effect on filing, it does not require the judge’s approval. The brief’s significance point is one of timing: it contends the plaintiffs used this self-executing mechanism to end the case two days before the court’s deadline for briefing whether a genuine “case or controversy” existed. “With prejudice” means the dismissed claims cannot be refiled.

Federal Rule of Civil Procedure 60(b) lets a court relieve a party from a final judgment or order in defined circumstances, including fraud and “any other reason that justifies relief.” Separately, courts have long recognized an inherent power to set aside a judgment for “fraud on the court,” which Rule 60(d)(3) expressly preserves and which is not subject to the one-year time limit that applies to some Rule 60(b) motions. The motion this brief supports asks the court to reopen the case after its dismissal. Because the plaintiffs voluntarily dismissed the case (rather than the court entering a contested judgment), the movants invoke the court’s authority to set aside the dismissal on fraud-on-the-court and related grounds.

An amicus curiae (“friend of the court”) is a non-party that files a brief to offer the court information or perspective bearing on a pending matter. Amici do not control the litigation and are not bound by its outcome; their briefs are permitted at the court’s discretion. Here, the 23 state attorneys general are not parties to Trump v. IRS — they filed this brief in support of a separate motion (by other movants) asking the court to reopen the dismissed case. The brief identifies the amici’s stated “interest” as their experience defending sovereign entities and settling claims on behalf of their states. The signatories are listed in the appendix; the brief is led by the California Attorney General’s office.

A dismissal “with prejudice” is a final adjudication on the merits: the United States cannot cure the defect by amending its complaint and refiling the same claim in this court. The court chose this disposition because its ruling is a legal interpretation of the statute — an SVRL simply is not a record the CRA reaches — so no additional facts could rescue the claim. This is distinct from a dismissal “without prejudice,” which would have allowed refiling. A with-prejudice dismissal is an appealable final judgment, which is the ordinary route by which the government would seek review in the Fourth Circuit.

This case is one of roughly 30 suits the Department of Justice filed across the country beginning in fall 2025, seeking states’ full statewide voter registration lists — including sensitive fields such as dates of birth, driver’s license numbers, and partial Social Security numbers. The filings were announced in batches by the Civil Rights Division through late 2025 and early 2026. At the time of this opinion, every court to rule had ruled against the United States; the eight dismissals the court cites (in Wisconsin, Maine, Arizona, Rhode Island, Massachusetts, Michigan, Oregon, and California) were decided between February and May 2026. According to public trackers, the DOJ appealed several of these dismissals. A number of other states separately agreed to provide their full lists voluntarily.

The Office of Legal Counsel (OLC) is the component of the Department of Justice that issues authoritative legal opinions binding within the executive branch. The United States cited a May 2026 OLC opinion concluding that the CRA authorizes the federal government to compel production of SVRLs. The court gives it no weight, observing that it “will not interpret the CRA contrary to its text simply because an office of the party advancing that interpretation has adopted it” — that is, an OLC opinion is the litigant’s own internal legal position, not independent or binding authority on a court. OLC opinions carry significant weight inside the executive branch but do not bind courts, which owe them no deference in litigation.

This is the court’s second, independent textual reason. Section 20702 makes it a federal crime to alter, deface, or destroy any record § 20701 requires to be retained. An SVRL, by design, is continuously updated — voters are added and removed, addresses change — and the NVRA and HAVA affirmatively require that ongoing maintenance. So if an SVRL were a § 20701 record, the routine list maintenance that federal law demands would simultaneously be a federal crime under § 20702. The court invokes the canon against absurdity — courts avoid readings of a statute that produce absurd or self-contradictory results — to reject that interpretation, reading § 20701 in harmony with the rest of the scheme.

This is the interpretive heart of the decision, quoted from the Michigan case United States v. Benson. The reasoning rests on a familiar tool of statutory construction — the rule against surplusage, which presumes every word in a statute does work and none is redundant. If § 20701 covered all records in an official’s possession, the qualifying phrase “come into his possession” would add nothing, since the state possesses its own SVRL. To give the phrase meaning, the court reads it to reach only records an official receives from an outside source (the voter), not records the state creates. The court reinforces this with dictionary definitions of “obtain,” “receive,” and “acquire,” and with a string of federal statutes that use “come into possession” to describe things a person gets rather than makes.

Kennedy v. Lynd (5th Cir. 1962) is the principal authority the United States relied on for its procedural argument. There, the Fifth Circuit described a CRA enforcement action as “a special statutory proceeding” conducted apart from the Federal Rules of Civil Procedure. The court gives two reasons for not following it: first, that the Supreme Court’s Powell decision two years later points the opposite way on materially identical statutory language; and second (in a footnote), that even Lynd itself acknowledged a court may decide whether a particular record falls within the CRA’s scope — which is exactly the dispositive question here. Lynd is a Fifth Circuit decision and is not binding on a district court in Maryland, which sits in the Fourth Circuit.

United States v. Powell (1964) is the Supreme Court decision the court finds controlling on the procedural question. Powell involved IRS administrative summonses under statutes (26 U.S.C. §§ 7402(b), 7604(a)) that, like the CRA’s § 20705, grant district courts jurisdiction to compel production “by appropriate process” but specify no particular procedure. The Court held that because those statutes contained no procedural specification, the Federal Rules of Civil Procedure apply, and enforcement proceeds by complaint, answer, and hearing. Judge Gallagher reasons that the CRA uses the identical “appropriate process” language and likewise specifies no procedure, so Powell dictates that the Federal Rules apply here too.

Federal Rule of Civil Procedure 81 lists specialized proceedings and the extent to which the ordinary civil rules apply to them. Subdivision (a)(5) provides that the Federal Rules govern proceedings to compel production of documents under a federal-agency subpoena “except as otherwise provided by statute, by local rule, or by court order.” The United States argued that a CRA enforcement action is a special summary proceeding to which the ordinary rules — including the right to file a motion to dismiss — do not apply. The court rejects that argument, finding nothing in the CRA or the local rules displacing the Federal Rules. The practical stakes were high: if the government were right, defendants could not have moved to dismiss, and the court would have proceeded directly toward compelling production.

This describes the “list maintenance” obligation of the National Voter Registration Act of 1993 (NVRA), 52 U.S.C. § 20507(a)(4) — the “Motor Voter” law. Together with the Help America Vote Act of 2002 (HAVA), 52 U.S.C. § 21083, which requires each state to keep a single, uniform, centralized, computerized statewide voter registration list, these statutes are what actually create and govern the SVRL. The court’s point is structural: the SVRL is a creature of the NVRA and HAVA, statutes enacted decades after the 1960 CRA, and the CRA’s retention provision was not written with such a database in mind. The United States notably chose not to sue under these statutes.

Section 20703 is the mechanism the United States invoked: it lets the Attorney General demand, in writing, that a record custodian make covered records available for inspection and copying. The final sentence — requiring the demand to state its “basis and purpose” — has been a second front in this nationwide litigation. In the parallel Oregon case, the court dismissed partly because it found the DOJ’s demand letter failed to adequately state a basis and purpose tied to the statute’s anti-discrimination function. Judge Gallagher does not reach that question here, resolving the case instead on the threshold issue of whether an SVRL is a covered record at all.

This is the operative language of 52 U.S.C. § 20701, and the phrase “come into his possession” is the hinge of the entire decision. The court reads it to cover only records an election official receives from an outside source — chiefly, materials voters themselves submit (registration forms, applications) — not records the state itself generates. The reference to “payment of poll tax” reflects the statute’s 1960 vintage: poll taxes were still in use at the time and were not abolished in federal elections until the Twenty-Fourth Amendment (1964) and in state elections until Harper v. Virginia Board of Elections (1966).

Title III of the Civil Rights Act of 1960 (now codified at 52 U.S.C. §§ 20701–20706) was enacted to combat the systematic destruction and concealment of voter records by Southern officials, which had frustrated federal efforts to document racial discrimination in voter registration. It requires election officers to retain voting records for 22 months and authorizes the Attorney General to demand access to those records. Notably, this is the only claim the United States brought in this case — it did not sue under the NVRA or HAVA, the two modern statutes that more directly govern voter-list maintenance. That choice matters because, as the court explains, the CRA’s retention provision contains limiting language (“come into [their] possession”) that the NVRA’s broader disclosure provision does not.

A writ of mandamus is an extraordinary remedy by which a higher court orders a lower court (or a government official) to perform a duty correctly or to undo a clear abuse of discretion. In the litigation context here, it is the mechanism an appellate court uses to quash an improper discovery or testimony order before it takes effect, rather than waiting for a final judgment to appeal. Mandamus is granted sparingly; the petitioner must generally show a clear and indisputable right to relief and no other adequate means to obtain it (Cheney v. U.S. District Court, 2004). The government cites these mandamus rulings to argue that orders compelling testimony from senior officials are the kind of error appellate courts step in to correct — implying the district court's declaration demand would be vulnerable to the same challenge.

The string of cases cited here (In re Cheney, In re Musk, In re Paxton, In re U.S. Dep't of Education, and In re United States (Jackson)) all apply what is known as the "apex doctrine." Under this doctrine, courts generally protect high-ranking government and corporate officials from being compelled to testify or sit for depositions unless the party seeking the testimony shows the official has unique, first-hand knowledge that cannot be obtained from other sources. The Supreme Court's decision in Cheney v. U.S. District Court (2004) established that discovery directed at senior Executive Branch officials raises separation-of-powers concerns and should be narrowly limited. Each of the cited "In re" cases involved an appellate court using a writ of mandamus to block a lower court's order compelling such testimony. The government invokes this line of authority to argue the court cannot condition mootness on sworn statements from three senior officials.

Section 1001 is the federal false-statements statute, which criminalizes knowingly making materially false statements to the federal government, with a maximum penalty of five years imprisonment. The government cites it here to argue that its representations — the Acting Attorney General's congressional testimony and counsel's court filings — were already made under threat of criminal penalty, so separate sworn declarations would add nothing to the mootness analysis. One distinction the filing does not draw: a § 1001 violation for false congressional testimony would, as a practical matter, be prosecuted by the Department of Justice itself, and the other authority cited (Federal Rule of Civil Procedure 11) governs attorney conduct and carries primarily monetary sanctions rather than criminal liability. A sworn declaration under penalty of perjury (28 U.S.C. § 1746) is the specific form of commitment the court requested.

A case is "moot" when there is no longer a live controversy for a court to decide — for example, when the challenged conduct has stopped and cannot reasonably be expected to recur. Federal courts lack jurisdiction over moot cases under Article III. The government argues the Fund's cancellation moots this lawsuit. The complication is a doctrine called "voluntary cessation": when a defendant voluntarily stops the challenged conduct, the case is not automatically moot, because the defendant could resume the conduct after the suit is dismissed. The party claiming mootness bears a "heavy burden" of showing the conduct cannot reasonably be expected to recur (Friends of the Earth v. Laidlaw, 2000). This is why the court sought sworn declarations rather than accepting the government's litigation statements — and why the President's post-testimony comments that he still wanted the Fund were significant to the court's analysis.

The Anti-Weaponization Fund was a $1.776 billion fund the Justice Department announced on May 18, 2026, to compensate individuals who claimed they were targeted or harmed by federal "weaponization" or "lawfare." It was created as part of the settlement of President Trump's lawsuit against the IRS over the leak of his tax returns; under the settlement, the Trump plaintiffs dropped their claims in exchange for the Fund's creation. The money was to come from the Judgment Fund, a permanent Treasury appropriation (31 U.S.C. § 1304) used to pay settlements and judgments against the federal government. The Fund drew bipartisan criticism — including concern that it could pay January 6 defendants — and several lawsuits, of which this case (Floyd v. DOJ) is one. Critics, including some members of Congress, argued the spending was not authorized by Congress, raising separation-of-powers and appropriations questions.

This is the Court's own characterization of what it has done: it declines to add the petitioner's finality requirement but insists it is not enlarging the doctrine either, leaving Rooker-Feldman confined to the Exxon formulation — suits by state-court losers, complaining of injuries from a state-court judgment rendered before the federal suit, that ask the federal court to review and reject that judgment. The practical holding is that the doctrine applies regardless of whether the state judgment is still subject to further appeal. The dissent reads the result differently, arguing the Court has in fact expanded the doctrine past Exxon's line — while noting, with some relief, that the majority repeatedly calls the doctrine "narrow," language lower courts should heed.

Supreme Court Rule 14.1(a) provides that only the questions "set out in the petition, or fairly included therein," will be considered by the Court. The majority invokes it to explain a deliberate limit on the decision: although it reads much of the petitioner's (and the dissent's) argument as ultimately resting on the view that Rooker and Feldman were wrongly decided and should be cut back or overruled, that broader question was not presented in the petition for certiorari, so the Court does not address it. The decision therefore takes the doctrine's validity as a given and resolves only the narrower finality question.

Cox Broadcasting Corp. v. Cohn (1975) is the leading case on what counts as a "final judgment" for Supreme Court review under § 1257. It recognizes that finality is not purely mechanical: the Court described a "pragmatic" approach identifying four sets of circumstances in which a state-court judgment may be treated as final for review even though further state proceedings technically remain. The majority cites it to make a practical point against the petitioner's rule — determining § 1257 finality is itself complicated, so building Rooker-Feldman around that requirement would add confusion rather than reduce it.

This citation marks the circuit split the Court took the case to resolve. After Exxon, most courts of appeals read its "after the state proceedings ended" language to mean Rooker-Feldman applies only once the state litigation has functionally "ended" — and the First Circuit in Federación de Maestros built a three-part test to decide when that has happened. By the dissent's count, seven circuits adopted some version of an "ended" requirement; only the Sixth Circuit (and now the Fourth, in this case) held that an unfinished, still-appealable state judgment can trigger the doctrine. Notably, the majority observes that these "ended"-test circuits did not adopt the stricter § 1257-finality rule the petitioner urged — the "ended" approach and the petitioner's rule are not the same thing.

"Comity" refers to the mutual respect courts of different systems extend to one another's proceedings — here, the deference federal courts owe to ongoing state-court litigation. The Court invokes Ruhrgas AG v. Marathon Oil Co. (1999) for the principle that cooperation and comity underpin the federal system, and ties that to § 1257: Congress let the Supreme Court intervene in state litigation only after a State's highest court has rendered a final judgment, which the Court reads as a deliberate limit on federal intrusion into state affairs. The dissent sharply disputes that federalism does any work here, noting that neither Rooker nor Feldman used the word "federalism."

These are two of the federal abstention doctrines — judge-made rules under which a federal court with jurisdiction nonetheless declines to exercise it out of deference to state proceedings. Younger v. Harris (1971) bars federal courts from enjoining ongoing state criminal prosecutions and certain civil enforcement proceedings. Colorado River (1976) permits a federal court to stay or dismiss a case in favor of parallel state litigation, but only in "exceptional circumstances," given courts' "virtually unflagging obligation" to hear cases within their jurisdiction. The petitioner and the dissent argue these doctrines (plus preclusion) already do Rooker-Feldman's work; the majority responds that they may not even apply where a plaintiff attacks a judgment rather than relitigating claims — and the district court here found none of them applicable.

"Void ab initio" is a Latin phrase meaning "void from the beginning" — a request to treat the consent order as if it had never had any legal effect, rather than merely setting it aside going forward. The Court treats the nature of this requested relief as decisive: asking a federal district court to declare a state-court judgment void from inception, and to enjoin its enforcement, is in substance a request for appellate review of that judgment. Because that is what triggers Rooker-Feldman, the specific words of T. M.'s prayer for relief — not the § 1983 label on her complaint — are what place her suit inside the doctrine.

Lance v. Dennis (2006) is a per curiam decision — an unsigned opinion of the Court — issued the year after Exxon. It reinforced Exxon's narrowing message, stressing that Rooker-Feldman "applies only in limited circumstances" where a party "in effect seeks to take an appeal of an unfavorable state-court decision to a lower federal court," and reiterating that the doctrine does not bar suits by those who were not parties to the state-court judgment. Both the majority and the dissent cite Lance — the majority for its "in effect . . . an appeal" framing, the dissent for the proposition that post-Exxon cases describe the doctrine without relying on § 1331.

Johnson v. De Grandy (1994) is cited for two propositions that recur in Rooker-Feldman law. First, the doctrine bars federal suits that seek "what in substance would be appellate review of [a] state judgment" — the inquiry looks to the substance of the relief sought, not the statute invoked. Second, De Grandy is the source of the rule (developed in later cases) that Rooker-Feldman does not bar a federal plaintiff who was not a party to the state-court judgment, because such a person could never have sought review of it in the Supreme Court. The majority leans on the first point; the "substance over form" framing is what lets it treat this suit as an attempted appeal.

Exxon Mobil Corp. v. Saudi Basic Industries Corp. (2005) is the pivotal modern case, and the parties read it in opposite ways. By 2005 the lower courts had expanded Rooker-Feldman well beyond its origins, using it to dismiss many suits that merely overlapped with state litigation. Exxon reined this in, holding the doctrine is "confined to cases of the kind from which the doctrine acquired its name" — suits by state-court losers, complaining of injuries from a state-court judgment, filed after that judgment, and asking the federal court to review and reject it. The dispute in this case is whether Exxon's reference to suits filed "after the state proceedings ended" is itself a requirement (the dissent's view) or merely a description of the facts in Rooker and Feldman (the majority's view).

Feldman (1983) is the doctrine's second namesake. Applicants who had been denied admission to the D.C. bar (in part because they had not graduated from an accredited law school) sued in federal district court. The Supreme Court drew a line that still governs: a district court has no jurisdiction to review the D.C. high court's specific decision denying these applicants a waiver — that is appellate review reserved to the Supreme Court — but it does have jurisdiction to hear a general constitutional challenge to the bar-admission rule itself, because reviewing a rule is not reviewing a judicial decision. That distinction (challenging a judgment versus challenging a rule) remains central to applying Rooker-Feldman.

Rooker v. Fidelity Trust Co. (1923) is the first of the doctrine's two namesake cases. After William and Dora Rooker lost in the Indiana courts — a judgment affirmed by the Indiana Supreme Court — they filed a new suit in federal district court asking it to declare the state judgment unconstitutional and void. In a brief unanimous opinion, the Supreme Court held the district court had no jurisdiction: setting aside a state-court judgment for legal error is an exercise of appellate jurisdiction, and the district courts' jurisdiction is "strictly original." The proper path, the Court said, was appeal through the state system and ultimately to the U.S. Supreme Court.

Section 1331 is the general federal-question statute: it gives district courts "original jurisdiction" over civil cases arising under federal law. "Original" jurisdiction means the power to hear a case in the first instance, as opposed to "appellate" jurisdiction, the power to review another court's decision. The two are treated as mutually exclusive. The Court's reasoning is that because § 1331 confers only original jurisdiction, a district court asked to undo a state-court judgment would be exercising appellate jurisdiction it does not possess. The dissent counters that a collateral attack on a judgment is not the functional equivalent of an appeal, so § 1331's original-jurisdiction grant should be enough.

Section 1257(a) of Title 28 is the statute that gives the U.S. Supreme Court jurisdiction to review state-court decisions, and it does so only for "final judgments or decrees rendered by the highest court of a State in which a decision could be had." Two features of that text drive this case: review is limited to final judgments, and only from a State's highest available court. The petitioner's proposed rule would have imported both limits into Rooker-Feldman — barring federal district court suits only where § 1257 would actually let the Supreme Court step in. The majority rejects that link; the dissent would embrace it, treating Rooker-Feldman as essentially a mirror of § 1257 (what it calls "the § 1257 Rule").

The Rooker-Feldman doctrine holds that the lower federal courts (district courts and courts of appeals) have no power to sit in review of state-court judgments — a losing party in state court cannot file a new suit in federal district court asking it to declare the state judgment wrong or void. The only federal court that may review a state-court judgment is the U.S. Supreme Court, and only by writ of certiorari. The doctrine takes its name from the two cases in which the Supreme Court applied it: Rooker v. Fidelity Trust Co. (1923) and District of Columbia Court of Appeals v. Feldman (1983). The precise formulation the Court quotes here comes from Exxon Mobil Corp. v. Saudi Basic Industries Corp. (2005), the modern decision that sharply narrowed the doctrine.

"Draw" is a term of art in vicarious copyright liability, the Fourth Cause of Action. Vicarious infringement has two elements: (1) the defendant had the right and ability to supervise or control the infringing activity, and (2) the defendant derived a direct financial benefit from it. The "draw" concept addresses the second element: under cases like Fonovisa, Inc. v. Cherry Auction (9th Cir. 1996), a financial benefit is "direct" when the availability of the infringing material acts as a draw that attracts customers, even if it is not the primary draw. The complaint alleges the availability of Eminem's music drew users to Meta's platforms and thereby increased advertising and subscription revenue — the factual basis for the direct-financial-benefit element.

This is the standard for contributory copyright infringement, the Third Cause of Action. Contributory liability has two elements: (1) knowledge of the direct infringement, and (2) a material contribution to it (inducing, causing, or materially contributing to the infringing conduct). The doctrine traces to Gershwin Publishing Corp. v. Columbia Artists Management (2d Cir. 1971) and was applied by the Supreme Court in Sony Corp. v. Universal City Studios (1984), the "Betamax" case, which held that a product capable of substantial non-infringing uses does not by itself create contributory liability. Contributory infringement differs from inducement (Count II) in that it does not require proof of intent to promote infringement, and from vicarious liability (Count IV) in that it turns on knowledge and contribution rather than financial benefit and control.

The Second Cause of Action is for inducement of copyright infringement. The governing standard comes from the Supreme Court's decision in MGM Studios, Inc. v. Grokster, Ltd. (2005), which held that one who distributes a product or service "with the object of promoting its use to infringe copyright, as shown by clear expression or other affirmative steps taken to foster infringement," is liable for the resulting infringement by third parties. Inducement requires more than merely knowing a product can be used to infringe; it requires affirmative acts encouraging infringement. The complaint's allegations that Meta promoted music features, declined to implement filtering, and advertised the tools are the factual hooks for the Grokster elements — though whether they meet the "affirmative steps to foster infringement" bar is a question for the court.

These are two specific Meta platform features the complaint singles out. "Original Audio" lets a user extract the audio track from an existing post or Reel (or upload their own) and label it as their own original sound, which other users can then reuse. "Reels Remix" lets a user take a portion of another user's Reel — keeping its audio — and build new content around it. The complaint's theory is that these features cause "exponential" infringement: one unlicensed composition, once stripped and relabeled as "original," propagates across many additional videos by many additional users, each constituting a further alleged reproduction and synchronization. This mechanism underlies the inducement and contributory-infringement counts.

Cambridge Analytica was a British political consulting firm that, in a scandal that became public in March 2018, obtained the Facebook profile data of tens of millions of users — harvested via a third-party personality-quiz app — without their consent, and used it for political advertising and voter-targeting work. The fallout led to a $5 billion FTC penalty against Facebook in 2019, a $725 million class-action settlement, and Cambridge Analytica's own bankruptcy. The complaint cites the Michigan-specific data-sharing figure to support its argument that Meta has extensive contacts with Michigan for purposes of personal jurisdiction; the reference is offered as evidence of Meta's reach into the state, not as a copyright allegation.

In 2013, the same plaintiffs sued Facebook in this District over the song "Under the Influence," which they alleged was used without authorization in a promotional video ("Airplane") for the "Facebook Home" software launch. The ad agency Wieden+Kennedy had reportedly taken the position that no copyright could be asserted because the underlying music resembled an earlier Michael Jackson recording. Facebook pulled the video and the case was ultimately resolved by a stipulated dismissal. The complaint invokes this prior case for a specific purpose: to argue that Meta already submitted to personal jurisdiction in the Eastern District of Michigan once before, and to show Meta was on notice of the plaintiffs' selectivity about licensing Eminem's music.

Under 17 U.S.C. § 504(c), a copyright owner may elect statutory damages instead of proving actual damages. The range is $750 to $30,000 per work infringed, rising to a maximum of $150,000 per work where the infringement is willful — the figure the complaint uses. The plaintiffs' arithmetic ($150,000 × 243 works × 3 platforms) yields roughly $109.3 million, the headline exposure figure widely reported in connection with the case. Two caveats a reader should know: statutory damages are awarded per work infringed, and whether the same composition made available on three platforms supports three separate awards (rather than one) is a contested legal question, not a settled multiplier; and courts have discretion to set any amount within the statutory range, so the maximum is a ceiling, not an entitlement.

Audiam is a music-rights administration company founded in 2013 by Jeff Price (a co-founder of TuneCore) that locates and collects digital mechanical royalties, with an early emphasis on YouTube. Its ownership has changed hands several times: SOCAN (a Canadian collecting society) acquired it in 2016, and SESAC acquired a controlling stake in August 2021. The Harry Fox Agency (HFA), the largest U.S. mechanical-rights licensing agent, is also owned by SESAC, which acquired HFA in 2015 — so Audiam and HFA are corporate siblings under common SESAC ownership rather than HFA being Audiam's direct parent. SESAC has been majority-owned by the investment firm Blackstone since 2017. The complaint's relevance for Audiam is narrower: it alleges Meta licensed certain catalogs through Audiam but that Audiam never represented, and had no authority to license, the Eight Mile Style catalog.

The DMCA's safe harbor (17 U.S.C. § 512) shields online service providers from monetary liability for copyright infringement committed by their users, provided the provider meets certain conditions — most importantly that the infringing material is stored "at the direction of a user," that the provider lacks actual knowledge of specific infringement, and that it removes infringing material upon proper notice (the "notice-and-takedown" system). The complaint's recurring argument is that the safe harbor does not apply here because Meta itself — not its users — allegedly copied and stored the compositions in its Music Libraries, making Meta a direct infringer rather than a passive host. Whether storage was "at the direction of a user" is typically a central fought-over question in these cases.

The "synchronization" or "sync" right is the right to reproduce a musical composition in timed relation with visual images — for example, pairing a song with a video. It is part of the bundle of exclusive reproduction rights a composition's copyright owner holds under 17 U.S.C. § 106. Sync licenses are negotiated individually between the user and the rights-holder (there is no compulsory or statutory sync license, unlike the compulsory mechanical license that exists for making audio-only recordings). This is central to the complaint's theory: when a Meta user pairs an Eight Mile composition with a video in a Reel or Story, that is a synchronization, and the complaint alleges no sync license was ever granted.

The full list appears in Schedule A (Exhibit A) at the end of the filing. Each entry shows a song title and an "EMS / MA Controlled Share" percentage — the fraction of the composition's copyright that Eight Mile Style or Martin Affiliated controls, which ranges from a fraction of a percent to 100%. This matters because music compositions are frequently co-owned by multiple songwriters and publishers; a 100% share means the plaintiffs control the entire composition, while smaller shares reflect partial ownership alongside other rights-holders. The catalog includes many of Eminem's best-known works (for example, "Lose Yourself," "The Real Slim Shady," "Cleaning Out My Closet," and "Mockingbird" are all listed at 100%).

This is a copyright infringement suit filed May 30, 2025, in the U.S. District Court for the Eastern District of Michigan. The plaintiffs, Eight Mile Style and Martin Affiliated, are the music publishing entities that co-own and administer a catalog of compositions recorded by Marshall Mathers (Eminem). They allege that Meta — through Facebook, Instagram, and WhatsApp — reproduced, stored, and distributed 243 Eminem compositions in its in-app "Music Libraries" without a license, and built tools (Original Audio, Reels Remix) that let users incorporate the music into their own videos. The complaint pleads four counts: direct copyright infringement, inducement, contributory infringement, and vicarious infringement. The same plaintiffs previously sued Facebook in this District in 2013 over an Eminem composition used in a "Facebook Home" advertisement, and separately litigated against Spotify over streaming of the same catalog.

The decision was 8–1. Justice Kagan wrote for the Court, joined by Chief Justice Roberts and Justices Alito, Sotomayor, Gorsuch, Kavanaugh, Barrett, and Jackson. There were three concurrences and one dissent. Justice Gorsuch (joined by Sotomayor and Jackson) concurred to argue the decision should go further and to question whether prospective appeal waivers are valid at all. Justice Kavanaugh (joined by Alito and Barrett) concurred to emphasize that the miscarriage-of-justice exception sets a high bar, expressly disagreeing with what he read as Gorsuch’s lower threshold. Justice Barrett concurred separately to address the Court’s “supervisory power” over lower courts, which she — like the dissent — doubts exists, locating the decision instead in the common law of waiver. Justice Thomas dissented alone, arguing the Court identified no source of law for its new exception and that Hunter’s knowing, voluntary waiver should simply be enforced.

This phrase, quoted from Cutter v. Wilkinson (2005), expresses a routine principle of appellate practice: the Supreme Court generally decides only questions that lower courts have already passed on, rather than resolving issues “in the first instance.” Here it explains why the Court announced the miscarriage-of-justice standard but declined to apply it to Hunter’s medication challenge — the Fifth Circuit had never reached that question because its own precedent foreclosed it. The practical effect is that Hunter has won the legal rule but not yet his case: the dispute returns to the Fifth Circuit to decide, under the new standard, whether enforcing his waiver would be a miscarriage of justice.

Supervised release is a period of court-monitored conditional liberty that a federal defendant serves after completing a prison term — it replaced traditional parole in the federal system under the Sentencing Reform Act of 1984. During supervised release, the defendant must comply with conditions set by the court (here including the disputed medication requirement); violating those conditions can result in revocation and a return to prison after a separate hearing. As the dissent notes, a defendant may move to modify supervised-release conditions “at any time” under 18 U. S. C. §3583(e)(2), and a court must hold separate revocation proceedings under §3583(e)(3) before imprisoning someone for a violation — points the dissent raises to argue Hunter’s challenge is premature.

Aiding and abetting is not a separate crime but a theory of liability, codified at 18 U. S. C. §2, under which a person who helps commit a federal offense is punishable as a principal — that is, as if he had committed the offense himself. Wire fraud itself (18 U. S. C. §1343) criminalizes the use of interstate wire communications to carry out a scheme to defraud, and carries a statutory maximum of 20 years (30 years where a financial institution is affected, as charged here). Because aiding and abetting carries the same penalty as the underlying offense, Hunter’s guilty plea to one count of aiding and abetting wire fraud exposed him to the same statutory maximum as a principal wire-fraud count.

An appeal waiver is a provision in a plea agreement by which the defendant gives up, in advance, the statutory right to appeal his conviction, his sentence, or both. The right to appeal a federal sentence is not constitutional — it was created by statute (18 U. S. C. §3742) and is of relatively modern origin. Appeal waivers are now standard in federal plea agreements: studies cited in the opinions indicate that as of 2003, roughly two-thirds of federal plea agreements contained them, and they have become more common since. A defendant typically signs the waiver before the sentence is known, which is the feature the concurrences and dissent debate most pointedly — the defendant is waiving the right to challenge a sentence that has not yet been imposed.

These are the statutory provisions the parties dispute over the medication condition. Section 3583(d) governs the discretionary conditions a court may impose on supervised release; subsection (d)(2) requires that any such condition involve “no greater deprivation of liberty than is reasonably necessary” to serve the statutory sentencing purposes. Hunter cites this provision to argue the forced-medication condition is excessive. The Government counters with 18 U. S. C. §3563(b)(9), which expressly lists, as an available condition, that a defendant “undergo available medical, psychiatric, or psychological treatment.” The interplay is central to the remand: §3563(b)(9) authorizes treatment conditions in general terms, while §3583(d)(2) and the constitutional liberty interest in refusing medication (see Sell v. United States, 539 U. S. 166 (2003)) supply the limits.

Federal Rule of Criminal Procedure 11 governs guilty pleas in federal court. Rule 11(c)(1) confirms that plea agreements are negotiated between the prosecutor and the defendant, with the court forbidden from participating in those discussions. Rule 11(c)(3)(A), which the opinion also cites, gives the district court three options when presented with a plea agreement: accept it, reject it, or defer a decision. Separately, Rule 11(b)(1)(N) — discussed elsewhere in the opinion and the concurrences — requires the judge to confirm that the defendant understands the terms of any appeal waiver before accepting the plea. This rule, adopted in 1996, figures in the dissent’s argument: Justice Thomas notes that the rulemaking committee has repeatedly considered and declined to adopt limits on appeal waivers beyond this disclosure requirement.

United States v. Olano (1993) is the case that fixed the now-standard distinction between “waiver” and “forfeiture” in federal criminal procedure. Waiver is the “intentional relinquishment or abandonment of a known right”; forfeiture is “the failure to make the timely assertion of a right.” The distinction matters because a forfeited error can still be reviewed on appeal for “plain error,” while a waived right generally cannot be raised at all. The majority uses this distinction to explain why the Government’s silence at sentencing was neither: it was not an affirmative abandonment (waiver), and because the proper time to enforce an appeal waiver is after a notice of appeal is filed — not at sentencing — the silence was also not a failure to timely assert the right (forfeiture).

This memorable hypothetical comes from United States v. Mezzanatto (1995), where the Supreme Court held that a defendant may waive the protections of the plea-statement evidentiary rules. In the passage the Court quotes, it borrowed an illustration from an opinion by Judge Richard Posner (United States v. Josefik, 7th Cir. 1985): even if a defendant consented to be tried by twelve orangutans, the conviction would be invalid, because “some minimum of civilized procedure is required by community feeling regardless of what the defendant wants.” The orangutan recurs throughout the Hunter opinion — and at oral argument, counsel debated whether a sentence chosen by an orangutan could be insulated from appeal by a waiver. It serves as the Court’s shorthand for the principle that some procedural defects are so fundamental that no agreement by the parties can require a court to honor them.

Santobello v. New York (1971) is the foundational modern case on the enforceability of plea agreements. There, the prosecution promised as part of a plea deal to make no sentencing recommendation, but a different prosecutor later recommended the maximum sentence, which the court imposed. The Supreme Court held that when the government breaks a promise that induced a guilty plea, the defendant is entitled to a remedy — either withdrawal of the plea or specific performance of the agreement. Santobello is also the case in which the Court declared plea bargaining “an essential component of the administration of justice” and “highly desirable” — language Justice Gorsuch’s concurrence revisits critically as the turning point when the Court embraced plea bargaining. The majority cites it here for the narrower proposition that a district court’s decision whether to accept a plea agreement is committed to “sound judicial discretion.”

In Class v. United States (2018), the Supreme Court held 6–3 (Breyer, J.) that a defendant who pleads guilty does not automatically give up the right to argue on appeal that the statute he was convicted under is unconstitutional. As relevant here, the plea agreement contained no appeal waiver, but during the plea colloquy the district judge mistakenly told the defendant he was “giving up [his] right to appeal.” The Court held that the defendant’s agreement to that misstatement “neither expressly nor implicitly waived his right to appeal.” The majority treats this case as the mirror image of Hunter’s: in Class a court’s misstatement did not create an appeal waiver, and here a court’s misstatement (plus the prosecutor’s silence) did not undo one. The operative principle in both is that a judge’s offhand remark during a hearing cannot, by itself, alter the parties’ agreement about appeal rights.

The Court is explicit about the limits of its holding, expressly reserving several questions: (1) disarming those "addicted to" a controlled substance (the separate clause of § 922(g)(3), not at issue because the government never alleged Hemani was an addict); (2) those presently intoxicated; (3) future laws targeting users of a particular drug Congress finds especially risky; (4) § 922(g)(1)'s felon ban; and (5) a § 922(g)(3) prosecution supported by individualized proof of dangerousness, or proof that a specific drug always renders users dangerous. The decision is as-applied to Hemani on these facts — regular marijuana use, nothing more.

The Court quotes the dissent of then-Judge Amy Coney Barrett in Kanter v. Barr (7th Cir. 2019), a case decided before her elevation to the Supreme Court. The Kanter majority upheld the felon-in-possession ban (§ 922(g)(1)) as applied to a man convicted of a nonviolent fraud offense; Barrett dissented, arguing through extensive historical analysis that legislatures may disarm people who are dangerous, but not categories of people based on status alone absent dangerousness. That "dangerousness, not status" principle — and her warning that a broad power to designate groups as dangerous could "quickly swallow" the right — is what the majority borrows here.

This citation is to a DOJ/DEA final order published April 28, 2026 (91 Fed. Reg. 22714), issued after oral argument in this case. The order moved a narrow set of marijuana products — those in FDA-approved drug products, and marijuana handled under a state medical-marijuana license — from Schedule I to Schedule III. It did not reschedule marijuana generally; recreational and other marijuana remained on Schedule I, and a separate DEA administrative hearing on broader rescheduling was set to begin June 29, 2026. The Court uses the order as evidence that the federal government's own regulatory posture toward marijuana is inconsistent with treating every marijuana user as categorically dangerous.

This describes § 922(g)(8), the statute upheld in Rahimi. The Court draws a deliberate contrast between two kinds of historical surety law. Surety-of-the-peace laws — the ones Rahimi relied on — required an individualized showing that a person posed a threat of violence before a bond could be demanded. The government here instead invokes surety-of-good-behavior laws, which targeted "scandals against good morals" (drunkenness, eavesdropping, frequenting bawdy houses) and did not require any showing of violence. The opinion notes the government does not rely on the surety-of-the-peace tradition — and explains why that tradition would not fit a defendant whose only conduct is regular drug use.

This passage supports the Court's historical claim with figures on early-American alcohol consumption. By most accounts per-capita consumption of distilled spirits in the early nineteenth century was far higher than today; the era is sometimes called "the Alcoholic Republic" after W.J. Rorabaugh's 1979 study of the same name, which the opinion cites. The Court's point is comparative: if merely regular heavy drinking had triggered disarmament-style restrictions, a large share of the founding generation — by its examples, including several Presidents — would have qualified, which it did not.

Benjamin Rush (1746–1813) was a Philadelphia physician, a signer of the Declaration of Independence, and an early figure in American medicine. The treatise the Court cites, his Medical Inquiries and Observations Upon the Diseases of the Mind (1812), is often described as the first American textbook of psychiatry. Rush was also an early proponent of the view that habitual drunkenness was a disease rather than merely a moral failing — context that reinforces the Court's point that the historical category described a condition of incapacity.

In re Tracy (N.Y. Ch. 1829) is an early American case the Court uses to fix the historical meaning of "habitual drunkard." The quoted standard — intoxicated for a considerable part of one's time, to a degree that deprives one of ordinary reasoning faculties — illustrates the opinion's central historical finding: founding-era law reserved the label for those rendered functionally incapacitated, not for ordinary or even frequent drinkers. "Paige Ch." refers to Paige's Chancery Reports, which compiled decisions of the New York Court of Chancery, an equity court that existed before New York merged law and equity in 1847.

"Why" and "how" are the two metrics Bruen identified as playing a central role in deciding whether a modern law is "relevantly similar" to a historical one: the why is the purpose or justification behind the law (what problem it addresses), and the how is its operation or mechanism (the manner and degree of burden it imposes on the right). The structure of the majority opinion tracks these two axes — Parts A, B, and D test the government's "why" comparisons, and Part C tests the "how."

United States v. Rahimi (2024) was the Court's first major application of Bruen. It upheld § 922(g)(8), which disarms persons subject to domestic-violence restraining orders, by analogy to historical surety and "going armed" laws. Rahimi clarified that the government need not produce a "historical twin" or a precise founding-era precursor; it is enough that a modern law is "relevantly similar" to historical regulation in why it burdens the right and how it does so. The Court here applies that more forgiving standard and still finds the government's analogy wanting.

This is the test from New York State Rifle & Pistol Assn. v. Bruen (2022). Step one: if the Second Amendment's plain text covers the conduct, the Constitution presumptively protects it. Step two: the burden shifts to the government to show the regulation is consistent with the Nation's historical tradition of firearm regulation. Bruen displaced the "two-step" means-end balancing (involving tiers of scrutiny) that most circuit courts had used after Heller — the very framework Justice Jackson's concurrence here urges the Court to reconsider. Throughout this case the government concedes step one, so the entire dispute is litigated at step two.

For scale: § 922(g) as a whole produces thousands of federal convictions per year, the large majority under § 922(g)(1) (felon-in-possession). The unlawful-user provision at issue here, § 922(g)(3), accounts for only about 5% of that total. The Court attributes the figure to a Congressional Research Service report, Guns and Drugs: A Brief History of 18 U.S.C. § 922(g)(3) (2026).

The CSA sorts controlled substances into five schedules. Schedule I is the most restrictive — substances with a high potential for abuse and (in the statute's terms) no currently accepted medical use, such as heroin and LSD. The schedules descend to Schedule V, the least restrictive. Placement is made by the DEA in consultation with HHS and turns on statutory factors in 21 U.S.C. § 811(c), including current scientific knowledge, history of abuse, and risk to public health — factors that need not relate to whether a drug makes its users violent. That is the structural point the Court develops later: the gun ban's reach is defined by a list assembled for public-health reasons, not dangerousness.

This is the penalty provision the opinion repeatedly references. 18 U.S.C. § 924(a)(8) sets the maximum prison term for a § 922(g) violation at 15 years. That ceiling was raised from 10 years to 15 years by the Bipartisan Safer Communities Act of 2022. The "disarmament for life" the Court mentions follows because a § 922(g)(3) conviction is itself a felony, which then triggers § 922(g)(1)'s separate, permanent ban on firearm possession by convicted felons.

Section 922(g)(3) is one of nine prohibited-possessor categories in the federal Gun Control Act of 1968. It bars firearm possession by anyone who is an "unlawful user of" or "addicted to" a controlled substance. Critically, the provision does not define "controlled substance" itself — it borrows the definition from the Controlled Substances Act (CSA), so the scope of the gun ban automatically tracks whatever Congress and the DEA place on the CSA's five schedules. This incorporation is central to the Court's reasoning: because a substance can be scheduled for reasons unrelated to violence, the firearm ban sweeps in users of any scheduled drug regardless of dangerousness.

District of Columbia v. Heller (2008) was the Supreme Court's landmark decision recognizing that the Second Amendment protects an individual right to keep and bear arms for self-defense, unconnected to militia service. The same opinion expressly cautioned that the right is "not unlimited" and listed examples of "presumptively lawful" regulations — including bans on possession by felons and the mentally ill, and prohibitions on carrying in sensitive places. The Court here draws on that limits language; the page it cites (554 U.S. at 626) is the well-known passage cataloguing those presumptively lawful measures.

This footnote invokes one of the foundational principles of American antitrust law: the statutes protect competition as a process, not individual competitors from the effects of competition. Brunswick (1977) involved bowling alley operators who argued they were harmed when Brunswick acquired failing competing alleys and kept them open, preventing the plaintiffs from gaining market share. The Supreme Court held that this was not the kind of harm the antitrust laws were designed to prevent — the plaintiffs were harmed by more competition, not less. The underlying principle comes from Brown Shoe Co. v. United States (1962), which is quoted in the footnote. The DOJ deploys this distinction to address complainants who argued the merger would harm them — implying that their objections reflect competitive self-interest rather than genuine harm to competition or consumers.

General Dynamics (1974) is a landmark Supreme Court antitrust case that established that historical market share data, standing alone, is not sufficient to sustain a presumption that a merger will harm competition. The case involved the merger of two coal producers. Although the combined company held a significant market share, the Court found that one company's coal reserves were largely committed under long-term contracts, meaning its future competitive significance was far less than its current market share suggested. The DOJ cites General Dynamics here to signal that even if the combined Paramount/WBD entity would hold a significant share of theatrical film production, the "highly dynamic" nature of the industry — with new entrants like A24 and shifting distribution models — means historical shares are an unreliable predictor of future competitive harm.

In March 2019, The Walt Disney Company completed its $71.3 billion acquisition of 21st Century Fox's entertainment assets, including the 20th Century Fox film studio, FX Networks, National Geographic, and Fox's stake in Hulu. The deal reduced the number of major film studios from six to five. Critics of the Paramount/WBD merger argued that the Disney/Fox deal provided an "event study" — a before-and-after comparison — showing that studio consolidation leads to reduced theatrical output. The DOJ rejects this analogy on two grounds: first, the Disney/Fox deal closed in early 2019, just months before the COVID-19 pandemic fundamentally disrupted the film industry, making any output comparison unreliable; and second, Disney's business model (built around franchise IP that it monetizes across theme parks, merchandise, and streaming) gives it different output incentives than a "pure-play media business like Paramount."

These three companies represent a new generation of studios that have disrupted the traditional Hollywood model. A24, founded in 2012, has become a cultural force with critically acclaimed films including Everything Everywhere All at Once (2022, which won seven Academy Awards) and Moonlight (2016, Best Picture). NEON, founded in 2017, similarly built a reputation for prestige content, distributing the Best Picture winner Parasite (2019). Blumhouse Productions, founded in 2000 by Jason Blum, pioneered a low-budget, high-return model for horror films (Get Out, The Purge franchise) and has since expanded into larger productions. The DOJ's point is that these companies — none of which existed in the era of the traditional "Big Six" studios — demonstrate that the theatrical film market is more competitive and accessible than historical market share data would suggest.

In antitrust merger investigations, documents and information that companies produce to the DOJ under a Hart-Scott-Rodino (HSR) filing or a civil investigative demand are subject to statutory confidentiality protections. The DOJ generally cannot share this material with state attorneys general without the merging parties' consent. A "voluntary waiver" is the mechanism by which the parties agree to let the DOJ and state AGs share investigative materials. While technically optional, refusing to grant a waiver can signal a lack of cooperation and may draw additional scrutiny. Here, the waivers allowed state AGs to attend and participate in the DOJ's depositions of company executives — a meaningful level of state involvement in a federal merger review.

Each of these prior deals shaped the entity that Paramount is now acquiring. AOL's merger with Time Warner in 2001 was valued at $164 billion and is widely regarded as one of the worst corporate mergers in history — AOL's dial-up internet business declined rapidly, and the combined company wrote down nearly $100 billion in value. AT&T's $85 billion acquisition of Time Warner in 2018 was the subject of a landmark antitrust challenge by this same Division, which sued to block the deal as an anticompetitive vertical merger. The DOJ lost at trial before Judge Richard Leon in the D.C. District Court, and the D.C. Circuit unanimously affirmed. It was the first vertical merger case the Division had tried in forty years. AT&T subsequently reversed course and spun off WarnerMedia, which merged with Discovery Inc. in 2022 to form Warner Bros. Discovery. The Division's reference to "challenges that arise when the commercial rationale for a deal lacks clear alignment" is a pointed allusion to these failed combinations.

In December 2025, Netflix agreed to acquire WBD's studio and streaming divisions (including Warner Bros. Studios, HBO, and Max) in a deal initially structured as cash and stock, later revised to all cash, valued at approximately $72 billion ($82.7 billion including WBD debt). The agreement would have spun off WBD's linear television networks as a separate entity. The Netflix deal drew significant concern within the entertainment industry about further streaming consolidation. In December 2025, Paramount launched a rival all-cash tender offer for the entirety of WBD — not just its studio and streaming assets. After a contractual waiver allowed WBD to negotiate with Paramount, the WBD board determined on February 26, 2026, that Paramount's revised offer of $31 per share (approximately $110.9 billion total enterprise value) was a "Company Superior Proposal." Netflix declined to match and withdrew. The DOJ's statement notes that it reviewed both proposals, benefiting from the "comparative perspectives" each bidder presented on the media landscape.

Paramount Skydance Corporation (NASDAQ: PSKY) was formed in August 2025 when Skydance Media completed its merger with Paramount Global. Skydance was founded in 2006 by David Ellison, the son of Oracle co-founder Larry Ellison. The Ellison family and RedBird Capital Partners backed the transaction, with David Ellison becoming chairman and CEO of the combined company. The Skydance-Paramount deal was itself a major media consolidation event, combining Skydance's production capabilities (Top Gun: Maverick, Mission: Impossible) with Paramount's studio, broadcast network (CBS), cable channels, and streaming service (Paramount+). Just six months after completing that merger, the new entity pivoted to acquiring Warner Bros. Discovery — the transaction at issue here.

Section 106 of the National Historic Preservation Act is the federal government's primary mechanism for considering the effects of its projects on historic properties before spending federal money. Under the process, a federal agency must identify historic properties that might be affected by a proposed project, assess potential adverse effects, and consult with the State Historic Preservation Officer, tribal officials, and other interested parties before proceeding. The Advisory Council on Historic Preservation (ACHP) oversees the process and may comment on the adequacy of the agency's review. Importantly, Section 106 does not give the ACHP or any other body a veto over federal projects — it requires only that the agency "take into account" the effects on historic properties. However, the consultation process frequently leads to project modifications that mitigate damage to historic resources. The legal question here is whether the Kennedy Center — which is not technically a "Federal agency" under the APA — is nonetheless required to follow Section 106 by virtue of a 2003 statute that brought the Smithsonian within the process for certain types of projects.

The East Wing of the White House — which housed the offices of the First Lady and the White House social secretary, among other functions — was demolished in early 2026 as part of President Trump's renovation plans for the executive residence. The demolition was particularly controversial because President Trump had previously promised that construction would "pay total respect to the existing building." The East Wing demolition prompted its own federal lawsuit — National Trust for Historic Preservation v. National Park Service (D.D.C., Judge Lamberth) — which is cited elsewhere in this opinion. Judge Cooper references the East Wing episode here to explain why there were legitimate fears that the Kennedy Center might face a similarly drastic fate despite assurances from its leadership.

Judge Cooper is extending his "Hail Mary pass" metaphor with a reference to one of the most famous plays in NFL history. On December 28, 1975, in an NFC divisional playoff game, Dallas Cowboys quarterback Roger Staubach threw a desperate 50-yard touchdown pass to wide receiver Drew Pearson with 24 seconds remaining, defeating the Minnesota Vikings 17–14. After the game, Staubach told reporters he closed his eyes and said a "Hail Mary" — popularizing the now-universal term for any last-second, low-probability deep pass. Judge Cooper's point: just as Staubach's improbable pass found its receiver in the end zone, Beatty's long-shot ultra vires claim connects.

Ultra vires is a Latin phrase meaning "beyond the powers." In administrative law, an ultra vires claim argues that a government entity or official has acted entirely outside the authority Congress granted — not merely that the action was unwise or even illegal, but that it was fundamentally unauthorized. The doctrine provides a narrow but important avenue for judicial review when no other mechanism (such as the Administrative Procedure Act) is available. As the court notes, ultra vires claims are deliberately hard to win: the Supreme Court has "strictly limited" their scope, requiring the plaintiff to show that the challenged action plainly disregards a specific, unambiguous statutory command. The standard is so demanding that courts sometimes call it a "Hail Mary pass" — a metaphor Judge Cooper extends with evident relish in the next paragraph.

In Dong (1997), the D.C. Circuit held that the Smithsonian Institution is not a federal "agency" as defined by the Administrative Procedure Act, because it was established as an independent trust instrumentality rather than an executive department or government corporation. This ruling has cascading consequences throughout this case. Because the Kennedy Center is a bureau within the Smithsonian, it too falls outside APA coverage — which is why the court grants summary judgment to the Defendants on Beatty's APA claim (Count Six). The ruling also means the Kennedy Center may not qualify as a "Federal agency" under the National Historic Preservation Act, potentially exempting its renovation from the Section 106 historic preservation review process — a gap that Congress only partially addressed through the Smithsonian Facilities Authorization Act of 2003.

Thole (2020) was a 5–4 Supreme Court decision holding that beneficiaries of a defined-benefit pension plan lacked Article III standing to sue under ERISA for alleged mismanagement of plan assets. Because the plaintiffs were guaranteed the same monthly payments regardless of how the plan's investments performed, the Court found they had no concrete stake in the lawsuit. Critically, however, the majority distinguished trustees and other fiduciaries from mere beneficiaries, noting that fiduciaries have independent legal obligations that give them standing to sue to protect trust assets. Judge Cooper relies on this distinction to support Representative Beatty's standing: as a trustee rather than a beneficiary, she has fiduciary duties that give her a personal stake the Thole plaintiffs lacked.

Loper Bright (2024) is one of the most consequential Supreme Court decisions in recent decades. In a 6–3 ruling authored by Chief Justice Roberts, the Court overruled Chevron U.S.A., Inc. v. Natural Resources Defense Council (1984), eliminating the 40-year-old doctrine under which courts deferred to federal agencies' reasonable interpretations of ambiguous statutes they administered. The Court held that the Administrative Procedure Act requires courts to exercise independent judgment in interpreting federal statutes. Judge Cooper cites Loper Bright here for the specific point that the Kennedy Center's officers' internal interpretation of "trust funds" receives no judicial deference — but the citation signals the broader post-Chevron principle that courts, not administrators, are the final arbiters of statutory meaning.

The One Big Beautiful Bill Act was a comprehensive budget reconciliation bill passed by the 119th Congress and signed by President Trump in July 2025. The legislation — which exceeded 1,000 pages — addressed tax policy, immigration enforcement, energy production, defense spending, and numerous other subjects. The $257 million Kennedy Center appropriation was one of many discrete spending provisions within the broader bill. Budget reconciliation is a special legislative procedure that allows certain spending, revenue, and debt-limit legislation to pass the Senate by simple majority, bypassing the 60-vote threshold typically required to overcome a filibuster. The Kennedy Center appropriation is authorized through September 30, 2029 — a deadline that becomes significant later in the opinion when the Defendants argue that a preliminary injunction could jeopardize the project timeline.

Richard Grenell is a Republican political figure who served as U.S. Ambassador to Germany (2018–2020) and Acting Director of National Intelligence (February–May 2020) during President Trump's first term. His professional background was in public diplomacy, communications, and national security rather than arts administration or nonprofit management. As the opinion notes, Grenell was subsequently replaced as the Center's senior executive by Matthew Floca, whose background is in facilities and project management — meaning that neither of the two individuals who have led the Kennedy Center since the Board's reconstitution had prior experience in performing arts administration.

The REACH (an acronym for Renew, Experience, Art, Community, and Humanity) is a 72,000-square-foot expansion of the Kennedy Center campus that opened in September 2019 — the first major addition to the site since the original building opened in 1971. Designed by Steven Holl Architects, the REACH consists of three interconnected pavilion buildings set into the landscape south of the main building, connected by gardens and outdoor performance spaces. It houses rehearsal studios, classrooms, a video wall, and flexible event spaces. Because the REACH is a separate structure from the main building, it figures prominently in the Defendants' argument that some programming, educational, and memorial functions could continue there during the main building's closure.

Edward Durell Stone (1902–1978) was an American architect known for monumental public buildings in an ornate modernist style. His other major works include the U.S. Embassy in New Delhi (1954), the Museum of Modern Art's original building in New York (1939, with Philip Goodwin), and the General Motors Building in Manhattan (1968). The Kennedy Center, completed in 1971, is characteristic of Stone's later career — the grand marble façade and columned Hall of States and Hall of Nations are his design. The building was recognized for its architectural significance when it was listed on the National Register of Historic Places, a designation that is relevant to the historic preservation questions discussed later in this opinion.

A "trust instrumentality" is a category of federal entity that holds property in trust for the benefit of the American public. Unlike a typical federal agency, a trust instrumentality is not subject to the Administrative Procedure Act and operates with a degree of institutional independence. The Kennedy Center shares this classification with only a handful of other institutions, including the Smithsonian Institution (of which it is technically a bureau) and the National Gallery of Art. The classification is central to this opinion because it means the Center's Board members owe fiduciary duties — including duties of prudence, loyalty, and care — derived from the common law of trusts. These are the same duties that govern private trustees, and they provide the legal framework for every claim in this case.