One Judge, One Nation?

National Injunctions in the United States and the Search for Their Canadian Equivalent

Matthew Livingstone

July 2026

Abstract. This essay examines the rise, uses, and abuses of national (universal) injunctions in the United States, culminating in Trump v. CASA, Inc. (2025) and Trump v. Barbara (2026), and compares the American experience with Canada, where courts achieve nationwide effects through declarations of invalidity, disciplined interlocutory doctrine, and a unified appellate structure. It argues that broad remedies are sometimes necessary in a constitutional democracy, that they become dangerous when they are procedurally cheap, and that the two countries represent different calibrations of the same underlying trade-off between immediate protection and deliberative process.

Introduction

On January 20, 2025, hours after being sworn in for a second term, President Donald Trump signed Executive Order 14160, which purported to end birthright citizenship for children born in the United States to parents who were unlawfully or only temporarily present in the country (Congressional Research Service [CRS], 2025a). Within days, federal trial judges in Washington State, Maryland, and Massachusetts had blocked the order — not just for the families and states that sued, but for everyone, everywhere in the country. A single judge, sitting in a single courtroom, had frozen a signature presidential policy from coast to coast before it ever took effect.

Depending on your perspective, that sequence of events is either the constitutional system working exactly as designed or a symptom of something deeply broken. To supporters, these orders — commonly called national injunctions, nationwide injunctions, or universal injunctions — are an essential emergency brake. When the government adopts a policy that is plainly unlawful and that will harm hundreds of thousands of people at once, it makes little sense to protect only the handful of plaintiffs who managed to file a lawsuit first. To critics, the same orders represent a startling concentration of power: one unelected trial judge, potentially selected through strategic “forum shopping,” can override the policy judgment of a president elected by tens of millions of voters, and can do so provisionally, before any appellate court has weighed in.

This tension sits at the heart of modern American constitutional and administrative law. National injunctions have shaped fights over immigration, health care, environmental regulation, student loans, vaccine mandates, and citizenship itself. They have been wielded against Democratic and Republican administrations alike, and both parties have, with impressive symmetry, denounced them when in power and embraced them when in opposition. In June 2025, the U.S. Supreme Court intervened dramatically in Trump v. CASA, Inc., holding that federal courts generally lack the authority to issue universal injunctions at all (Trump v. CASA, Inc., 2025). Yet within hours of that ruling, litigants had pivoted to a class action, and one year later — on June 30, 2026 — the Supreme Court struck down the birthright citizenship order on the merits in Trump v. Barbara, in a case that reached the Court precisely through the class-wide pathway that CASA left open (Trump v. Barbara, 2026). The remedy changed its clothes; the underlying question of judicial power did not go away.

This essay examines that question from both sides, and then looks north. Canada makes an unusually instructive comparison. It is a federal constitutional democracy with a powerful, rights-enforcing judiciary, a written charter of rights, and courts that regularly strike down national legislation. Yet Canada has never had an American-style “national injunction” controversy — not because Canadian courts lack broad remedial powers (in some respects their powers are broader), but because Canada’s legal architecture channels those powers differently. Studying what Canada does instead — declarations of invalidity under section 52 of the Constitution Act, 1982, suspended declarations, a restrained approach to interlocutory relief against legislation, and a unified appellate structure — helps clarify which parts of the American debate are about judicial power in principle and which are about the particular plumbing of the American court system.

The essay proceeds in three movements. The first explains what national injunctions are, how they emerged, and why reasonable people defend and attack them. The second traces the American debate through its major cases and the current post-CASA landscape, including proposed reforms. The third turns to Canada, identifies the closest Canadian equivalents, asks whether those remedies generate similar controversy, and draws the two systems into direct comparison. The essay does not begin from the premise that national injunctions are good or bad. It ends, instead, with a balanced position: broad remedies are sometimes genuinely necessary in a constitutional democracy, but they are dangerous when they are cheap — easy to obtain, easy to shop for, and hard to review — and the real design question for any legal system is how to make them appropriately expensive.

What Are National Injunctions?

An injunction is a court order directing someone to do something or, more commonly, to stop doing something. If a factory is polluting your farmland, a court can order the factory to stop. If your former employee is about to leak trade secrets, a court can order them not to. Disobeying an injunction is contempt of court, punishable by fines or imprisonment. Injunctions are “equitable” remedies, meaning they descend from the flexible, conscience-based justice historically administered by England’s Court of Chancery rather than from the rigid forms of the common law courts (Bray, 2017).

A national injunction is an injunction against the government with a distinctive feature: it prohibits the government from enforcing a challenged law, regulation, or policy against anyone, not merely against the plaintiffs who brought the lawsuit (CRS, 2025a). Suppose a new federal rule affects ten million people, and a single advocacy group sues on behalf of its five thousand members. An ordinary injunction would block the rule as applied to those five thousand members. A national injunction blocks the rule as applied to all ten million people — including the 9,995,000 who never sued, never appeared in court, and may not even know the case exists.

Terminology matters here more than in most legal debates, because the labels quietly smuggle in arguments. The most common term, “nationwide injunction,” emphasizes geography: the order applies across the whole country. But scholars such as Samuel Bray have argued that geography is a red herring. Even a perfectly conventional injunction can have nationwide geographic reach — if a court orders the government not to deport a specific plaintiff, that protection follows the plaintiff to every state (Bray, 2017). What makes these injunctions distinctive is not where they apply but who they protect. For that reason, many scholars and now the Supreme Court prefer the term “universal injunction”: an injunction that runs universally, protecting parties and non-parties alike (Trump v. CASA, Inc., 2025). This essay uses “national injunction” and “universal injunction” interchangeably, following the essay prompt’s framing and common usage, while flagging that the defining feature is protection of non-parties.

Several neighbouring concepts must be kept distinct. A preliminary (or interlocutory) injunction is issued early in a case, before trial, to preserve the status quo while the litigation proceeds; a permanent injunction is issued at the end of a case, after the court has finally ruled on the merits. Most controversial national injunctions are preliminary — they pause national policy based on a prediction that the challengers are likely to win, sometimes years before any final judgment (Sidley Austin LLP, 2025). Declaratory relief is a court’s formal statement of the parties’ legal rights — for example, a declaration that a statute is unconstitutional — without a coercive order backed by contempt. Class-wide remedies protect non-parties in a different, more disciplined way: under Rule 23 of the Federal Rules of Civil Procedure, a court can certify a class of similarly situated people and issue an injunction protecting the entire class, but only after the plaintiffs satisfy procedural requirements such as commonality, typicality, and adequacy of representation (Federal Rules of Civil Procedure, r. 23). Finally, “vacatur” under the Administrative Procedure Act (APA) — a court order “setting aside” an unlawful agency rule — functionally removes the rule from the books for everyone, making it a close statutory cousin of the universal injunction, though its proper scope is itself hotly contested (Arnold & Porter, 2025).

Table 1

Four remedies compared: who is protected, and at what procedural price

Feature Ordinary injunction National (universal) injunction Class action injunction Declaratory judgment

Who is protected The plaintiffs before the court Everyone affected by the policy, party or not All members of a certified class Formally, the parties; practically, guides government conduct

When available Preliminary or final Historically preliminary or final; sharply limited by CASA (2025) Preliminary (provisional certification) or final Usually at final judgment

Procedural screen Standing + injunction factors Same as ordinary injunction (the core criticism) Rule 23: commonality, typicality, adequacy, cohesion Standing + actual controversy

If challengers lose Plaintiffs bound by judgment Non-party beneficiaries not bound; can sue again elsewhere Entire class bound, win or lose Parties bound by declaration

Enforcement Contempt power Contempt power Contempt power No contempt; relies on government compliance

Note. Compiled from Bray (2017), Frost (2018), Federal Rules of Civil Procedure r. 23, and Trump v. CASA, Inc. (2025).

Why does this vocabulary matter? Because the debate is often less about whether courts may ever protect non-parties (nearly everyone concedes they sometimes may, for instance through class actions) and more about which procedural doorway they must walk through to do it. Calling an order a “nationwide injunction” makes it sound like a question of scale; calling it a “universal injunction” makes it sound like a question of judicial power over strangers to the litigation; calling the alternative a “class action” makes it sound like a mere technicality. As we will see, after CASA the entire American fight has effectively become a fight about doorways.

How Injunctions Normally Work

To understand why national injunctions are contested, it helps to start with what courts traditionally do. The classic model of adjudication is bilateral: a plaintiff claims the defendant has wronged them, the court decides, and the remedy runs between those parties. As the Supreme Court has put it, injunctive relief should ordinarily “be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs” (Califano v. Yamasaki, 1979). The judgment binds the parties; the court’s reasoning binds no one, though it may persuade other courts as precedent (Trump v. CASA, Inc., 2025).

A homely analogy: think of a court as a referee in a dispute between two neighbours over a fence. The referee can order this neighbour to move this fence. What the referee traditionally cannot do is announce that, having studied the matter, all fences of this type on every street in the country must come down — including fences belonging to people who never asked for a ruling and might quite like their fences. If other neighbours want relief, they bring their own cases, and over time the accumulation of decisions — especially decisions of appellate courts — settles the law for everyone. This is how the common law generalizes: through precedent, not through the remedial scope of any single order.

The bilateral model works smoothly when the defendant is a private party. It becomes awkward when the defendant is the national government and the challenged conduct is a uniform national policy. Three features create the awkwardness. First, scale: a single executive order can affect millions of people identically, and it seems arbitrary — even perverse — for its lawfulness to be adjudicated one plaintiff at a time. Second, equality: if a court holds that a policy is unconstitutional as applied to the plaintiff, allowing the government to keep applying the same policy to identical non-parties looks like unequal justice. The government would be enforcing, against most of the population, a policy that a court has already declared unlawful. Third, irreversibility: some harms cannot be undone. A person deported under an unlawful policy, or a child denied citizenship documents at birth, may suffer consequences that no later judgment can fully repair (Frost, 2018).

On the other hand, broad remedies short-circuit the very process by which the legal system corrects its own errors. If the first court to hear a challenge freezes the policy for the whole country, other courts never get to weigh in, the issue never “percolates” through multiple circuits, and the Supreme Court must resolve the question on a rushed emergency posture rather than with the benefit of considered, conflicting appellate opinions (Bray, 2017). A narrow remedy protects the plaintiff and lets the law develop; a broad remedy protects everyone and stops the law’s development cold. That trade-off — protection now versus deliberation over time — recurs throughout this essay.

The Rise of National Injunctions in the United States

National injunctions are a genuinely modern phenomenon, though the precise birthdate is disputed. The Supreme Court’s majority in CASA, drawing on Bray’s historical work, concluded that the universal injunction was “conspicuously nonexistent for most of our Nation’s history,” absent from eighteenth- and nineteenth-century equity practice and essentially unknown until the mid-twentieth century (Trump v. CASA, Inc., 2025; Bray, 2017). In the early republic, courts even refused to entertain generalized challenges to federal statutes: in Frothingham v. Mellon (1923), the Court held that a taxpayer could not sue to block a federal spending program merely as one member of the affected public, reflecting the view that courts redress particular injuries rather than supervise government at large (Frothingham v. Mellon, 1923). Other scholars, notably Mila Sohoni, have pushed back, identifying early twentieth-century cases in which federal courts enjoined the enforcement of statutes in terms that protected non-parties, and arguing that the practice has a longer pedigree than critics allow (Sohoni, 2020). The historical debate matters because, as discussed below, the Supreme Court ultimately made history the touchstone of its analysis.

Whatever its origins, the modern explosion is undeniable and, importantly, bipartisan in its targets. Then-Attorney General William Barr stated in 2019 that federal courts issued only 27 nationwide injunctions in the entire twentieth century (CRS, 2025a). A Harvard Law Review study counted six nationwide injunctions against the George W. Bush administration, twelve against the Obama administration, sixty-four against the first Trump administration, and fourteen during the first three years of the Biden administration (CRS, 2025a). By CRS’s count, courts issued nationwide injunctions in 34 separate cases in just the first five months of the second Trump administration — between January 20 and June 27, 2025 — targeting policies ranging from birthright citizenship to federal funding freezes, mass layoffs, refugee admissions, and deportations under the Alien Enemies Act (CRS, 2025a).

The trajectory tells a story about litigation strategy as much as judicial philosophy. During the Obama years, Republican state attorneys general perfected the playbook: sue in a favourable district (often in Texas), obtain a nationwide preliminary injunction from a sympathetic judge, and stall the policy — as happened to the Deferred Action for Parents of Americans (DAPA) immigration program in 2015 and to Obama-era rules on overtime pay and transgender students. During the first Trump administration, Democratic attorneys general and civil liberties groups ran the same playbook in reverse, obtaining nationwide injunctions against the travel ban, the rescission of the Deferred Action for Childhood Arrivals (DACA) program, border wall funding, and asylum restrictions, typically filing in districts within the Ninth Circuit. Under President Biden, the geography flipped again: single-judge divisions in Texas produced nationwide orders against vaccine mandates, immigration enforcement priorities, and — most dramatically — the Food and Drug Administration’s decades-old approval of the abortion medication mifepristone (Stanford Law School, 2025). Each side’s outrage at the other’s injunctions was sincere; each side’s use of them was enthusiastic. As Justice Barrett dryly noted in CASA, by the end of the Biden administration the United States had reached “a state of affairs where almost every major presidential act was immediately frozen by a federal district court,” and the solicitors general of both parties had asked the Court to rein the practice in (Trump v. CASA, Inc., 2025).

Three structural features of the American system fed this growth. First, the United States has 94 federal district courts staffed by roughly 677 district judges, each empowered to hear constitutional challenges to federal action, and some districts have single-judge divisions where plaintiffs know in advance exactly which judge will hear their case. Second, preliminary relief is available early and reviewed deferentially, so a favourable trial judge can deliver most of the practical value of final victory within weeks of filing. Third, there is no fast, unified mechanism — short of the Supreme Court’s discretionary and increasingly strained emergency docket — to reconcile conflicting district court views quickly. Put those together and the incentive structure practically writes itself: whoever loses a national election heads to their favourite courthouse (Stanford Law School, 2025).

Why National Injunctions May Be Necessary

The affirmative case for national injunctions begins with a simple observation: sometimes plaintiff-only relief is not merely inadequate — it is incoherent. The birthright citizenship litigation supplies the cleanest example. If courts could protect only named plaintiffs, then citizenship at birth would depend on whether a newborn’s parents happened to have joined a lawsuit. Two babies born in the same hospital on the same day would have different constitutional statuses. Justice Sotomayor pressed exactly this point in dissent in CASA: some rights are, by their nature, indivisible or administrable only as uniform rules, and a remedy that fractures them produces chaos rather than justice (Trump v. CASA, Inc., 2025). Even the CASA majority conceded that broader relief is permissible when necessary to give the plaintiffs themselves “complete relief” — acknowledging, in effect, that the bilateral model sometimes cannot contain the dispute (CRS, 2025b).

Second, national injunctions respond to the practical limits of individual litigation. Amanda Frost, the leading academic defender of the practice, argues that universal relief is appropriate when the affected population cannot realistically get to court: undocumented immigrants facing removal, children, people who do not know a policy applies to them until it is too late, or people whose individual stakes are too small to justify a federal lawsuit even though the aggregate harm is enormous (Frost, 2018). A policy that unlawfully cuts a benefit by $40 a month for two million people inflicts nearly a billion dollars of annual harm that no rational individual will ever litigate. Without some mechanism for aggregate relief, unlawful government action in this category is effectively unreviewable.

Third, there is the equality argument. The rule of law is usually understood to include the principle that like cases be treated alike. When a court has determined — even preliminarily — that a policy is unlawful, permitting the government to continue enforcing it against everyone except the plaintiffs institutionalizes unequal treatment based on nothing more than litigation status (Frost, 2018). It also creates perverse incentives for the government: an administration could lose the same issue in fifty separate suits and keep enforcing the policy against the millions who had not yet sued, extracting value from unlawfulness at scale. Defenders argue the injunction merely makes the government stop doing, to anyone, the thing a court has found unlawful — which is what a law-abiding government should do anyway.

Fourth, national injunctions economize on judicial resources. Requiring thousands of identical suits to establish the same point multiplies cost, delay, and the risk of inconsistent outcomes across regions — a particular problem for regulated industries and for administrative schemes that presuppose national uniformity, such as immigration, taxation, and interstate commerce (CRS, 2025a). Federal immigration law in particular has long been described by courts as demanding a uniform national rule, which is one reason immigration cases dominate the national injunction docket.

Finally, defenders invoke the emergency-brake function. Some executive actions inflict harms that are irreversible before appellate review can occur: deportations to danger, destruction of records, expenditure of appropriated funds, the dismantling of an agency. In such cases, a preliminary injunction limited to the plaintiffs simply fails to preserve the court’s ability to decide the case in any meaningful sense. On this view, the universal injunction is not judicial imperialism but judicial self-defence — a way of ensuring that the eventual, considered answer of the appellate courts still matters by the time it arrives (Frost, 2018). The judiciary’s answer in Trump v. Barbara took a year to arrive; the injunctions (and later the class-wide order) are the only reason that answer governed a live policy rather than a fait accompli (American Civil Liberties Union [ACLU], 2026).

How National Injunctions Can Be Abused

The critique begins where the defence ends: with power. A national injunction gives one district judge — the most junior tier of the federal judiciary, sitting alone, often ruling within days on a thin record — the ability to suspend the policy of a co-equal branch of government for the entire nation. Critics call this “government by district judge.” The asymmetry is stark: the challengers need to win only once, anywhere, to freeze the policy everywhere, while the government must win everywhere to implement it anywhere. Justice Gorsuch has described this as allowing plaintiffs to play the litigation lottery with house money: file in multiple districts, lose nine times, win the tenth, and the tenth controls (Bray, 2017; Trump v. CASA, Inc., 2025).

That asymmetry produces the second pathology: forum shopping. Because the payoff to finding a sympathetic judge is national, litigants invest heavily in judicial selection. Conservative litigants during the Biden years famously filed in single-judge divisions in Texas — most notoriously the Amarillo division, where any filed case was virtually certain to be assigned to one particular judge — while progressive litigants during Trump’s terms clustered filings in districts in Washington, Massachusetts, Maryland, and California (Stanford Law School, 2025). The Judicial Conference of the United States was sufficiently alarmed that in March 2024 it issued guidance urging random, district-wide assignment of cases seeking to enjoin state or federal policies (Judicial Conference of the United States, 2024). Forum shopping does more than skew outcomes; it corrodes the perception of courts as neutral arbiters, converting the map of federal districts into a partisan chessboard and inviting the public to view judicial rulings as products of judge-selection rather than law.

Third, national injunctions distort the appellate process that normally disciplines judicial error. The federal system is designed for “percolation”: multiple district courts and courts of appeals consider a question, disagreement sharpens the arguments, and the Supreme Court resolves the conflict with the benefit of that record. A universal injunction ends percolation at the first ruling. Worse, it shifts enormous consequential decisions onto the Supreme Court’s emergency (“shadow”) docket, where the justices rule quickly, often without full briefing, argument, or reasoned opinions — a forum poorly suited to the most important questions in the country (Coglianese & Wiener, 2025). The government, facing a nationwide freeze, cannot wait years for orderly review; the Court, facing a flood of emergency applications, cannot give each the deliberation it deserves.

Fourth is the democratic objection. Preliminary national injunctions can stall a policy for most of a presidential term without any final determination that it is actually unlawful. Whatever one thinks of any given policy, a system in which the practical ability of an elected government to govern depends on winning a race against dozens of strategically filed lawsuits raises legitimate concerns about accountability. The policy dies or survives based on interim procedure rather than democratic judgment or considered constitutional decision. Critics add a subtle point about judicial behaviour: the availability of universal relief changes what litigants ask for and how courts see their role, encouraging judges to opine on abstract legality of policies rather than resolve concrete disputes — turning courts, in Bray’s phrase, into multiple competing chancellors superintending the executive (Bray, 2017).

Finally, there is a symmetry warning that honest partisans on both sides have learned to take seriously. A remedy powerful enough to stop a president you fear is powerful enough to stop a president you support. Nationwide injunctions halted DAPA under Obama, the travel ban under Trump, vaccine mandates and student loan relief under Biden, and dozens of second-term Trump initiatives — and in the mifepristone litigation, a single judge’s order threatened to suspend nationwide access to a medication the FDA had approved for over twenty years (FDA v. Alliance for Hippocratic Medicine, 2024). The tool does not care who holds it. That is either its greatest virtue (it is genuinely neutral) or its greatest danger (it arms whoever is angriest), and often both.

Major U.S. Case Studies

The abstract arguments become vivid in the cases. The five episodes below span immigration, national security, administrative law, public health, and constitutional rights, and they involve injunctions against presidents of both parties.

Texas v. United States (2015): DAPA and the modern template

In 2014, President Obama announced the DAPA program, which would have deferred deportation for roughly four million undocumented parents of American citizens and permanent residents. Twenty-six states, led by Texas, sued in the Southern District of Texas, alleging the program exceeded executive authority under the immigration statutes and the APA. Judge Andrew Hanen issued a nationwide preliminary injunction blocking DAPA before it began, reasoning that immigration policy demands national uniformity, and the Fifth Circuit affirmed; an equally divided Supreme Court (4–4 after Justice Scalia’s death) left the injunction in place, and the program never took effect (United States v. Texas, 2016). DAPA established the modern template — a coalition of state attorneys general, a carefully chosen district, a preliminary injunction with national scope — and it demonstrated the remedy’s decisive power: a major presidential initiative was killed by interim relief without any final merits judgment from the Supreme Court. Defenders note the states raised serious legal objections that appellate courts endorsed; critics note that a program affecting millions died in procedural limbo, chosen forum and all.

Trump v. Hawaii (2018): the travel ban and the first judicial alarm

President Trump’s 2017 executive orders restricting entry from several majority-Muslim countries were immediately blocked by nationwide injunctions from district courts in Washington, Hawaii, and Maryland. The litigation ping-ponged through the Ninth and Fourth Circuits until the Supreme Court upheld the third version of the policy in 2018 (Trump v. Hawaii, 2018). The case is a landmark for two reasons. Substantively, it showed nationwide injunctions operating as a real-time check on a contested national security policy — and showed the government ultimately winning on the merits, which critics cite as proof that interim universal relief had blocked a lawful policy for over a year. Institutionally, Justice Thomas’s concurrence delivered the first sustained attack on universal injunctions from within the Court, arguing they were “legally and historically dubious” and inconsistent with founding-era equity — the opening brief, as it turned out, for CASA seven years later (Trump v. Hawaii, 2018).

Department of Homeland Security v. Regents (2020): DACA and administrative law

When the first Trump administration moved to rescind DACA in 2017, district courts in California, New York, and the District of Columbia issued nationwide orders keeping the program alive. The Supreme Court ultimately held the rescission unlawful — not because ending DACA was substantively forbidden, but because the government’s explanation was arbitrary and capricious under the APA (Department of Homeland Security v. Regents of the University of California, 2020). Regents illustrates the administrative law wing of the debate: much of what looks like constitutional warfare is actually fought under the APA, where courts “set aside” unlawful agency action — a remedy that, by its nature, tends to operate against the rule itself rather than merely between the parties. It also illustrates the necessity argument at its strongest: a plaintiff-by-plaintiff remedy for hundreds of thousands of DACA recipients was never realistic, and nationwide relief preserved the status quo long enough for orderly Supreme Court review.

Alliance for Hippocratic Medicine v. FDA (2023–2024): mifepristone and the shoe on the other foot

In 2023, an association of anti-abortion physicians sued the FDA in the Amarillo division of the Northern District of Texas — where case assignment made the presiding judge a near-certainty — seeking to invalidate the agency’s 2000 approval of mifepristone. The district court issued an order suspending the approval nationwide, which would have removed the drug from the market in every state, including states where abortion remained legal. The Supreme Court stayed the order and, in 2024, held unanimously that the plaintiffs lacked standing (FDA v. Alliance for Hippocratic Medicine, 2024). For many progressives who had defended nationwide injunctions during the Trump years, mifepristone was the moment the weapon swung around: a single strategically selected judge had come within a stay of overriding a two-decade-old scientific judgment for the entire country, at the request of plaintiffs the Supreme Court unanimously found had no legal injury at all. The episode did more than any law review article to make forum shopping a mainstream concern (Stanford Law School, 2025).

The birthright citizenship litigation (2025–2026): CASA and Barbara

Executive Order 14160 produced the confrontation that changed the law. Judges in three districts issued universal preliminary injunctions; the courts of appeals declined to narrow them; and the administration, in a deliberate strategic choice, asked the Supreme Court to address only the scope of the injunctions, not the constitutionality of the order (Sidley Austin LLP, 2025). The result, discussed in detail below, was CASA’s near-abolition of the universal injunction — followed immediately by a nationwide class action, Barbara v. Trump, filed by the ACLU in New Hampshire within hours of the CASA decision. The district court certified a nationwide class of affected children and issued a class-wide preliminary injunction on July 10, 2025; the Supreme Court took the case before judgment (Congressional Research Service, 2026) and, on June 30, 2026, struck down the executive order by a 6–3 vote, with Chief Justice Roberts writing that the Fourteenth Amendment’s Citizenship Clause guarantees citizenship to children born in the United States to parents unlawfully or temporarily present (Trump v. Barbara, 2026; ACLU, 2026). The full arc — universal injunctions, their abolition, their functional resurrection as a class remedy, and a merits ruling against the government — is the single best case study of what the remedy does, what its critics achieved, and what its abolition did not change.

Table 2

A timeline of the American national injunction controversy

Year(s) Case / event Administration Significance

1923 Frothingham v. Mellon Harding Court refuses a generalized taxpayer challenge to federal spending; remedies remain tied to particular injuries.

1900s Twentieth century overall Various Roughly 27 nationwide injunctions issued in the entire century, by the Justice Department’s count.

2015–16 Texas v. United States (DAPA) Obama State coalition wins nationwide preliminary injunction; 4–4 Supreme Court leaves it in place; program dies without a merits ruling.

2017–18 Trump v. Hawaii (travel ban) Trump I Multiple nationwide injunctions; Supreme Court upholds final version; Justice Thomas questions universal injunctions.

2020 DHS v. Regents (DACA) Trump I Nationwide orders preserve DACA; rescission held arbitrary and capricious under the APA.

2021–24 Biden-era challenges Biden Nationwide orders halt vaccine mandates, immigration priorities, student-loan relief; 14 nationwide injunctions in first three years.

2023–24 AHM v. FDA (mifepristone) Biden Single judge in a hand-picked division suspends a 23-year-old drug approval nationwide; Supreme Court finds no standing.

Jan–Jun 2025 Second-term wave Trump II Nationwide injunctions issue in 34 cases in five months, including against Executive Order 14160 (birthright citizenship).

Apr 2025 No Rogue Rulings Act passes House Trump II House votes 219–213 to limit injunctions to parties; bill stalls in the Senate.

Jun 27, 2025 Trump v. CASA, Inc. Trump II Supreme Court, 6–3, holds universal injunctions exceed courts’ equitable authority; complete relief becomes the ceiling.

Jul 10, 2025 Barbara v. Trump class certified Trump II Nationwide class of affected children certified within two weeks of CASA; class-wide preliminary injunction issues.

Jun 30, 2026 Trump v. Barbara Trump II Supreme Court, 6–3, strikes down the birthright citizenship order on the merits.

Note. Injunction counts from Congressional Research Service (2025a); case outcomes from sources cited in text.

The Current U.S. Legal Debate

The current landscape is defined by Trump v. CASA, Inc. (2025), so its reasoning deserves careful attention. Writing for a 6–3 majority, Justice Barrett grounded the decision not in the Constitution but in statute: the Judiciary Act of 1789, which gives federal courts jurisdiction over “all suits … in equity,” authorizes only those remedies traditionally available in the English Court of Chancery at the founding. Surveying that history — and siding with Bray over Sohoni in the scholars’ long-running dispute — the Court concluded that founding-era equity knew only party-specific remedies, and that the universal injunction was “conspicuously nonexistent for most of our Nation’s history” (Trump v. CASA, Inc., 2025; Yale Journal on Regulation, 2025). Universal injunctions therefore “likely exceed” the equitable authority Congress granted. The Court stayed the birthright citizenship injunctions to the extent they were “broader than necessary to provide complete relief to each plaintiff with standing to sue” (Trump v. CASA, Inc., 2025).

The qualifications are as important as the holding. First, the complete-relief principle survives: a court may still protect non-parties where doing so is genuinely necessary to remedy the plaintiff’s own injury — the Court cited nuisance-style cases where the only workable injunction stops the conduct altogether (Sidley Austin LLP, 2025). The Court expressly declined to decide whether the state plaintiffs — who argued that children move across state lines, making a patchwork rule unadministrable for state benefit systems — needed nationwide relief for that reason, remanding the question (CRS, 2025b). Second, the decision rested on statutory rather than constitutional grounds, leaving Congress theoretically free to authorize universal injunctions by legislation. Third, every member of the Court appeared to accept that Rule 23(b)(2) class actions remain a lawful route to class-wide — even nationwide — injunctive relief (Yale Journal on Regulation, 2025). Fourth, the Court left entirely open the status of APA vacatur, the workhorse remedy of administrative law, over which the justices are openly split (Arnold & Porter, 2025).

The concurrences and dissents mapped the next battlefield with remarkable candour. Justice Thomas, joined by Justice Gorsuch, insisted complete relief is a ceiling that must not become a loophole. Justice Alito warned that lax class certification or expansive state standing “may render the Court’s decision little more than a minor academic exercise” — that the universal injunction could “return from the grave under the guise of nationwide class relief” (Trump v. CASA, Inc., 2025). Justice Kavanaugh, concurring, argued that there often should be a single national answer about whether a major federal policy operates during litigation — but that the answer should come from the Supreme Court through its rulings on interim relief, not from district courts. Justice Sotomayor’s dissent, joined by Justices Kagan and Jackson, accused the majority of rewarding governmental “gamesmanship” — the administration had conspicuously not asked the Court to hold the order constitutional — and predicted the decision would strip protection from the most vulnerable while the merits remained unresolved. Justice Jackson wrote separately to warn of a “zone of lawlessness” in which the executive could act unlawfully against everyone who had not personally sued (Trump v. CASA, Inc., 2025; CRS, 2025b).

Events since June 2025 have largely vindicated Justice Alito’s prophecy about substitution, if not his fears about abuse. Within hours of CASA, the CASA plaintiffs moved for class-wide relief in Maryland, and the ACLU filed Barbara v. Trump in New Hampshire; by mid-July 2025, courts in both districts had certified nationwide classes of affected children and issued class-wide preliminary injunctions (Crowell & Moring LLP, 2025; Arnold & Porter, 2025). Other litigants pivoted to APA vacatur — persuading courts to “set aside” agency rules implementing presidential directives, since executive orders themselves are not “agency action” reviewable under the APA (Coglianese & Wiener, 2025). Still others emphasized state plaintiffs and complete-relief arguments. Commentators across the spectrum converged on the same observation: CASA closed one doorway and redirected traffic to three others — complete relief, class actions, and vacatur — each of which will now generate its own scope litigation, its own forum shopping, and its own emergency applications (Sohoni, 2025). Meanwhile the merits question CASA avoided was answered a year later in Trump v. Barbara: five justices held the executive order unconstitutional under the Citizenship Clause, and Justice Kavanaugh, concurring in the judgment, would have invalidated it on statutory grounds under 8 U.S.C. § 1401(a), producing a 6–3 result against the government (Trump v. Barbara, 2026; National Constitution Center, 2026). The pairing of the two cases frames the unresolved questions elegantly: the Court has now said both that district courts may not lightly freeze national policy for everyone and that this particular national policy was unlawful all along. Whether the year-long interval — during which the order remained blocked only because a class action succeeded — represents the system working or a near-miss depends on which side of the original debate one occupies.

Possible Reforms in the United States

Reform proposals fall into two broad families: those that restrict broad relief and those that discipline it.

The restrictive family is exemplified by the No Rogue Rulings Act of 2025 (NORRA), which passed the House of Representatives 219–213 in April 2025 on a party-line vote. NORRA would prohibit district courts from issuing injunctive relief extending beyond the parties, with one exception: when two or more states from different circuits sue the executive branch, a randomly selected three-judge panel could issue a nationwide injunction, with direct appeal available to the Supreme Court (No Rogue Rulings Act, 2025; The Hill, 2025). The bill stalled in the Senate, where it faces the sixty-vote threshold, and CASA achieved much of its aim by judicial decision months later. The strength of the restrictive approach is that it directly attacks the one-judge problem; its weakness, as the House Judiciary Committee’s dissenting views argued, is over-breadth — a flat ban would forbid even injunctions (like school desegregation orders) whose non-party effects are inherent in complete relief for the plaintiffs (U.S. House of Representatives, 2025).

The disciplining family accepts that broad relief is sometimes warranted and focuses on who grants it, how, and with what review. Its main ideas include the following. First, channel broad relief through class actions: Rule 23 already requires courts to verify that the represented group is genuinely similarly situated, binds the class to the judgment win or lose (solving the asymmetry problem, since a class that loses is precluded, unlike the beneficiaries of a universal injunction), and provides appellate review of certification. Post-CASA practice is effectively running this experiment now; the open question, flagged by Justice Alito, is whether courts will apply certification rigorously or rubber-stamp sweeping classes (Trump v. CASA, Inc., 2025). Second, attack forum shopping directly: implement the Judicial Conference’s 2024 policy of random district-wide assignment for suits seeking to block national policies, or require such suits to be filed in the District of Columbia or before three-judge panels, as was once required for constitutional challenges to federal statutes (Judicial Conference of the United States, 2024). Third, require stronger justification and findings: courts issuing relief protecting non-parties could be required to make explicit findings on why narrower relief fails, with automatic expedited appellate review — converting the scope of relief from an afterthought into a litigated question. Fourth, fix the emergency docket: expedited certiorari procedures for cases in which any court has blocked a national policy, so the definitive answer arrives in months rather than years — essentially institutionalizing Justice Kavanaugh’s view that national uniformity during litigation should be the Supreme Court’s job (CRS, 2025b).

Each reform embodies a judgment about the underlying trade-off. Party-limited relief maximizes percolation and democratic latitude at the price of unequal interim treatment and unremedied harm to non-parties. Class actions restore discipline but add cost and delay precisely when time matters most — although Barbara demonstrated the pathway can move fast when courts cooperate. Venue reform reduces judge-selection games but cannot eliminate ideological variation across the bench. And centralizing interim decisions in the Supreme Court solves uniformity while concentrating in nine justices exactly the power critics feared in one district judge — merely relocating, rather than resolving, the question of who gets to pause a presidency.

The Canadian Situation: Does Canada Have National Injunctions?

Ask a Canadian litigator for the country’s leading “national injunction” case and you will get a puzzled look — followed, usually, by a more interesting answer: Canada does not really need them. Canadian courts unquestionably issue remedies with national effect. What Canada lacks is the distinctive American package — a lone trial judge, selected by strategic filing, provisionally freezing national policy for non-parties on an emergency basis — and the reasons it lacks that package are structural, not accidental.

Start with the main event: constitutional invalidity. Section 52(1) of the Constitution Act, 1982 — Canada’s supremacy clause — provides that the Constitution is “the supreme law of Canada, and any law that is inconsistent with the provisions of the Constitution is, to the extent of the inconsistency, of no force or effect” (Department of Justice Canada, 2025a). When a Canadian superior court concludes that a statute violates the Canadian Charter of Rights and Freedoms or the federal division of powers, it issues a declaration of invalidity: the offending provision is of no force or effect. Crucially, this remedy operates on the law itself, not on the parties. Nobody speaks of a declaration “protecting non-parties,” because that framing does not fit — an invalid law simply is not law, for anyone (Department of Justice Canada, 2025a). A declaration that a federal statute is invalid therefore has inherently national consequences, in the same way that American APA vacatur removes a rule from the books. In this sense, Canada’s ordinary, uncontroversial, everyday remedy for unconstitutional national legislation is more sweeping than the remedy the U.S. Supreme Court abolished in CASA.

Canadian courts also possess the full toolkit surrounding that core remedy: interlocutory (preliminary) injunctions and stays, including against the enforcement of legislation pending trial; permanent injunctions; personal remedies under section 24(1) of the Charter, which authorizes “such remedy as the court considers appropriate and just in the circumstances,” from damages to structural orders (Department of Justice Canada, 2025b); judicial review in the Federal Court, which has statutory power to quash decisions of federal boards and officials with effect across the country (Federal Courts Act, 1985); class actions, including multi-provincial and national classes (Western Canadian Shopping Centres Inc. v. Dutton, 2001); and even injunctions with global reach against private parties (Google Inc. v. Equustek Solutions Inc., 2017). The ingredients for an American-style controversy all exist. The next two sections examine why they combine so differently — and the answer, in brief, is that Canada makes broad remedies harder to get quickly, easier to review, and subject to explicit legislative response mechanisms, while making them entirely ordinary at the end of a case rather than the beginning.

The Closest Canadian Equivalents

No single Canadian remedy maps perfectly onto the American national injunction, but several come close along different dimensions.

Declarations of invalidity under section 52(1)

The declaration of invalidity is the closest functional equivalent in effect: it neutralizes a national law for everyone. It differs from the American injunction in three fundamental ways. First, it is typically a final remedy issued after a full trial on a complete evidentiary record — Canadian Charter litigation is famously fact-heavy — rather than a preliminary order issued weeks into a case. Second, it is a declaration, not a coercive order: Canadian constitutionalism relies on the strong convention that governments obey declarations, so there is no contempt machinery and no supervising judge. Third, its formal reach is disciplined by precedent rather than decree. In R. v. Sullivan (2022), the Supreme Court of Canada clarified that when a superior court declares a law unconstitutional under section 52(1), that ruling is not a formal erasure binding all other courts nationwide; it is a judicial precedent, binding within the province under ordinary stare decisis principles and persuasive elsewhere, until an appellate court settles the question (R. v. Sullivan, 2022). Sullivan is, in effect, Canada’s answer to the “one judge rules the nation” problem: a single trial judge’s constitutional ruling matters, but it propagates through the doctrine of precedent — with all its built-in appellate correction — rather than through the scope of an injunction.

Suspended declarations of invalidity

Canada’s most distinctive contribution to comparative remedies law runs in the opposite direction from the national injunction: instead of a court stopping a law before it is found unlawful, a Canadian court can find a law unconstitutional and yet keep it temporarily in force. The suspended declaration of invalidity, born in Re Manitoba Language Rights (1985) — where striking down ninety years of unilingual Manitoba statutes immediately would have created a legal vacuum — allows the court to delay the effect of its declaration, typically for six to eighteen months, so the legislature can craft a constitutional replacement (Bird, 2019). In Ontario (Attorney General) v. G (2020), the Supreme Court restated the framework: declarations should presumptively take immediate effect, and the government bears the burden of demonstrating that a compelling public interest, grounded in the Constitution, justifies a suspension (Ontario (Attorney General) v. G, 2020; Department of Justice Canada, 2025a). The suspended declaration is the mirror image of the universal injunction — a remedy designed to protect democratic space rather than to occupy it — and, as discussed below, it has attracted a mirror-image controversy.

Interlocutory injunctions and stays against legislation

The closest procedural equivalent — a preliminary court order pausing government policy pending trial — exists in Canada but is deliberately hard to obtain. The governing test from RJR-MacDonald Inc. v. Canada (Attorney General) (1994) asks whether there is a serious question to be tried, whether the applicant will suffer irreparable harm, and where the balance of convenience lies — and at the third stage, Canadian courts apply a heavy thumb on the scale for government: legislation duly enacted by a democratic legislature is presumed to serve the public interest, and suspending it pending trial is an extraordinary step (RJR-MacDonald Inc. v. Canada (Attorney General), 1994; Manitoba (Attorney General) v. Metropolitan Stores Ltd., 1987). The Supreme Court reinforced the point in Harper v. Canada (Attorney General) (2000), staying a lower court injunction against federal election spending limits and stressing that courts should not lightly deprive the public of the benefit of duly enacted legislation on an interlocutory basis (Harper v. Canada (Attorney General), 2000). The doctrine also distinguishes “exemption” cases (suspending a law only for the applicant) from “suspension” cases (suspending it for everyone), treating the latter as far more drastic — a distinction that is, in substance, the American party-specific versus universal debate, resolved decades earlier in favour of restraint (Manitoba (Attorney General) v. Metropolitan Stores Ltd., 1987).

Federal Court remedies, class actions, and global injunctions

Three further instruments complete the picture. First, the Federal Courts Act channels judicial review of federal boards, commissions, and officials into a single national court, which may quash or set aside federal administrative action — an order that, because the decision-maker is national, has automatically national effect, much like APA vacatur (Federal Courts Act, 1985, ss. 18, 18.1). Concentrating this jurisdiction in one specialized court, whose judgments are reviewed by one Federal Court of Appeal, structurally eliminates the inter-district races that drive American forum shopping. Second, Canadian class actions, recognized at common law in Western Canadian Shopping Centres Inc. v. Dutton (2001) and governed by provincial and federal legislation, permit national classes; but because a section 52 declaration already protects everyone, constitutional litigants rarely need class certification for injunctive purposes — Canadian class actions centre on damages (Western Canadian Shopping Centres Inc. v. Dutton, 2001). Third, Canadian courts have shown striking comfort with broad injunctions in private law: in Google Inc. v. Equustek Solutions Inc. (2017), the Supreme Court upheld an interlocutory injunction ordering Google — a non-party to the underlying trade secrets dispute — to de-index a defendant’s websites worldwide (Google Inc. v. Equustek Solutions Inc., 2017). Equustek proves the point that Canadian judicial restraint about enjoining legislation is a considered stance about democratic institutions, not a general timidity about remedial scope.

Which equivalent is closest overall? For effect, the section 52 declaration of invalidity; for procedure, the interlocutory suspension of legislation under RJR-MacDonald; for structure, Federal Court judicial review. The instructive point is that Canada disaggregates what the American national injunction bundles together: national effect (routine, but final and precedent-based), preliminary timing (available, but exceptional), and coercive universal scope (essentially absent against legislation).

Are Broad Remedies Controversial in Canada?

Yes — but the controversy has a different shape. Canada spent the 1990s and 2000s conducting a vigorous national argument about judicial power under the Charter, with critics charging that an activist Supreme Court and a “Court Party” of interest-group litigants were transferring policy questions from legislatures to judges, and defenders responding that Charter review is a form of institutional “dialogue”: courts identify constitutional defects, and legislatures respond with new legislation, invoking section 1’s reasonable-limits justification or, in the extreme, section 33’s notwithstanding clause (Hogg & Bushell, 1997). That debate is recognizably the cousin of the American democratic-legitimacy argument. But it has focused on whether courts should strike down laws at all — the merits of judicial review — rather than on the American question of remedial scope and timing. Nobody in Canada seriously argues that a declaration of federal invalidity should protect only the claimant; the in rem logic of section 52 makes that position almost unintelligible.

Where Canadian remedial controversy does exist, it clusters in three places. The first is the suspended declaration. Scholars, including Kent Roach and the University of Toronto’s Asper Centre, argued for years that Canadian courts had drifted into granting suspensions routinely and without justification — leaving claimants who had proven their rights were violated subject to the unconstitutional law for another year or more, a practice Brian Bird memorably called a “judicial notwithstanding clause” (Bird, 2019; University of Toronto Faculty of Law, 2020). The Supreme Court substantially accepted the critique in G, demanding principled justification and preserving individual exemptions for successful claimants (Ontario (Attorney General) v. G, 2020). Note the inversion: America’s remedial fight is about courts doing too much, too early; Canada’s has been about courts doing too little, too late, for the very people who won.

The second cluster concerns interlocutory injunctions in resource and Indigenous rights disputes — pipelines, logging, blockades — where courts issuing (or refusing) injunctions effectively allocate enormous economic and constitutional stakes before trial. This controversy is real and heated, but it concerns private-versus-public interest balancing and reconciliation with Indigenous peoples, not the scope of relief against national legislation. The third is Equustek-style extraterritoriality: ordering a global platform to act worldwide raised alarms about comity and free expression, and a U.S. federal court promptly declared the Canadian order unenforceable in the United States (Google Inc. v. Equustek Solutions Inc., 2017). Canada, in other words, has exported the very complaint Americans make domestically — one judge, everyone bound — to the international plane.

Is abuse structurally more or less likely in Canada? Less, for identifiable reasons. Forum shopping yields little: superior courts in every province apply the same national Charter jurisprudence, a single Supreme Court sits atop every appellate route and hears constitutional appeals relatively quickly, and Sullivan confines a lone trial ruling’s formal effect. Interim suspension of legislation is doctrinally disfavoured, so the preliminary-freeze dynamic barely exists. Mandatory notice to attorneys general of constitutional questions ensures governments are heard before laws are struck. And the legislature holds explicit response tools — section 1 justification, legislative sequels, and section 33 — that lower the temperature of any single judicial ruling because it is rarely the last word (Hogg & Bushell, 1997). The cost of this design is the mirror-image risk: a rights claimant facing an unconstitutional national law may wait years for a final declaration, and then wait again through a suspension. Canada trades speed of protection for stability of process.

Canadian Case Studies

Re Manitoba Language Rights (1985): inventing the suspended declaration

When the Supreme Court held that Manitoba’s failure to enact its laws in both English and French rendered ninety years of statutes invalid, an immediate declaration would have dissolved the province’s entire legal order. The Court declared the laws invalid but deemed them temporarily effective while Manitoba translated and re-enacted them, expressly invoking the rule of law to avoid “anarchy” (Bird, 2019). The case shows Canadian courts at their most creative — and most cautious: maximal constitutional principle, delivered with a remedy engineered to preserve governance. It is the founding precedent for the idea that how and when a broad remedy takes effect is itself a matter of constitutional judgment.

Schachter v. Canada (1992) and Carter v. Canada (2015): calibrating breadth

Schachter established the menu of section 52 remedies — striking down, striking down with suspension, severance, reading down, and reading in — and insisted the remedy match the extent of the inconsistency (Schachter v. Canada, 1992). Carter, which struck down the criminal prohibition on physician-assisted dying, illustrates the machinery under stress: the Court suspended its declaration for twelve months, later granted a four-month extension when Parliament could not legislate in time, and permitted individual exemptions during the extension so that suffering claimants were not held hostage to the legislative timetable (Carter v. Canada (Attorney General), 2015; Department of Justice Canada, 2025a). Carter is Canada’s clearest demonstration that nationwide invalidation of a criminal law can be managed as a structured, dialogic process — and its extension saga is the standard citation for critics who say suspensions let governments dawdle while rights violations continue.

Ontario (Attorney General) v. G (2020): disciplining the suspension

G unanimously held that Ontario’s sex offender registry discriminated against persons found not criminally responsible on account of mental disorder, contrary to section 15 of the Charter. On remedy, the majority restated first principles: declarations presumptively take immediate effect; the government must demonstrate a compelling interest to justify suspension; and successful claimants should ordinarily be exempted from any suspension, because they “help the public by uncovering unconstitutional laws” (Ontario (Attorney General) v. G, 2020; Supreme Court of Canada, 2020). G is the Canadian system consciously correcting a remedial drift — the same kind of self-correction CASA claimed to perform, but aimed at protecting claimants from judicial under-reach rather than protecting government from judicial overreach.

R. v. Sullivan (2022): one judge does not rule the nation

Two Ontario accused invoked a prior Ontario ruling that had declared a Criminal Code provision unconstitutional, arguing the provision was already a dead letter. The Supreme Court disagreed with the strongest version of that claim: section 52 declarations by superior courts are not formally binding erga omnes across the country; they operate through stare decisis, binding courts in the same province absent exceptional circumstances and persuading courts elsewhere (R. v. Sullivan, 2022). Sullivan directly engages the CASA problem — what happens when one trial judge declares a national law invalid — and resolves it with precedent doctrine rather than remedial surgery. The federal law’s fate is settled definitively only as the question ascends the unified appellate ladder, which in Canada it reliably and relatively rapidly does.

RJR-MacDonald (1994) and Harper (2000): the interim restraint tradition

Tobacco companies in RJR-MacDonald sought to stay federal plain-packaging regulations pending their Charter challenge; the Court articulated the tripartite interlocutory test and denied relief, emphasizing the public interest in enforcing duly enacted law (RJR-MacDonald Inc. v. Canada (Attorney General), 1994). In Harper, an Alberta judge had enjoined federal third-party election spending limits weeks before a possible election; the Supreme Court stayed the injunction, warning that interlocutory suspension of legislation deprives the public of measures the legislature deemed beneficial before any final finding of invalidity (Harper v. Canada (Attorney General), 2000). These two cases are the Canadian analogue of the entire American preliminary injunction controversy, decided in two paragraphs of principle: assume validity, preserve the law, and make claimants win at trial. An American reader will notice what this restraint costs — RJR-MacDonald-style deference would have left Executive Order 14160 operative during litigation had it been Canadian law — and what it purchases: no forum derby, no shadow docket, no government by preliminary order.

Google v. Equustek (2017): breadth without borders

Finally, Equustek rounds out the portrait. A small British Columbia technology company obtained an interlocutory injunction requiring Google to de-index a former distributor’s websites from search results worldwide. The Supreme Court upheld the order: where the internet is the instrument of harm, only a global order is effective (Google Inc. v. Equustek Solutions Inc., 2017). The case demonstrates that Canadian courts will issue extraordinarily broad injunctions — against non-parties, with worldwide effect, at the interlocutory stage — when the logic of effective relief demands it. The restraint Canadian courts show toward legislation is therefore best read not as remedial modesty but as a theory of democratic institutions: private wrongdoing gets whatever scope effectiveness requires; democratic enactments get the benefit of the doubt until final judgment.

Comparing the United States and Canada

The comparison can now be drawn systematically. The two systems share the fundamentals — federalism, constitutional supremacy, judicial review, independent courts — yet they distribute remedial power in nearly opposite patterns.

Table 3

Broad judicial remedies in the United States and Canada

Dimension United States Canada

Core remedy against an unlawful national law Injunction (party-centred equity); APA vacatur for agency rules Declaration of invalidity under s. 52(1): the law itself is of no force or effect

Protection of non-parties Contested; after CASA requires class action, complete-relief showing, or vacatur Automatic in substance — an invalid law binds no one; formal effect spreads through precedent (Sullivan)

Preliminary freeze of national policy Historically routine; now channelled through class-wide preliminary injunctions and stays Exceptional; strong presumption that duly enacted laws serve the public interest (RJR-MacDonald; Harper)

Forum shopping incentive High: 94 districts, single-judge divisions, national payoff for one win Low: uniform national jurisprudence, limited first-instance payoff, unified appeals

First-instance forums 94 federal district courts Provincial superior courts in every province; Federal Court for federal administrative action

Appellate unification 13 circuits; discretionary Supreme Court review; strained emergency docket Provincial courts of appeal and Federal Court of Appeal; one Supreme Court; advisory reference procedure

Timing of broad relief Front-loaded: most force at the preliminary stage Back-loaded: full force at final judgment, sometimes further suspended

Legislative response tools New legislation; constitutional amendment; venue and jurisdiction bills (e.g., NORRA) Legislative sequels under s. 1; notwithstanding clause (s. 33); ordinary amendment

Softening devices Stays pending appeal; administrative stays Suspended declarations of invalidity; individual exemptions; transition periods

Signature controversy One trial judge provisionally freezing national policy for everyone Courts keeping unconstitutional laws temporarily alive; legislatures overriding courts via s. 33

Note. Compiled from sources cited in text, including Trump v. CASA, Inc. (2025), R. v. Sullivan (2022), RJR-MacDonald (1994), and Department of Justice Canada (2025a).

Several differences deserve emphasis beyond the table. Court structure does the quietest but heaviest work. The United States distributes first-instance federal judicial power across hundreds of judges whose interim orders can have national effect, and unifies the law only at the top, slowly. Canada distributes first-instance power too — provincial superior courts everywhere can hear Charter challenges — but denies interim orders national coercive effect (Sullivan plus RJR-MacDonald), channels review of federal administration into one Federal Court system, and unifies constitutional law through a Supreme Court that sits as a general court of appeal and can hear references — advisory opinions requested by governments — that settle contested questions before or without litigation. The reference power is a genuinely foreign concept to American law and functions as a pressure-release valve: many disputes that in the United States would be fought through injunction races (the constitutionality of carbon pricing, for example, resolved in References re Greenhouse Gas Pollution Pricing Act (2021)) reach Canada’s top court directly as clean legal questions.

Constitutional text also matters. Section 52 makes invalidity the automatic consequence of inconsistency, giving Canada a law-centred remedy that needs no injunction; American courts, lacking an express supremacy remedy running against federal law, built their practice on party-centred equity — which is precisely why the scope of the party-centred remedy became the battlefield. Meanwhile, section 33 of the Charter gives Canadian legislatures a five-year override of most judicial rulings under the Charter — a democratic trump card with no American parallel. Whatever one thinks of the notwithstanding clause (its increasing provincial use is Canada’s own live legitimacy controversy), its existence changes the stakes of judicial review: a Canadian court that strikes down a law knows the legislature retains a lawful path to disagree, which arguably licenses both bolder rights rulings and calmer political reactions (Hogg & Bushell, 1997).

Finally, judicial culture and selection differ. American federal judges are appointed through an openly partisan process and are publicly sorted by appointing president; litigants act accordingly. Canadian judicial appointments, while hardly apolitical, do not generate reliable ideological sorting at the trial level, and Canadian legal culture treats the attorney general’s presence, intervener participation, and appellate correction as the norm in constitutional cases. The American national injunction debate is, in significant part, a debate about what happens when remedial doctrine collides with a judiciary the public perceives as politically legible. Canada’s calmer remedial politics reflect, in part, a judiciary that is harder to shop.

The Democratic Legitimacy Question

Beneath the doctrinal machinery lies the oldest question in constitutional theory: by what right do unelected judges stop the projects of elected governments? Alexander Bickel famously named the problem the “counter-majoritarian difficulty” (Bickel, 1986), and the national injunction debate is that difficulty in its most concentrated modern form — because the injunction does not merely invalidate a democratic choice after deliberation, it can suspend one immediately, provisionally, and at the instance of a single strategically selected judge.

The rule-of-law answer is that this framing gets the legitimacy question backwards. Constitutional democracy is not majoritarianism; it is majoritarianism within limits that majorities themselves adopted. When a court blocks an unlawful policy, it is not defeating democracy but enforcing democracy’s own higher-order commitments — the Fourteenth Amendment in Barbara, the Charter in Carter. On this view, the relevant comparison is not “one judge versus millions of voters” but “the Constitution versus a temporary official acting outside it,” and a remedy that leaves the government free to keep violating the Constitution against everyone who has not sued is not modest — it is a subsidy for lawbreaking at scale. Justice Jackson’s CASA dissent put the point sharply: a regime in which the executive must obey the law only as to those with the resources to litigate creates zones of lawlessness organized by wealth and access (Trump v. CASA, Inc., 2025).

The democratic-accountability answer replies that this account assumes what litigation is supposed to determine. At the preliminary stage, no one yet knows whether the policy is unlawful; the injunction rests on one judge’s prediction. If the prediction is wrong — as it was, on the Supreme Court’s eventual view, in the travel ban litigation — then the “rule of law” remedy has in fact nullified lawful democratic action for years. And prediction errors are not randomly distributed: forum shopping ensures the predicting judge is systematically the one most likely to rule against the government. There is also an accountability asymmetry. When a legislature errs, voters can remove it; when a president errs, elections and impeachment loom; when a district judge errs about national policy, the correction runs through years of appeals during which the error governs. Judicial review’s classic defence — that it is deliberate, reasoned, final, and rare — describes appellate merits decisions like Barbara. It describes a 2 a.m. temporary restraining order rather less well.

The most useful synthesis treats legitimacy as a function of process quality, not of judicial power as such. Broad judicial interference with government is most defensible when four conditions hold: the decision follows genuine adversarial testing on a developed record; the deciding court was not chosen by the winner; the decision is final or rapidly reviewable by a court that speaks for the whole system; and the political branches retain some lawful avenue of response. Measured against those conditions, the American preliminary universal injunction scored poorly on all four — which is why even many scholars sympathetic to strong judicial review welcomed discipline — while the final, appellately settled invalidation of a national law scores well, which is why almost no one in either country contests that power. Canada’s arrangements, whatever their other costs, are essentially an institutionalization of the four conditions: restraint before final judgment, national effect through precedent and appellate settlement, and section 33 and legislative sequels as the political response valve. The lesson is not that courts should be weak. It is that the legitimacy of a broad remedy is earned by the procedure that produces it.

Developing a Balanced Position

Are national injunctions necessary? Sometimes, genuinely, yes — in the narrow but important class of cases where rights are indivisible, harms are irreversible, and affected people cannot realistically litigate for themselves. The birthright citizenship episode is close to a controlled experiment: had relief been strictly party-limited between January 2025 and June 2026, an executive order that five justices ultimately held unconstitutional would have operated against every child whose parents had not sued, creating a cohort of documented-stateless newborns whose injuries no final judgment could fully repair. Any account of remedies that accepts that outcome as the cost of proceduralism has, it seems fair to say, mistaken the means of law for its ends.

When are they dangerous? When they are cheap. The pre-CASA equilibrium made the most powerful remedy in the system also the easiest to obtain: available from any of hundreds of judges, at the preliminary stage, on an undeveloped record, in a forum chosen precisely for its predicted sympathy, with asymmetric stakes that let challengers lose repeatedly and win once. That combination did identifiable damage — to percolation, to the Supreme Court’s docket, to public confidence in judicial neutrality — and the damage was ecumenical, striking Democratic and Republican governments alike. The strongest single argument against the universal injunction was never that non-parties must be left unprotected; it was that a system offering universal relief on those terms invites every losing political coalition to relitigate elections in trial courts.

What safeguards should exist? The analysis above suggests broad relief should be conditioned on procedural weight commensurate with its scope: class certification or an equivalent showing that the protected group is genuinely uniform; random or centralized venue for suits seeking to halt national policy; explicit findings on why narrower relief fails; and guaranteed expedited appellate review so that no single judge’s view governs the nation for long. Post-CASA American law is groping toward exactly this settlement — universal relief survives, but must now pass through the tollbooths of Rule 23, complete-relief findings, or APA vacatur — and Barbara demonstrated that the tolled road can still deliver nationwide protection within weeks when the case warrants it. The unresolved risk, flagged by Justice Alito, is that courts will wave traffic through the tollbooths, reproducing the old problem under new names; the countervailing risk, flagged by Justice Sotomayor, is that the tolls will price out precisely the plaintiffs who most need protection. Which risk materializes will be the story of the next decade of American public law litigation.

Does Canada offer a better model, a different model, or a less visible version of the same problem? Honest answer: a different model with real advantages and an unadvertised price, plus a small dose of the same problem in different clothes. Canada’s advantages are structural and enviable from an American vantage: no meaningful forum shopping, no preliminary freezes of national legislation, a unified and comparatively fast appellate settlement mechanism, and remedies aimed at laws rather than tug-of-wars over non-parties. Its price is temporal: claimants can live under unconstitutional laws for years awaiting final judgment, and then longer under suspended declarations — a cost borne by the very people whose rights the system exists to protect, and one Canadian doctrine has itself recognized as excessive and begun correcting in G. And Canada is not immune to the underlying dynamic: suspended declarations were criticized as routinized judicial policymaking, Equustek exported the one-judge-binds-everyone problem to the globe, and the growing casual use of the notwithstanding clause shows that reducing friction between courts and legislatures does not eliminate contests over who decides — it relocates them. The systems are best seen as two calibrations of the same dial: the United States historically set it toward immediate protection at the cost of process, Canada toward process at the cost of immediate protection, and each has recently reached toward the other’s setting — CASA adding process to American protection, G adding protection to Canadian process.

Conclusion

This essay began with a single executive order and the judges who stopped it. Eighteen months later, that story had produced two landmark Supreme Court decisions that answer, between them, the two questions the national injunction debate always entangled: who may block a national policy while its legality is contested, and whether this national policy was legal. CASA answered the first with restraint — trial courts may no longer protect the world by default. Barbara answered the second with force — the policy violated the Constitution, and a nationwide class of children was protected throughout. The sequence is a compact demonstration of the essay’s central tension: courts need tools strong enough that unlawful government action does not simply win by default and by delay, yet tools so strong, so fast, and so easily obtained that a single judge can govern the country are corrosive in a different way, converting courts into the continuation of elections by other means.

The Canadian comparison earns its place because it decomposes that tension into design choices. Canada shows that a legal system can invalidate national laws for everyone without any doctrine of universal injunctions; that interim restraint toward legislation is compatible with muscular final remedies; that appellate structure and precedent doctrine can do the uniformity work Americans ask of remedial scope; and that legislatures can be given explicit, lawful ways to answer courts, lowering the temperature of every individual ruling. It equally shows the costs: slower protection, suspended rights, and its own remedial controversies inverted. Neither country has solved the problem, because the problem — how much power any one institution, or any one official, should wield over everyone at once — is not solvable, only managed.

The forward-looking question, for both democracies, is therefore not whether courts should have broad remedies but how expensive broad remedies should be, and who should pay the price of error while the legal system makes up its mind. A democracy that makes sweeping relief too cheap invites government by the fastest litigant and the friendliest judge. A democracy that makes it too dear tells its most vulnerable members that their rights will be vindicated eventually, which can be another way of saying too late. Designing remedies that are strong enough to protect people and disciplined enough to deserve obedience — remedies whose authority is earned by the process that produces them — is not a technicality of civil procedure. It is, in the end, what separates the rule of law from the rule of whoever got to the courthouse first.