Why Immigration Law in the Supreme Court Works Differently

Immigration law in the Supreme Court is governed by a doctrine that has no real counterpart anywhere else in American constitutional law. In every other field this series covers, Congress legislates under an enumerated power, the Court measures the statute against constitutional limits, and the ordinary tools of judicial review do the work. In immigration, the Court has held since 1889 that the power to decide who may enter and who must leave belongs to the nation as an attribute of sovereignty itself, not to any clause of the Constitution, and that the political branches exercise that power with a latitude courts in no other field would tolerate. The question this article answers is what that latitude has actually produced: which decisions built it, where its edges were drawn, and what the governing holdings say.

The operative question is deceptively simple: why do courts defer to Congress on immigration when they defer on almost nothing else to the same degree? A litigant challenging an environmental rule or a tax provision can expect the court to parse text, weigh precedent, and test the government’s reasoning against constitutional boundaries. A litigant challenging an exclusion or deportation decision enters a courtroom where the controlling precedent says the political branches hold a power “largely immune from judicial control,” a phrase the Court used without apology in the middle of the twentieth century. Understanding immigration litigation means understanding why that sentence is possible, how far it reaches, and where later decisions carved exceptions into it without ever repudiating the core.

The problem Congress legislated against was not abstract. Through the nineteenth century, the United States had no comprehensive federal immigration code and no constitutional clause assigning the subject to Congress. The states regulated entry at their ports under their police powers, treaties with foreign nations set terms for the movement of their subjects, and the first federal interventions were narrow: the Page Act of 1875, the Chinese Exclusion Act of 1882, and a series of statutes that tested, one by one, whether the national government could close the door it had never been expressly given the key to. When those statutes reached the Supreme Court, the justices faced a genuine constitutional puzzle. The power to exclude aliens appeared nowhere in Article I. The Commerce Clause covered commerce with foreign nations, but human beings were an awkward fit for a commercial category. The Naturalization Clause governed citizenship, not admission. The Court had to locate the power somewhere, and the place it chose, national sovereignty itself, shaped everything that followed.

The litigation-as-pricing metaphor that organizes this series takes its sharpest form here. In most fields, litigation prices a statute’s breadth by testing its provisions against constitutional limits, and the resulting case law tells regulated parties what the statute can actually do. The price is set through adversarial testing, and it is usually substantial: environmental, labor, and civil rights statutes have all been narrowed, channeled, or fortified through decades of such testing. In immigration, the price is lowest, because the plenary power doctrine discounts most challenges before they begin. The political branches can legislate and enforce at the border with an assurance of judicial deference that no other field affords. That does not mean the statute goes untested. It means the testing concentrates where the doctrine permits it: on state action, on agency fidelity to text, on procedural regularity, on the constitutional boundaries the doctrine never swallowed. The case law of immigration is therefore a peculiar kind of price list, long on the margins and short at the center, and reading it requires knowing which shelf each decision sits on.

The decisions this article examines construe the Immigration and Nationality Act and the constitutional framework surrounding it. That framework is the true subject here, because in no other area of legislation does the constitutional background matter so much to the meaning of the statute. A reader who finishes this article should be able to explain the doctrine that makes immigration law constitutionally unlike every other field in this series: that Congress and the executive exercise a power the courts have treated since the nineteenth century as largely immune from ordinary constitutional review. That reader should also be able to trace how the doctrine was softened at the edges without being abandoned, and to state the holdings of the decisions that actually govern. That is the One Test this article is built to pass, and the arc that follows is organized to make it passable.

The statutory identity of this article is decisions construing the Immigration and Nationality Act and the constitutional framework surrounding it, and the relationship between the two deserves a final word before the arc begins. The Act supplies the provisions the Court interprets: the asylum and withholding standards, the detention authorities, the relief provisions, the removal machinery. The constitutional framework supplies the presumptions the Court brings to the interpretation: plenary power favoring the political branches, federalism limiting the states, due process and habeas constraining the margins. Neither can be understood without the other. A reader who knows the statute but not the framework will misread every close case, expecting text to decide what doctrine has already discounted. A reader who knows the framework but not the statute will mistake deference for lawlessness, missing the interpretive discipline the Court has imposed within the doctrine’s bounds. This article holds both in view at once, because the decisions themselves do.

The arc runs from the foundation to the present in the order the doctrine developed. It begins with the two nineteenth century decisions that located the immigration power in sovereignty rather than in an enumerated clause, and with the mid-century exclusion cases that extended deference to the executive at the border. It turns to the softening: the 1982 decision that barred states, though not the federal government, from denying public schooling to children present unlawfully. It takes up the structural case that began as a single deportation suspension and ended by invalidating the legislative veto across the entire United States Code. It works through the statutory interpretation decisions on asylum standards, habeas review, and retroactivity. It examines the detention pair, one reading a limit into post-order custody to avoid a constitutional problem and the other sustaining mandatory detention during proceedings. It maps the federalism decision that drew the preemption line for state enforcement statutes. And it closes with the two decisions most often invoked in current argument, one sustaining an entry suspension under a broad statutory delegation and the other resolving a rescission dispute on administrative procedure rather than on the policy’s legality, a distinction that matters enormously and is widely misunderstood.

Two cautions before the arc begins. First, several of these decisions are political symbols, and symbols invite readers to import meanings the opinions do not contain. This article states each holding from the opinions, presents majority and dissent with equal care, and describes the administrative-procedure ground of the 2020 decision precisely rather than as a ruling on the merits. Second, the field rewards precision about who is being constrained. Some of these decisions constrain Congress, some constrain the executive, and some constrain the states while leaving federal power untouched. Confusing those targets is the most common error in public discussion of this subject, and the structure below is designed to keep them distinct.

Each section below states the holding first, then the reasoning, then the dissent at its strongest, then the doctrinal effect. That order is deliberate. The holdings are what govern; the reasoning is what predicts; the dissents are what the next generation of litigants will try to revive; and the effect is what the ledger records. A reader who can reproduce that four-part structure for each decision, without notes, has mastered the field.

The constitutional vacuum the Court confronted deserves emphasis, because it explains why the doctrine took the shape it did. The framers wrote no immigration clause. For the first century of the republic, the movement of people was governed by a patchwork: states exercised police powers over arriving passengers, Congress passed narrow statutes aimed at particular problems, and treaties set terms with individual nations. When Congress began to legislate comprehensively in the late nineteenth century, first against Chinese laborers and then more broadly, there was no obvious textual home for the new federal power. The Court could have stretched the Commerce Clause to cover human migration, but that would have imported the clause’s limits and its doctrinal baggage. It could have treated admission as an aspect of foreign affairs, but that would have raised questions about the treaty power’s role. Instead the Court reached outside the enumerated powers entirely, to sovereignty itself, and in doing so created a power with no internal textual limit. That choice is the reason the field’s deference has no parallel: every other congressional power in this series comes with a clause that defines it, and every clause invites judicial interpretation of its boundaries. A power sourced in sovereignty invites no such boundary-drawing, because sovereignty is the thing boundaries are drawn around, not within.

The litigant’s perspective makes the abstraction concrete. Imagine two plaintiffs with identical complaints about fairness, one challenging a tax regulation and one challenging a deportation order. The tax plaintiff enters a courtroom where the judge will examine the statute, the regulation, the administrative record, and the constitutional limits on the taxing power, applying tools refined over two centuries. The immigration plaintiff enters a courtroom where the judge begins from the premise that the political branches hold plenary authority over the subject, and where the question is not whether the decision was wise or even fair but whether the political branches acted within a domain the Court has largely committed to them. The difference is not one of degree. It is a different constitutional universe, and this article maps it decision by decision.

The Nineteenth Century Foundation: Sovereignty Instead of an Enumerated Clause

The doctrine began with a laborer, a treaty, and a certificate. Chae Chan Ping had lived and worked in the United States before departing for a visit to China, carrying the reentry certificate that federal law then required of Chinese laborers who left the country. While he was abroad, Congress passed the Scott Act of 1888, which voided outstanding reentry certificates and barred the return of Chinese laborers who had departed. When Chae Chan Ping arrived at San Francisco and presented his certificate, he was refused admission. He challenged his exclusion through habeas corpus, arguing that the 1888 statute could not override the Burlingame Treaty of 1868 between the United States and China, under which Chinese subjects had been promised the right of free migration.

Justice Stephen Field, writing for a unanimous Court on May 13, 1889, rejected the challenge in Chae Chan Ping v. United States, 130 U.S. 581 (1889). The opinion located the power to exclude aliens not in any enumerated grant but in the sovereignty of the nation itself. Exclusion of foreigners, Field wrote, was an incident of sovereignty belonging to the government of the United States as part of the sovereign powers delegated by the Constitution. The reasoning had two steps with lasting consequences. First, the power to determine who may enter the territory is inherent in every sovereign nation and needs no textual hook beyond nationhood. Second, because treaties and statutes stand on equal footing as the supreme law of the land, a later statute can override an earlier treaty: the last-in-time rule. The Scott Act, enacted after the Burlingame Treaty, controlled, and the certificate Chae Chan Ping carried could not defeat it.

The unanimity of the 1889 ruling carried a signal of its own. A divided Court might have suggested that the sovereignty rationale was one plausible theory among others, open to revision as membership changed. A unanimous Court, speaking through Justice Field with no recorded dissent, presented the rationale as the law’s settled foundation, and later justices treated it that way. When the Court extended the doctrine to expulsion four years later, the six to three division concerned the application, not the premise; no justice in the Fong Yue Ting majority or dissent questioned that exclusion itself belonged to the sovereign. The premise had become common ground within a single presidential term, and common ground is the hardest kind of precedent to dislodge.

Four years later the Court extended the logic from exclusion to expulsion. The Geary Act of 1892 required Chinese laborers already present in the United States to register and carry certificates of residence, with deportation as the penalty for failure to comply, and it imposed a notorious evidentiary rule requiring proof of lawful residence by “at least one credible white witness.” Three laborers who refused to register challenged the statute. In Fong Yue Ting v. United States, 149 U.S. 698 (1893), decided May 15, 1893, Justice Horace Gray wrote for a 6 to 3 majority that the power to expel was “as absolute and unqualified” as the power to exclude. The right to expel aliens, the opinion held, rested on the same sovereign foundation as the right to exclude them, and Congress could entrust its exercise to executive officers with only the procedures Congress chose to provide. The Geary Act was upheld.

The dissents in Fong Yue Ting deserve attention because they stated the position the majority would reject for more than a century. Three justices argued that persons lawfully present in the United States were entitled to the protections of due process before being expelled, and that deportation without judicial trial offended constitutional guarantees. The majority’s answer was that deportation was not punishment in the constitutional sense but an exercise of the sovereign power to determine membership in the national community. That characterization, that removal is civil rather than criminal, administrative rather than punitive, became one of the load-bearing walls of the entire field. It explains why the full apparatus of criminal procedure has never attached to deportation proceedings, and why the Court could treat exclusion and expulsion as political acts committed to the political branches.

The two decisions together established what lawyers call the plenary power doctrine: the proposition that the political branches possess complete, or plenary, authority over immigration, subject to only the thinnest judicial review. The word plenary does not appear as a term of art in the Constitution. It is the Court’s own label for a power it considers so deeply tied to sovereignty, foreign relations, and national self-definition that judges lack both the competence and the constitutional warrant to second-guess its exercise. In practice the doctrine meant that when Congress or the executive made a decision about admission or removal, the Court would ask only whether the political branches had acted, not whether they had acted wisely, fairly, or even consistently.

Why could a later statute override a treaty in Chae Chan Ping?

Because the Constitution places treaties and federal statutes on the same legal plane, the later enactment controls when the two conflict. Field’s opinion added that exclusion was an incident of sovereignty itself, so Congress could revise the nation’s treaty commitments by statute. The holding made immigration policy a matter for current majorities rather than for past diplomats.

The statutory framework these decisions construe, in its modern form, is the Immigration and Nationality Act as reorganized by the 1965 amendments, whose text, categories, and ceilings are examined in the companion guide to the Immigration and Nationality Act of 1965. The nineteenth century cases predate that codification, but the doctrine they announced governs the reading of every provision it contains. When the modern Court interprets a detention provision or a relief provision of the Act, it does so against the background of Chae Chan Ping and Fong Yue Ting, and the background often decides the case before the text is even parsed. That is what makes the foundation worth lingering over: it is not history for its own sake but the operating assumption of the present law.

The treaty at the center of Chae Chan Ping repays attention because it shows what the sovereignty rationale displaced. The Burlingame Treaty of 1868 had promised the subjects of China and the citizens of the United States reciprocal rights of free migration and residence, a commitment the United States made when it wanted Chinese labor for railroad construction and Pacific trade. Within two decades, anti-Chinese politics in California and the western states had made the treaty a dead letter in practice, and Congress responded with escalating statutes: the Page Act of 1875, the Chinese Exclusion Act of 1882, and successive amendments tightening the restrictions. The Scott Act of 1888 was the breaking point. It did not merely restrict future migration; it voided the reentry certificates of laborers already abroad, converting lawful departures into permanent exile. Chae Chan Ping, holding such a certificate, became the test of whether Congress could do that to a man traveling under a treaty promise.

Field’s opinion answered with two moves that each did independent work. The first was the sovereignty source: the power to exclude foreigners, “as an incident of sovereignty,” belonged to the United States government as part of the sovereign powers delegated by the Constitution, and it could not be surrendered or bargained away by treaty.

The certificate system that the Scott Act destroyed deserves explanation, because it shows how the exclusion regime operated in practice. Under the 1882 Act and its amendments, Chinese laborers already in the United States who wished to travel abroad could obtain a reentry certificate documenting their prior lawful residence, which entitled them to return. The certificate was, in substance, a government promise: leave with this document and you may come back. Thousands of laborers relied on it, planning voyages home around its assurance. The Scott Act voided every outstanding certificate by legislative fiat, stranding certificate holders abroad without individualized fault or hearing. Chae Chan Ping’s challenge thus presented the Court with a reliance interest as sympathetic as any in the reports, a man who had done everything the law asked and found the rules changed while he was at sea. That the Court upheld the statute anyway, unanimously, measures the strength of the sovereignty rationale. Reliance, fairness, and treaty commitment all yielded to the proposition that the nation’s power over admission could not be constrained by its prior promises. The second was the last-in-time rule: because the Constitution makes treaties and federal statutes alike the supreme law of the land, with no hierarchy between them, the later in time controls when they conflict. Either move alone would have decided the case. Together they made immigration policy the creature of current legislative majorities, revisable at will, rather than a commitment binding across generations. The combination is what gave the plenary power doctrine its plenary character: a power sourced outside the enumerated clauses and adjustable by simple statute.

Fong Yue Ting tested whether the same logic governed people already inside the country, and the answer the Court gave defined deportation law for the next century. The Geary Act of 1892 required Chinese laborers to register within a year and to carry certificates of residence, with deportation for those who failed to comply. Its evidentiary provision demanded proof of lawful residence by “at least one credible white witness,” a requirement that made the racial logic of the statute explicit on its face. The three challengers, Fong Yue Ting among them, refused to register and argued that deportation without the protections of a judicial trial violated due process.

Gray’s 6 to 3 majority rejected the argument by characterizing deportation as something other than punishment. Expulsion, the opinion held, was “as absolute and unqualified” as exclusion: both were exercises of the sovereign power to determine the composition of the national community, and neither triggered the constitutional protections that attach to criminal punishment.

The registration scheme the Court sustained was designed to be difficult to satisfy, and its mechanics reveal the statute’s purpose. Laborers had one year to register, and the certificate of residence they received became the exclusive proof of lawful presence. Anyone without a certificate was presumed unlawfully present and subject to deportation, with the burden of proving otherwise falling on the laborer. The white-witness rule then made that burden nearly insurmountable for many: a laborer who had lost his certificate, or never obtained one, needed a white witness to vouch for his residence, in communities where few white residents would do so. The scheme’s operation was thus exclusionary in effect whatever its registration label, converting a documentation requirement into a removal engine. The majority’s willingness to sustain it, without scrutinizing its practical operation, demonstrated what plenary power meant in application: the Court would not look behind Congress’s chosen means to assess their fairness, only to confirm that the end, regulating the presence of aliens, lay within the sovereign power. Congress could therefore entrust deportation to executive officers, provide whatever procedures it chose, and condition continued residence on compliance with registration requirements that would have been unthinkable as conditions on citizens. The three dissenters answered that presence in the country, lawful at the time of entry, created a reliance interest that due process protected, and that banishment without judicial trial was punishment by another name. The majority’s characterization won, and its consequences radiated outward: because deportation was civil and administrative rather than criminal and punitive, the rights to jury trial, to appointed counsel, to proof beyond a reasonable doubt, and to protection against ex post facto laws never attached to removal proceedings. The wall between criminal procedure and deportation procedure, which still structures every removal case, was built in Fong Yue Ting.

The one nineteenth century boundary the justices drew against federal power came not from the plenary power line at all but from the Constitution’s own text. United States v. Wong Kim Ark, 169 U.S. 649 (1898), held that a child born in the United States to parents domiciled there was a citizen at birth, by a vote of six to two, with Justice McKenna taking no part. Justice Gray, who had written the expulsion opinion five years earlier, wrote for the majority that the Fourteenth Amendment’s Citizenship Clause decided the question. The ruling issued on March 28, 1898. Its placement matters for the whole field. Where the Constitution speaks directly, in the Citizenship Clause, the plenary power doctrine yields. Where the Constitution is silent, the doctrine governs. That distinction explains why the birthright citizenship question belongs to constitutional law rather than to any statute, and why no later immigration opinion has disturbed it.

The dual authorship is worth pausing over. Justice Gray wrote both the 1893 expulsion opinion and the 1898 citizenship opinion, five years apart, reaching opposite results about federal power. The pairing refutes the lazy reading of the plenary power justices as result-oriented restrictionists. Gray’s method was consistent across both: find the governing legal source and apply it. Where the source was sovereignty, the political branches won; where the source was the Fourteenth Amendment’s text, the individual won. The doctrine, in its founders’ hands, was a theory of legal sources, not a preference for exclusion. That source-driven method is what later majorities inherited, and it explains why the same Court that defers to Congress on admission enforces the Citizenship Clause against Congress without any sense of contradiction.

For decades after 1893, the doctrine sat largely undisturbed, cited as settled law while Congress built the modern immigration code on top of it. The cases were not relitigated because there was little to relitigate: once the power was plenary and its exercise non-punitive, challenges to the substance of exclusion and deportation statutes had nowhere to go.

The consolidation took the form of citation rather than reasoning. When later Courts confronted immigration questions, they reached for Chae Chan Ping and Fong Yue Ting as established premises, reciting the sovereignty rationale in a sentence before moving to the dispute at hand. The doctrine thus acquired the authority of repetition: each new citation made the next challenge harder, not because the reasoning grew stronger but because the weight of precedent grew heavier. By the mid-twentieth century, a litigant arguing that an exclusion decision violated due process was arguing against seventy years of unanimous and near-unanimous precedent, and the argument failed at the threshold. This is how plenary power differs from ordinary doctrine. Ordinary doctrine is refined through challenge; plenary power was insulated from challenge, and insulation compounded over time into something approaching constitutional bedrock.

The action moved to the margins, to procedure, to the states, to statutory interpretation, which is where the rest of this article’s arc takes place. But every marginal dispute was argued in the shadow of the foundation, and the foundation always won the arguments about the center.

The foundation also explains a feature of the field that puzzles newcomers. Immigration lawyers and scholars sometimes describe their subject as “exceptional,” meaning that the normal rules of constitutional law apply differently or not at all. The exception has a birthdate and a rationale. It was born in 1889 and 1893, and its rationale was that a sovereign nation must be able to decide who belongs to it without submitting that decision to judicial balancing. Whether that rationale remains persuasive is a question for argument. That it remains the law, softened but unreversed, is a question the rest of this article answers case by case.

The Border Exception: Deference at Its Maximum

If the nineteenth century cases located the power in sovereignty, the mid-century exclusion decisions tested how far that sovereignty extended when the executive acted alone at the border. The leading decision is Shaughnessy v. United States ex rel. Mezei, 345 U.S. 206 (1953). Ignatz Mezei had lived in the United States for a quarter century before leaving to visit his dying mother in Hungary. When he tried to return, he was excluded on security grounds without a hearing, on the basis of confidential information the government would not disclose. No other country would admit him. He spent twenty-one months detained on Ellis Island while the government and the courts worked out what, if anything, the law required for a man the nation would not admit and no other nation would take.

Justice Tom Clark, writing for the Court, held that Mezei’s exclusion was “largely immune from judicial control.” The opinion drew the sharpest version of the entry fiction the Court has ever endorsed: an alien stopped at the border, even one detained for months on American soil while exclusion proceedings continued, was to be treated as if stopped at the threshold, without the constitutional protections that attach to persons within the territory. The executive’s determination that Mezei should be excluded on security grounds was, in the Court’s telling, a political judgment about admission that courts were not equipped to review. The detention that accompanied the exclusion was treated as part of the exclusion itself rather than as a separate deprivation requiring its own justification.

The threshold theory extended the sovereignty logic to its most austere form. The opinion treated physical presence on Ellis Island as legally irrelevant: the alien had been stopped at the border, had never effected an entry in the legal sense, and therefore stood outside the community of persons to whom the Constitution’s protections run in full. Entry, on this view, is not a geographical fact but a legal construct, conferred by the sovereign’s permission rather than by the traveler’s footsteps. The twenty-one month detention followed from the construct. If the alien had never entered, his confinement was not imprisonment within the country but continued exclusion at its edge, and the political branches could maintain it while seeking a country willing to receive him. The reasoning is formalist, and deliberately so; it refuses to let the fortuity of where the government houses an excluded alien determine the constitutional rule.

Mezei sits at the outer edge of the doctrine, and later decisions would distinguish rather than overrule it, which is itself revealing. The Court in Zadvydas v. Davis, discussed below, would carefully separate Mezei’s situation, an alien stopped at the border seeking initial admission, from the situation of aliens ordered removed after having entered and lived in the country. The distinction between exclusion and deportation, between those seeking entry and those being expelled after entry, became one of the doctrine’s most important internal boundaries. Mezei governs the first category at maximum deference. The deportation cases, beginning with Fong Yue Ting’s insistence that expulsion was equally absolute, complicated the second, and the detention cases of 2001 and 2003 would draw the line in its modern form.

Why did the Court treat exclusion as nearly immune from judicial control?

Because admission was classified as a political act of sovereignty rather than an adjudication of rights. Mezei’s twenty-one months on Ellis Island illustrated the cost: with no country willing to receive him and no hearing required, the executive’s exclusion decision stood unreviewed. The holding marked the high-water line of deference, and later cases distinguished rather than extended it.

The mid-century line matters for a second reason: it established the executive, not just Congress, as a holder of plenary power. Chae Chan Ping and Fong Yue Ting had spoken mainly of the legislative power to exclude and expel. Mezei extended the deference to executive determinations at the border, holding that the political branches together, and the executive in its delegated sphere, could act on admission with minimal judicial oversight. When the modern Court reviews executive action on entry, from visa denials to entry suspensions, it works in Mezei’s shadow.

The entry fiction that did the work in Mezei deserves careful explanation, because it is one of the strangest doctrines in American law and it remains operative. Mezei was physically present on Ellis Island, in New York Harbor, for twenty-one months. Under ordinary constitutional logic, a person detained by the government on American soil for nearly two years would be entitled to challenge the detention. The entry fiction answered that he was not, legally speaking, here at all: an alien stopped at the border and held pending a determination on admission is treated as if still standing at the threshold, outside the territory, for constitutional purposes. The fiction allowed the Court to hold simultaneously that Mezei was detained by the United States and that the Constitution’s protections for detained persons did not reach him. Whether that is a defensible use of legal fiction or an evasion is a question scholars still debate. That it is the law, undisturbed at its core, is not debated.

The detention holding followed from the exclusion holding as a matter of logic. If the executive could exclude Mezei without a hearing on undisclosed security grounds, and if no other country would receive him, then some detention was the inevitable incident of the exclusion; release into the country would have defeated the exclusion itself. The Court therefore treated the twenty-one months not as a separate deprivation of liberty requiring its own justification but as the administrative consequence of an unreviewable admission decision. This is the reasoning Zadvydas would later cabin: once the context shifted from exclusion at the threshold to removal after presence, the detention could no longer ride on the exclusion’s coattails, and it needed its own constitutional accounting.

Mezei also fixed the executive as an independent holder of plenary power. The nineteenth century cases had spoken of Congress’s power to exclude and expel. Mezei’s exclusion was an executive determination, made on confidential security information, without statutory hearing procedures, and the Court sustained it as an exercise of the political branches’ shared sovereign prerogative. The “largely immune from judicial control” formulation thus covered both branches: Congress when it legislated admission criteria, the executive when it applied them at the border. That dual coverage is what makes the doctrine so resistant to challenge. A litigant who cannot get past Congress finds the executive standing behind the same shield.

Congress later reorganized the vocabulary without disturbing the doctrine’s substance. The 1996 amendments replaced the old exclusion and deportation proceedings with a single removal proceeding, retiring the terms exclusion and deportation from the statute’s procedural architecture. The constitutional distinction survived the renaming: courts continued to treat arriving aliens seeking admission differently from persons present in the country, applying Mezei’s deference to the former and the calibrated protections of the later detention cases to the latter. The vocabulary changed; the two-track doctrine did not. That persistence is itself evidence for the complication this article addresses. When Congress rewrote the procedural machinery, the Court mapped the old constitutional categories onto the new statutory ones rather than revisiting them, which is what courts do with doctrines they consider settled rather than with doctrines they consider ripe for reconsideration.

The Softening at the Edges: Plyler and the State Federal Divide

The first great softening of the doctrine did not limit Congress or the president at all. It limited the states. In 1975, Texas enacted a statute that withheld state funds from school districts for the education of children who had entered the United States unlawfully, and authorized districts to deny enrollment to such children. Families in Tyler, Texas challenged the law on behalf of their children. The case reached the Supreme Court as Plyler v. Doe, 457 U.S. 202 (1982), decided June 15, 1982.

Justice William Brennan wrote for a 5 to 4 majority that included Justices Marshall, Blackmun, Powell, and Stevens, with separate concurrences. The opinion struck down the Texas law on equal protection grounds. Its reasoning had several interlocking parts. The children had not themselves violated the law; they were present because of decisions their parents made, and penalizing children for their parents’ conduct offended basic principles of justice. Education was not a fundamental right under the Constitution, but it was not merely another government benefit either; denying it to a discrete group of children would impose a lifetime of disadvantage and create a permanent underclass. The state’s asserted interests, preserving resources and deterring unlawful immigration, did not survive the heightened review the Court applied: Texas could not show that excluding these children would improve educational quality for the rest or that the exclusion meaningfully advanced immigration enforcement, a federal responsibility in any event.

Chief Justice Warren Burger dissented, joined by Justices White, Rehnquist, and O’Connor. The dissent argued that the Court was creating a constitutional entitlement to public education for persons unlawfully present, a step the Constitution did not require and the political process had not taken. Immigration policy belonged to the federal government, and Texas was entitled to allocate its limited educational resources to lawful residents. The dissent warned that the majority’s reasoning had no natural stopping point: if unlawful presence could not justify exclusion from schools, it was unclear what public benefits a state could withhold.

Plyler’s place in the plenary power story is precise and frequently misstated. The decision constrained states, not the federal government. Nothing in Brennan’s opinion suggested that Congress could not set the terms of admission, removal, or federal benefits. The equal protection analysis applied to Texas because Texas was acting in an area where the federal government held primary authority and where the state’s discrimination lacked adequate justification. The recurring error, flagged in the brief for this article, is treating Plyler as a federal-power case. It is a federalism case wearing equal protection dress: it says that whatever latitude the plenary power doctrine gives Washington, it does not give the states a license to visit the consequences of unlawful presence on children through the schoolhouse door.

The school-access dimension of Plyler belongs to the companion article on education law in the Supreme Court, which develops the education doctrine in full; this article owns the immigration question, which is why the plenary power doctrine constrains federal review far less than it constrains the states. Plyler also illustrates the asymmetric pattern that runs through the entire field. When the federal government acts on immigration, courts apply the plenary power doctrine and defer. When a state acts on immigration, courts apply preemption and equal protection and scrutinize. The same field thus produces sweeping deference in one direction and firm limits in the other, and that asymmetry is not an accident of case selection. It is the structure of the doctrine.

The decision left open the questions its logic seemed to invite. If Texas could not deny K through 12 schooling, could it deny other benefits? Could the federal government deny benefits that states could not? Later decisions would answer some of these, but Plyler itself was deliberately narrow: public elementary and secondary education, for children, under a state law, on equal protection grounds. Its narrowness is part of its durability. By deciding little beyond the case before it, the majority produced a holding that survived decades of political controversy without being revisited.

The equal protection framework Brennan employed was itself unusual, and it explains both the decision’s reach and its limits. Education was not a fundamental right under the Court’s precedents, so strict scrutiny did not apply. But the classification, children penalized for their parents’ conduct, combined with the interest at stake, schooling that determined life prospects, led the majority to require more than the minimal rationality review that usually governs economic and social legislation. Texas had to show that its discrimination furthered a substantial state interest, and it could not. The cost rationale failed because the state produced no evidence that excluding undocumented children would improve the education of the rest or save meaningful sums. The deterrence rationale failed because immigration enforcement was a federal responsibility, and a state school-enrollment rule was a poor and indirect instrument for it. The opinion thus applied a form of heightened review without naming a new tier of scrutiny, a middle path that gave the holding force while keeping it confined to its facts.

The separate concurrences, by Justices Marshall, Blackmun, Powell, and Stevens, are worth noting for what they signaled about the majority’s fragility and its care. A 5 to 4 decision with multiple separate writings is a coalition, not a consensus, and each concurrence emphasized a different strand: the injustice of penalizing children, the special status of education, the federalism concern about states intruding on immigration enforcement. The dissent, led by Chief Justice Burger, pressed the institutional objection with equal force: the Court was ordering states to spend their limited educational resources on persons whose presence violated federal law, creating an entitlement the political branches had never enacted and the Constitution did not command. Presented with equal care, the two sides describe a genuine constitutional disagreement, not a lapse by either. The majority saw innocent children facing permanent disadvantage; the dissent saw judges displacing democratic control of public resources and immigration policy.

What Plyler left open is as instructive as what it decided. The opinion addressed public elementary and secondary education, not higher education, not health care, not other public benefits. It addressed children, not adults. It addressed state action, not federal action, leaving untouched the question whether Congress could condition federal benefits on immigration status. Later controversies over each of those boundaries were litigated in Plyler’s shadow, with litigants on both sides claiming its logic. The decision’s careful confinement to K through 12 schooling for children under state law is what allowed it to endure: by refusing to announce a general principle about the rights of unlawfully present persons, the majority denied future courts a broad precedent to extend and denied critics a broad target to attack.

Later litigants tested those boundaries, and the results confirmed the confinement. Challenges invoking Plyler for other benefits met the limits of its reasoning: the opinion’s special solicitude for elementary and secondary education did not transfer automatically to every public program, and its focus on children did not extend to adults. Defenders of restrictive state measures, meanwhile, learned to draft around it, targeting benefits and age groups the opinion had not addressed. The decision thus functioned less as a broad charter than as a fixed point: unmoved itself, but surrounded by contested territory where its logic applied by analogy or not at all depending on the court’s assessment. That is the fate of narrow holdings in controversial fields. They survive by deciding little, and they govern by the persuasive force of their reasoning rather than by the compulsion of their rule.

The Structural Case: How Chadha Killed the Legislative Veto

No immigration case has ever had a larger blast radius than Immigration and Naturalization Service v. Chadha, 462 U.S. 919 (1983). The case began with one man. Jagdish Rai Chadha, a student whose visa had expired, faced deportation. An immigration judge found that deportation would cause him extreme hardship and suspended it under section 244(a)(1) of the Immigration and Nationality Act, which authorized the Attorney General to grant such relief. Section 244(c)(2) of the same Act then gave either House of Congress the power to overturn the suspension by resolution. The House of Representatives did exactly that, voting to deport Chadha without debate, without a stated reason, and without presenting the resolution to the Senate or the president.

The suspension practice that produced the case was itself a product of plenary power’s harshness. Deportation under the Act was mandatory for broad categories of aliens, including long-term residents whose equities were strong but whose statutory defenses were exhausted. Section 244(a)(1) gave the Attorney General a safety valve: upon a finding of extreme hardship, deportation could be suspended and the alien’s status adjusted. Immigration judges granted thousands of such suspensions, exercising delegated discretion case by case. The one-House veto was Congress’s retained check on that discretion, a way to police the safety valve without repealing it. Chadha’s case showed the veto operating at its most arbitrary, a legislative body overturning an individualized hardship determination without explanation. The constitutional question was whether that arbitrariness mattered. The majority held that it did not need to reach the fairness of the House’s action, because the procedure itself was unconstitutional regardless of how fairly or unfairly it was used in any particular case.

Chadha challenged the one-House veto as unconstitutional. On June 23, 1983, the Court agreed by a 7 to 2 judgment. Chief Justice Burger wrote the majority opinion, joined by Justices Brennan, Marshall, Blackmun, Stevens, and O’Connor. The reasoning was structural and, in the majority’s telling, straightforward. The House resolution had the purpose and effect of altering the legal rights and duties of Chadha and the Attorney General. An action with that effect is legislation. The Constitution prescribes a single, exclusive procedure for legislation: passage by both Houses and presentment to the president, the bicameralism and Presentment Clauses of Article I, Section 7. A resolution passed by one House alone and never presented to the president could not satisfy that procedure, and no statute could authorize Congress to bypass it.

The constitutional mechanics behind the holding are worth stating in full, because they explain why it could not be contained. Article I permits Congress to act legislatively only through the concurrence of both Houses and presentment to the president, who may sign or veto. The framers designed that procedure as a deliberate obstacle: bicameralism forces a measure to survive two different majorities elected on different cycles, and presentment forces it to survive executive review, with the two-thirds override as the safety valve. The procedure is slow, public, and accountable, and its slowness is the point. A one-House resolution bypasses every element: one chamber acts alone, the president never sees the measure, and no override is possible because there is nothing to override. The majority’s insight was that the label on the action cannot change its nature. If the resolution alters legal rights and duties beyond the legislative branch, it is legislation whatever Congress calls it, and legislation must run the constitutional gauntlet.

Justice Powell concurred in the judgment on narrower grounds, arguing that the House had exercised a judicial function by overturning an individualized adjudication, and that the case could be resolved on separation of powers principles without reaching every legislative veto. Justice White dissented at length, joined in part by the practical argument that the legislative veto was an indispensable instrument of modern governance: Congress had written veto provisions into nearly two hundred statutes, covering some two hundred separate provisions, as the price of delegating broad authority to executive agencies. White’s dissent, at 462 U.S. 967 to 968, catalogued the scale of what the majority was destroying. Chief Justice Rehnquist also dissented, joined by Justice White, on severability grounds, arguing that the veto provision could not be separated from the suspension authority and that both should fall together.

The majority’s logic could not be contained to immigration. If a one-House veto of a deportation suspension violated bicameralism and presentment, so did every other one-House veto, and so did the two-House vetoes that bypassed presentment. The decision invalidated the legislative veto across the entire United States Code, in fields from arms sales to trade to environmental regulation. That is why the case appears in every constitutional law course and almost no immigration course: its immigration origins are a historical curiosity next to its structural consequences. But the origins matter for this article’s thesis. The broadest judicial check on congressional power ever to emerge from an immigration file did not limit the substance of immigration law at all. It limited the procedures Congress could use to control the executive, and it did so in terms that applied far beyond the field.

Chadha also clarified something about the plenary power doctrine by negative implication. The doctrine gives Congress extraordinary latitude over the substance of immigration law: who may enter, who must leave, on what terms. It does not give Congress latitude over the constitutional procedures for exercising legislative power. Plenary over substance, constrained on structure. That distinction would recur.

The legislative veto had been Congress’s answer to a dilemma of the administrative state, and understanding the dilemma explains why White’s dissent treated the majority opinion as a disaster. Across the twentieth century, Congress delegated sweeping authority to executive agencies: to regulate, to spend, to suspend deportations, to manage trade and arms sales. Delegation on that scale created an accountability problem, because the same Congress that granted the power could not realistically repeal or rewrite the underlying statute every time an agency used the power in a way the current Congress disliked. The veto solved the problem elegantly. Congress would grant the authority and reserve a check: if the executive’s action offended, one or both Houses could nullify it by resolution, without the cumbersome process of passing a new law over a presidential veto. By the early 1980s, as White’s dissent catalogued, the device appeared in nearly two hundred statutes, a load-bearing element of the modern regulatory bargain.

The majority’s answer was that elegance could not cure unconstitutionality. Burger’s opinion treated the question as one of first principles rather than of practical governance. The Constitution’s lawmaking procedure was exclusive and deliberate: the framers had considered and rejected shortcuts, and they had settled on bicameralism and presentment as the price of legislation. A resolution that altered Chadha’s legal status, directing his deportation after the Attorney General had suspended it, was legislation in substance whatever label Congress gave it, and legislation required both Houses and the president’s opportunity to veto. The opinion acknowledged the practical cost and dismissed it as irrelevant: the Constitution’s procedures were not suggestions to be waived when inconvenient.

Powell’s concurrence in the judgment offered a narrower path the majority declined to take. Powell argued that the House had acted as a court, overturning an individualized adjudication of hardship without the procedural protections of adjudication, and that the case could be resolved on separation of powers grounds without invalidating every legislative veto in the Code. White’s dissent made the opposite case for breadth, defending the veto as a necessary and long-accepted instrument of congressional oversight. Rehnquist’s dissent, joined by White, added the severability point: if the veto fell, the suspension authority it checked should fall with it, since Congress would not have granted unreviewable suspension power standing alone. The majority rejected all three alternatives, severed the veto, preserved the suspension authority, and let the consequences fall across the statute books. Congress adapted by substituting report-and-wait requirements and other oversight devices that stopped short of nullification, but the veto as a legal instrument was finished.

The aftermath illustrated the difference between a legal rule and a political practice. Congress continued to write provisions that looked like vetoes, requiring agencies to notify committees and wait a specified period before acting, and agencies continued to treat committee objections as practically binding even without legal force. The form changed; the underlying desire for control did not. What Chadha removed was the legal entitlement to nullify, not the political reality of oversight. For immigration specifically, the consequence was that the Attorney General’s suspension decisions, once subject to one-House reversal, became final exercises of executive discretion reviewable only through the ordinary channels. The plenary power over substance remained with the political branches, but the mechanism for dividing that power between them had to be rebuilt on constitutional foundations, through statutes that both Houses passed and presidents signed. When the Court later reviewed immigration statutes for compliance with due process, habeas, and non-retroactivity principles, it was applying the same lesson: the political branches decide the policy, but the Constitution still polices the process.

How did the 2002 reorganization change the names in these case captions?

The Homeland Security Act of 2002 abolished the Immigration and Naturalization Service effective March 1, 2003, splitting its work among USCIS, ICE, and CBP. Earlier disputes are therefore captioned against the INS, while later ones name the Department of Homeland Security. The reorganization guide maps which successor inherited each authority.

Statutory Interpretation: Asylum Standards, Habeas, and Retroactivity

The late 1980s and the turn of the century produced a pair of decisions that look, at first glance, like ordinary statutory interpretation cases. They are, but in immigration law statutory interpretation is never only about text. The background presumption of plenary power means that ambiguous provisions are often read to favor the government, so when the Court reads them the other way, the reasoning repays close attention.

INS v. Cardoza-Fonseca, 480 U.S. 421 (1987), decided March 9, 1987, concerned the standard for asylum. The Refugee Act of 1980 had created two forms of protection with two different standards. Section 208 of the Act, governing asylum, required an applicant to show a “well-founded fear of persecution.” Section 243(h), governing withholding of deportation, required a showing that persecution was a “clear probability,” meaning more likely than not. The Board of Immigration Appeals had effectively merged the two, applying the stricter clear-probability test to asylum claims as well. Luz Marina Cardoza-Fonseca, a Nicaraguan national, challenged that conflation.

Justice Stevens wrote an opinion joined by only four other justices, Brennan, Marshall, Blackmun, and O’Connor, with Justice Blackmun adding a separate concurrence and Justice Scalia concurring in the judgment, producing a 6 to 3 judgment for reversal. Justice Powell dissented, joined by Chief Justice Rehnquist and Justice White. The majority held that Congress had deliberately chosen different language for the two provisions and meant different things by them. The opinion walked through text, structure, and legislative history: the well-founded fear standard came from the United Nations Protocol Relating to the Status of Refugees, which the Refugee Act was written to implement, and the Protocol’s standard was understood internationally as more generous than a more-likely-than-not test. A ten percent chance of persecution, the opinion suggested by way of illustration, could support a well-founded fear even though it fell far short of clear probability. The BIA’s conflation was rejected, and the asylum standard was restored to its independent, more protective meaning.

The asylum provision at the heart of Cardoza-Fonseca was created by the 1980 statute whose humanitarian framework is examined in the companion guide to the Refugee Act of 1980. The decision matters beyond its immediate parties because it established that even within a field of extraordinary deference, the Court would enforce Congress’s textual choices against the executive’s narrowing construction. The plenary power doctrine did not license the agency to rewrite the statute’s standards. Deference to the political branches meant deference to Congress’s enacted text, not to the executive’s preferred reading of it.

The opinion’s interpretive method is as important as its result. The majority began with text, noting that Congress had used different phrases in the two provisions and must be presumed to have meant different things by them. It turned to history, tracing the asylum provision to the 1980 legislation and the legislative record showing Congress’s intent to conform domestic law to the international refugee definition. It consulted the United Nations convention and protocol, not as binding law but as evidence of what Congress meant when it borrowed the convention’s language. Only after that did it address the agency’s reading, and it found the reading unworthy of deference because the agency had not interpreted an ambiguity but had erased a distinction the statute plainly drew. The method, text, then history, then the international source, then the agency, in that order, became a template for immigration interpretation.

The international background of the asylum standard explains why Congress chose the words it did. The Refugee Act of 1980 was written to bring United States law into conformity with the United Nations Protocol Relating to the Status of Refugees, which defined a refugee as a person with a well-founded fear of persecution. The Protocol’s standard was understood, in the international practice Congress meant to adopt, as a generous one: it asked whether a reasonable person in the applicant’s position would fear persecution, not whether persecution was more likely than not. When the Board of Immigration Appeals applied the clear-probability test to asylum claims, it was therefore not merely misreading a domestic statute. It was defeating the international commitment the statute was enacted to implement, replacing a fear-based standard with a probability-based one that Congress had deliberately rejected for asylum while retaining it for withholding.

The deference question lurking beneath the case concerned the agency’s claim to interpretive authority. The Board argued, in effect, that its construction of the ambiguous relationship between the two standards deserved judicial respect. Stevens’s opinion answered with text, structure, and history rather than with deference doctrine: the two provisions used different language, occupied different sections, and served different functions, asylum as a discretionary benefit reflecting humanitarian judgment, withholding as a mandatory protection against return to persecution. Where Congress had spoken distinctly, the agency could not merge. Scalia’s concurrence in the judgment approached from the deference side, agreeing with the result while emphasizing the proper framework for reviewing agency interpretations. Powell’s dissent, joined by Rehnquist and White, would have sustained the Board’s reading as a permissible construction of related provisions. The lineup matters because it shows the decision was not an ideological bloc voting on immigration sympathies. It was a disagreement about how to read a statute, resolved by the ordinary tools of textual analysis, in a field where those tools are often thought to yield to deference.

The practical stakes were substantial. Under the Board’s conflated standard, asylum seekers had to prove that persecution was more likely than not, a burden many genuine refugees could not carry, since persecution is often threatened rather than certain and evidence from repressive regimes is scarce. Under the restored well-founded fear standard, an applicant who showed a reasonable possibility of persecution, illustrated by the opinion’s example of a one-in-ten chance, could qualify. The decision thus expanded the class of persons eligible for asylum without altering the Attorney General’s ultimate discretion to grant or deny it, a distinction the opinion preserved: eligibility was a legal question the Court could decide, while the discretionary grant remained with the executive.

The asylum-withholding distinction the case restored continues to structure protection claims. Asylum, under the well-founded fear standard, is discretionary: even an applicant who meets the standard may be denied as a matter of discretion, though in practice grants follow eligibility. Withholding, under the clear-probability standard, is mandatory once the standard is met: the government may not return a person who shows persecution is more likely than not. The two forms of protection thus differ in both threshold and consequence, generous threshold with discretionary relief on one side, demanding threshold with mandatory relief on the other. The Board’s conflation had collapsed both dimensions, applying the demanding threshold while retaining discretion. Cardoza-Fonseca restored the statute’s design, and in doing so demonstrated the Court’s willingness to enforce congressional calibrations against executive simplification, even in a field of plenary power.

INS v. St. Cyr, 533 U.S. 289 (2001), decided June 25, 2001, presented two questions arising from the 1996 amendments to the immigration laws. Enrico St. Cyr, a lawful permanent resident, had pleaded guilty to a criminal offense before the 1996 changes and then faced removal under provisions that, among other things, repealed section 212(c) of the Act, the discretionary relief that had previously allowed long-term residents in his position to seek a waiver. He sought habeas corpus review in federal court. The government argued that the 1996 statutes had stripped the courts of habeas jurisdiction and that the repeal of section 212(c) applied to his pre-amendment plea.

Justice Stevens wrote for a 5 to 4 majority, joined by Justices Kennedy, Souter, Ginsburg, and Breyer. Justice O’Connor dissented, joined by Chief Justice Rehnquist and Justices Scalia and Thomas. The opinion reached two holdings, each grounded in a clear-statement rule. First, the 1996 provisions had not repealed habeas corpus jurisdiction under 28 U.S.C. section 2241, because Congress had not spoken with the clarity required to displace the historic writ; the Suspension Clause concerns hovering over any such repeal demanded explicit language the statutes lacked. Second, the repeal of section 212(c) could not be applied retroactively to aliens who had pleaded guilty before the repeal, because plea bargains involve a quid pro quo in which the defendant gives up constitutional rights in reliance on the law as it stood, and applying the new law to undo that reliance would attach new legal consequences to completed conduct under the Landgraf framework.

The second holding requires a precision the fact record for this article insists upon. St. Cyr did not hold that the 1996 amendments’ expanded definition of aggravated felony could never apply retroactively. Section 321(b) of IIRIRA expressly stated that the expanded definition applied regardless of when the conviction occurred, and the St. Cyr opinion acknowledged at 533 U.S. 319 that this provision “clearly states” its retroactive reach. The retroactivity limit the Court imposed concerned the repeal of discretionary relief as applied to pre-statute guilty pleas, where reliance interests were at their peak, not a wholesale invalidation of retroactive application across the 1996 package. Collapsing the two is one of the recurring errors in summaries of the decision.

The 1996 amendments that generated St. Cyr, along with much of the modern detention and removal litigation, are examined in the companion guide to the 1996 immigration act, which covers the provisions most of the modern cases construe. St. Cyr’s significance for the plenary power story is that it used interpretive canons, the presumption against implied repeal of habeas and the presumption against retroactivity, to constrain the executive’s reading of a harsh statute without questioning Congress’s power to write a harsh statute. Had Congress spoken clearly, the outcome would have differed. The doctrine survived; the agency’s overreading did not.

The habeas holding turned on one of the oldest interpretive principles in American law: courts do not infer the repeal of habeas corpus jurisdiction from ambiguous language. The 1996 amendments contained elaborate provisions channeling judicial review of removal orders into the courts of appeals and restricting district court jurisdiction, and the government argued that this architecture left no room for habeas petitions under the general habeas statute. Stevens’s majority answered that displacing the historic writ required Congress to speak with unmistakable clarity, and nothing in the 1996 provisions did so. Behind the clear-statement rule stood the Suspension Clause, which permits suspension of the writ only in cases of rebellion or invasion. Reading the statutes to eliminate habeas without clear language would have forced the Court to decide whether the resulting scheme violated the Clause. The avoidance principle counseled against that confrontation: construe the statutes to preserve the writ, and the constitutional question never arises.

The retroactivity holding applied the framework of Landgraf v. Rivers, which asks first whether Congress clearly prescribed a statute’s temporal reach and then, if not, whether applying it to past conduct would attach new legal consequences to completed transactions. The repeal of section 212(c) failed at the second step for one specific class: aliens who had pleaded guilty before the repeal. A guilty plea is a bargained exchange in which the defendant surrenders the right to trial, the right against self-incrimination, and other constitutional protections, in reliance on the legal consequences the plea carries at the time. To retroactively eliminate the discretionary relief available when the plea was entered was to change the price of the bargain after it was struck. The opinion was careful to confine this logic to pleas, where reliance was concrete and documented, rather than extending it to all pre-amendment conduct. And it preserved the aggravated-felony expansion’s express retroactivity under section 321(b), which the opinion acknowledged “clearly states” its reach. The result was a calibrated retroactivity doctrine: reliance-protecting where pleas were involved, statute-respecting where Congress had spoken plainly.

O’Connor’s dissent, joined by Rehnquist, Scalia, and Thomas, read the 1996 jurisdiction provisions as a comprehensive scheme that left no office for habeas review and would have applied the repeal to St. Cyr’s case. The disagreement, presented with the care the subject requires, was about how much clarity the clear-statement rules demanded and whether the 1996 Congress had supplied it. Both sides agreed on the underlying principle that Congress could legislate harshly and retroactively if it did so explicitly. They divided only on whether it had.

The practical consequence of the habeas holding was to preserve a forum. Without habeas, aliens raising the retroactivity claim would have been left to whatever review the 1996 channeling provisions allowed, which the government argued was none. With habeas preserved, federal courts could hear the claim that the repeal was being applied unlawfully to pre-statute pleas, and the retroactivity holding gave them the rule to apply. The two holdings thus worked as a pair: the first kept the courthouse door open, the second supplied the law to apply inside. For the plenary power story, the pairing illustrates the doctrine’s resilience. The Court constrained the executive’s reading of the statute and protected reliance interests without questioning for a moment Congress’s authority to repeal discretionary relief prospectively, to expand the aggravated-felony definition retroactively where it spoke clearly, or to legislate the entire enforcement scheme in the first place. Deference to the legislative power survived intact; only the executive’s attempt to stretch ambiguous text beyond what Congress had clearly authorized failed.

The agency deference question runs underneath both interpretation decisions and deserves to be named directly, because it is where immigration law meets administrative law most sharply. The Board of Immigration Appeals decides hundreds of thousands of cases, and the question of how much respect its readings deserve determines who, as a practical matter, makes immigration law day to day. Cardoza-Fonseca gave a two-part answer. Where the agency interprets a genuine ambiguity, its reasonable reading receives respect; where the agency erases a distinction the statute plainly draws, no deference is owed, because there is nothing to defer to. The dissent would have applied the first half more generously, treating the relationship between the two standards as exactly the kind of interstitial question agencies exist to resolve. The majority treated it as a question the statute had already answered. The disagreement recurs everywhere the Board’s readings are challenged: is this an ambiguity the agency may fill, or a command the agency must obey. St. Cyr added the clear statement overlay, which functions as a thumb on the scale against the agency where background law is strong. Habeas and retroactivity presumptions do not merely guide the Court; they constrain the agency, because the Court will not sustain an agency reading that displaces them without plain congressional authorization.

The practical consequence is that immigration agencies lose interpretation cases more often than agencies in other fields, not because the Court is hostile but because the field’s interpretive presumptions are unusually strong. The asylum standard case, the habeas case, and the detention case all rejected agency-favored readings, while the detention-during-proceedings case sustained the agency’s position only where the statute spoke in mandatory terms. A practitioner briefing an interpretation dispute should therefore ask first whether any background presumption favors the client’s reading, because the Court will apply it before reaching deference, and second whether the statute’s text genuinely leaves room for the agency’s view, because invented ambiguity gets no respect. Those two questions decide more immigration interpretation cases than any theory of deference.

Detention: A Limit Read In, and Detention Sustained

Two decisions, three years apart, drew the modern boundaries of immigration detention. They point in opposite directions, and the reason they can coexist is that they concern different stages of the process and different liberty interests.

Zadvydas v. Davis, 533 U.S. 678 (2001), decided June 28, 2001, three days after St. Cyr, concerned detention after a final order of removal. Kestutis Zadvydas had been ordered removed, but no country would accept him, and the government held him in indefinite post-order custody under section 241(a)(6) of the Act, codified at 8 U.S.C. section 1231(a)(6), which authorized detention beyond the ninety-day removal period for certain categories of aliens. Justice Breyer wrote for a 5 to 4 majority, joined by Justices Stevens, O’Connor, Souter, and Ginsburg. Justices Scalia and Kennedy dissented separately.

The majority did not strike down the statute. It construed it, invoking the canon of constitutional avoidance: because indefinite detention of a person already ordered removed would raise serious due process concerns, the Court read the provision to authorize detention only for a period “reasonably necessary” to bring about removal. To give that phrase administrable content, the opinion established a framework. Six months was deemed a “presumptively reasonable” period. After that point, an alien who could show “no significant likelihood of removal in the reasonably foreseeable future” was entitled to release, subject to the government’s opportunity to rebut the showing. The statute thus carried an implicit temporal limit, derived not from its text alone but from the constitutional doubt that an unlimited reading would create.

The opinion carefully distinguished Mezei. The mid-century case had concerned an alien stopped at the border and seeking initial admission, the exclusion context where deference ran strongest. Zadvydas concerned aliens who had entered, lived in, and been ordered removed from the country, the deportation context where the liberty interest was weightier and the constitutional avoidance canon had more purchase. The distinction between exclusion and deportation, seeded in the nineteenth century cases and sharpened at mid-century, did decisive work here. It allowed the Court to limit detention without repudiating the deference that Mezei embodied.

The avoidance canon that produced the limit operates as a negotiation between the Court and Congress, and its mechanics are worth stating precisely. The canon provides that when a statute is susceptible to two readings, one raising serious constitutional doubts and one avoiding them, courts choose the avoiding reading. In Zadvydas, the serious doubt was whether indefinite detention of a person the government could not remove violated due process. The avoiding reading was that section 1231(a)(6), which authorized continued detention without specifying a duration, implicitly authorized only detention for the period reasonably necessary to effect removal. The canon thus allowed the majority to impose a limit without declaring the statute unconstitutional, preserving Congress’s power to write a clearer law while refusing to assume Congress had written an unconstitutional one. Scalia’s dissent objected that the statute contained no temporal limit to find, and that the majority was amending rather than interpreting. The disagreement was about the judicial role under the canon, not about the existence of plenary power.

The six-month framework gave the limit administrable shape. For the first six months after the removal period, detention was presumptively reasonable, and the alien bore the burden of showing otherwise. After six months, an alien who demonstrated no significant likelihood of removal in the reasonably foreseeable future shifted the burden to the government to rebut the showing or release the alien, typically under supervision. The framework balanced the government’s legitimate interests, in detaining dangerous persons and in preserving the ability to remove when a receiving country could be found, against the liberty interest in not being held indefinitely for a removal that would never happen. It was, in substance, a judicially crafted release valve for a statute that had none, built from the constitutional doubt the statute’s silence created.

The reasonably foreseeable future standard has proved generative in practice. The alien’s burden is to show no significant likelihood of removal within that horizon, a showing typically made through evidence that no country will accept him, that travel documents cannot be obtained, or that diplomatic efforts have failed over time. The government’s rebuttal typically offers evidence of ongoing negotiations, changed country conditions, or new removal arrangements. The six-month presumption structures the inquiry without deciding it: before six months, the government need not justify continued detention; after six months, the alien’s showing shifts the practical burden. The standard is deliberately open-textured, inviting case-by-case adjudication rather than categorical rules, and that openness is characteristic of the field.

Demore v. Kim, 538 U.S. 510 (2003), decided April 29, 2003, sustained the other side of the line. Hyung Joon Kim, a lawful permanent resident convicted of criminal offenses, was held in mandatory detention during his removal proceedings under section 236(c) of the Act, codified at 8 U.S.C. section 1226(c), which required detention of certain criminal aliens pending the outcome of their cases. Chief Justice Rehnquist wrote for the Court, sustaining the provision by a 5 to 4 vote on the constitutional holding. The opinion emphasized the brevity and purpose of the detention: it lasted only for the duration of proceedings, which in the mine-run case meant a matter of months, and it served the legitimate purposes of preventing flight and protecting the public during the pendency of removal.

Justice Kennedy’s concurrence in the judgment supplied the deciding fifth vote and the opinion’s most quoted qualification. Kennedy agreed that mandatory detention during proceedings was constitutional in the ordinary case but warned that an unreasonable delay or a prolonged detention without adequate justification would raise different constitutional questions. The concurrence thus built a safety valve into the holding: the statute survived as applied to detention of the expected, limited duration, not as a blank check for indefinite custody during endlessly protracted proceedings. By tying approval to the typical brevity of removal proceedings, the concurrence converted a five-vote judgment into a narrow one. The government won the authority to detain mandatorily during proceedings, but it did not win authority to detain indefinitely while proceedings stall.

The categorical nature of section 1226(c) is what made Demore a hard case. The provision did not merely authorize detention; it required it, for criminal aliens as defined by the statute, without an individualized bond hearing at which the alien might show he posed no flight risk or danger. In ordinary due process analysis, deprivations of liberty trigger individualized determinations, and mandatory detention without a hearing is difficult to sustain. Rehnquist’s majority answered that the immigration context changed the calculus: Congress had found that criminal aliens as a class posed heightened risks of flight and recidivism during proceedings, the detention was limited to the pendency of the removal case, and the plenary power doctrine gave Congress latitude to legislate categorically in this field that it would not enjoy elsewhere. The 5 to 4 division on the constitutional holding showed how close the question was, and Kennedy’s concurrence narrowed the majority’s rationale to its facts. The controlling understanding of Demore, the one lower courts applied afterward, was Kennedy’s: mandatory detention during proceedings of ordinary duration was permissible, but detention that became prolonged or unjustified would be judged under the constitutional doubts Zadvydas had mapped.

Taken together, the detention pair maps the field’s internal geography. At the border seeking admission, deference is at its maximum and Mezei governs. During proceedings to determine removability, detention is permitted as a brief incident of the process and Demore governs, with Kennedy’s duration caveat. After a final order, when removal cannot be effected, detention must end within a reasonable period and Zadvydas governs. The plenary power doctrine was not abandoned in any of these; it was channeled. The Court deferred where the political branches’ judgment was at its most legitimate, at admission and during the adjudicative process, and it intervened where detention had become untethered from any removal that was actually foreseeable.

The functional test the two decisions share can be stated as a single question: what is the detention for. Detention pending a removal decision serves the proceeding; it ensures the alien appears, prevents flight, and protects the community during the interval before the government decides. Detention after a removal order serves the removal; it holds the alien while the government arranges travel and transfer. When the purpose is active, the detention is an incident of the immigration power and receives deference. When the purpose has evaporated, because removal is not reasonably foreseeable, the detention becomes purposeless confinement, and purposeless confinement is what due process forbids the government to impose without criminal process. The test is functional rather than categorical: it does not ask what label the statute bears but what work the custody performs. That is why Zadvydas and Demore, reaching opposite results under the same 1996 legislation, are consistent. The purposes differ, so the outcomes differ.

The dissents in both cases are worth stating with the same care as the majorities. In Zadvydas, Justice Scalia argued that the Court had rewritten a statute that contained no temporal limit, usurping a legislative judgment committed to Congress under the plenary power doctrine. Justice Kennedy argued that the majority’s framework invited endless litigation over foreseeability and undervalued the government’s interest in detaining dangerous aliens. In Demore, the dissenters argued that mandatory detention without an individualized bond hearing violated due process for permanent residents with substantial ties to the country. The debate in both cases was not about whether the plenary power doctrine existed but about what it permitted at the margins, which is where all the modern action in this field takes place.

The doctrine ledger

The artifact the brief requires is a single table stating, for each decision, the year, the provision or power at issue, the holding in one sentence, and whether the decision expanded, constrained, or merely applied the plenary power framework.

Decision	Year	Provision or power at issue	Holding in one sentence	Plenary power effect
Chae Chan Ping v. United States, 130 U.S. 581	1889	Congressional power to exclude aliens; later statute overrides earlier treaty	Exclusion is an incident of national sovereignty rather than an enumerated power, so a later statute may override an earlier treaty.	Expanded
Fong Yue Ting v. United States, 149 U.S. 698	1893	Congressional power to expel aliens under the Geary Act	The power to expel is as absolute and unqualified as the power to exclude, and Congress may entrust deportation to executive officers.	Expanded
Shaughnessy v. United States ex rel. Mezei, 345 U.S. 206	1953	Executive exclusion and detention at the border	Exclusion decisions are largely immune from judicial control, and detention incident to exclusion requires no separate justification.	Expanded
Plyler v. Doe, 457 U.S. 202	1982	Texas denial of public K through 12 schooling; equal protection	A state may not deny free public schooling to children present unlawfully, a limit that binds states rather than the federal government.	Constrained
INS v. Chadha, 462 U.S. 919	1983	One-House legislative veto of deportation suspension, INA section 244(c)(2)	A one-House veto that alters legal rights is legislation subject to bicameralism and presentment, invalidating the veto device across the Code.	Constrained
INS v. Cardoza-Fonseca, 480 U.S. 421	1987	Asylum standard, INA section 208; withholding standard, section 243(h)	Well-founded fear for asylum is a more generous standard than clear probability for withholding, and the agency may not conflate them.	Applied
INS v. St. Cyr, 533 U.S. 289	2001	Habeas under 28 U.S.C. section 2241; retroactive repeal of INA section 212(c)	Habeas survives absent a clear statement of repeal, and the section 212(c) repeal cannot apply retroactively to pre-statute guilty pleas.	Constrained
Zadvydas v. Davis, 533 U.S. 678	2001	Post-order detention, INA section 241(a)(6), 8 U.S.C. section 1231(a)(6)	Post-removal detention is authorized only for a period reasonably necessary to effect removal, with six months presumptively reasonable.	Constrained
Demore v. Kim, 538 U.S. 510	2003	Mandatory detention during proceedings, INA section 236(c), 8 U.S.C. section 1226(c)	Mandatory detention of criminal aliens during removal proceedings of ordinary duration is constitutional, with prolonged detention raising different questions.	Applied
Arizona v. United States, 567 U.S. 387	2012	Arizona S.B. 1070 enforcement provisions; federal preemption	Three state enforcement provisions are preempted, while the status-inquiry provision survives facial challenge under a narrow construction.	Constrained
Trump v. Hawaii, 585 U.S. 667	2018	Entry suspension, INA section 212(f), 8 U.S.C. section 1182(f)	Proclamation 9645 falls squarely within the delegated suspension authority and survives rational-basis review of the religion claim.	Applied
DHS v. Regents of the University of California, 591 U.S. 1	2020	Rescission of deferred action; Administrative Procedure Act review	The rescission memoranda were arbitrary and capricious for ignoring the forbearance-benefits distinction and reliance interests, without reaching the policy’s legality.	Applied
United States v. Wong Kim Ark, 169 U.S. 649	1898	Fourteenth Amendment Citizenship Clause	A child born in the United States to domiciled parents is a citizen at birth under the Citizenship Clause, a status Congress cannot legislate away.	Constrained

Federalism: Arizona and the Preemption Line

Arizona v. United States, 567 U.S. 387 (2012), decided June 25, 2012, is the federalism bookend to Plyler’s equal protection holding. In 2010 Arizona enacted S.B. 1070, the Support Our Law Enforcement and Safe Neighborhoods Act, the most aggressive state immigration enforcement statute of its generation. The United States sued to enjoin four provisions before they took effect, arguing that federal immigration law preempted them. Justice Kennedy wrote for a 5 to 3 majority, joined by Chief Justice Roberts and Justices Ginsburg, Breyer, and Sotomayor; Justice Kagan took no part, having recused herself.

The caption carries its own doctrinal information. Arizona v. United States is the rare immigration dispute in which the federal government sued a state, rather than an alien suing the government. The posture matters because it shows preemption operating as a sword for federal supremacy rather than as a shield for individual rights. The United States did not argue that Arizona’s statute violated anyone’s equal protection; it argued that the statute intruded on the federal government’s exclusive authority over immigration policy and enforcement. That framing let the Court decide the case without reaching any question about the rights of the persons the statute targeted, and it illustrates the federalism line’s distinctive character. State immigration measures can fall even when no individual claimant shows a personal constitutional injury, because the injury the Supremacy Clause redresses is to the federal structure itself.

Three provisions fell. Section 3 created a state-law crime for failure to carry an alien registration document. The Court held it preempted by field occupation: Congress had constructed a comprehensive registration scheme, and the federal framework left no room for a parallel state penalty. Section 5(C) made it a state misdemeanor for an unauthorized alien to seek or engage in work. The Court held it preempted as an obstacle to the federal design, because Congress had deliberately chosen to penalize employers rather than employees, and Arizona’s employee-side criminal penalty stood as an obstacle to that calibrated federal choice. Section 6 authorized state officers to make warrantless arrests where they had probable cause to believe the person had committed a public offense making the person removable. The Court held it preempted because it gave state officers greater arrest authority than Congress had given trained federal officers, creating an obstacle to the federal removal system Congress had constructed.

One provision survived. Section 2(B), the most publicly debated section, required state officers conducting a lawful stop, detention, or arrest to make a reasonable attempt to determine the immigration status of any person they stopped where reasonable suspicion existed that the person was unlawfully present. The Court rejected the facial challenge to this provision, and on this point all eight participating justices agreed the facial challenge failed. The majority construed the provision narrowly: it required status inquiries during otherwise lawful encounters, it did not by its terms authorize prolonged detention or independent arrest authority, and Arizona courts had not yet given it an authoritative construction that would raise constitutional problems. Because a facial challenge requires showing that the provision is unconstitutional in all its applications, and because the provision could be applied consistently with federal law, the challenge failed. As-applied challenges, based on how the provision actually operated in practice, were left open.

The narrow construction that saved Section 2(B) turned on the distinction between inquiry and detention. The provision required officers, during stops already lawful on independent grounds, to make a reasonable attempt to verify immigration status when reasonable suspicion of unlawful presence existed. It did not authorize officers to prolong a stop to await verification, to arrest on suspicion alone, or to detain persons the federal authorities declined to take into custody. The majority emphasized that Arizona courts had not yet construed the provision, and that the federal scheme itself contemplated and encouraged information sharing between state and federal officers. A provision that did no more than mandate communication during lawful encounters could be applied without conflicting with federal law, even if particular applications might later prove unconstitutional. The facial challenge therefore failed for a reason internal to facial-challenge doctrine: the challenger had to show no set of circumstances under which the provision could operate lawfully, and the Court identified several. Whether the provision as enforced in practice produced prolonged detentions or discriminatory stops was a question for as-applied litigation, which the decision expressly preserved.

Justices Scalia and Thomas dissented in part, and Justice Alito dissented in part as well, each on different grounds. Justice Scalia argued that Arizona retained a sovereign prerogative to exclude unwanted persons from its territory, a power he traced to the states’ pre-constitutional authority. Justice Thomas argued for a narrower preemption analysis that would have sustained more of the statute. Justice Alito would have sustained Section 5(C) and parts of Section 6. The partial dissents matter because they show that even within a decision striking down most of a state statute, the justices divided over how much enforcement room the states retained.

Arizona’s central argument deserves a fair statement, because it was not frivolous. The state contended that its provisions merely mirrored federal law: the registration crime tracked the federal registration requirement, the work crime tracked the federal prohibition on unauthorized employment, and the arrest provision tracked federal removal priorities. On this view, Arizona was not building a parallel regime but lending its officers to the enforcement of federal choices, and preemption should not punish cooperation. Kennedy’s majority answered that mirroring was not enough. A state criminal penalty for conduct Congress had chosen to address differently, or not to criminalize at all, was still a state penalty, and it carried the state’s independent enforcement discretion with it. Congress’s choice to penalize employers rather than employees reflected a calibrated judgment about how to structure the labor market’s incentives; Arizona’s employee-side crime upset that calibration. Congress’s allocation of arrest authority to trained federal officers reflected a judgment about how removal discretion should be exercised; Arizona’s broader arrest power displaced it. Cooperation, the majority held, meant acting within the federal design, not replicating federal prohibitions under state authority with state consequences.

Enforcement discretion is the thread that ties the three preemptions together and the concept most often missed by readers. The federal immigration system does not merely prohibit and permit; it prioritizes. With finite resources and competing goals, the executive decides which removable aliens to pursue first, which to defer, and which to leave alone, and those choices are themselves an exercise of the immigration power. A state measure that compels action where the executive has chosen inaction, or that authorizes arrests the federal scheme withholds, does not assist federal enforcement; it substitutes a state enforcement policy for the federal one. The opinion is explicit that such substitution is impermissible. Cooperation is welcome where the federal government invites it, through agreements, task forces, and information sharing, but independent state enforcement agendas are preempted. The distinction between invited cooperation and freelance enforcement is the line Arizona crossed, and it is the line future state measures will be tested against.

The provisions the United States did not challenge are as instructive as those it did. Information sharing between state and federal officers, notification requirements, and enforcement assistance consistent with federal priorities were left standing, and the opinion described them as the legitimate sphere of state involvement. The line the Court drew was thus functional rather than categorical: states could help the federal government do what it had chosen to do, but they could not use federal choices as raw material for independent state enforcement regimes. Scalia’s partial dissent rejected even that line, arguing from the states’ pre-constitutional sovereign authority that Arizona retained an inherent power to exclude unwanted persons from its territory. Thomas and Alito, in their separate partial dissents, accepted the preemption framework but would have applied it more narrowly, sustaining provisions the majority struck. The three partial dissents, read together, map the full spectrum of disagreement: from no state role, to a cooperative role, to an inherent sovereign role. The majority chose the middle position, and it has governed since.

Arizona’s statute had a declared strategy, attrition through enforcement: the idea that sufficiently rigorous state-level enforcement would make unlawful presence so difficult that affected persons would leave on their own. The strategy’s premise was that the federal government was under-enforcing its own laws and that the state could supply the missing pressure. The preemption analysis answered that premise directly. Under the Constitution’s allocation, the judgment about how much enforcement is enough belongs to the federal political branches, exercised through the plenary immigration power and the executive’s enforcement discretion. A state cannot declare federal enforcement inadequate and build its own parallel machinery to compensate, because the inadequacy determination is itself a federal policy judgment. The attrition strategy therefore failed not on its facts but on its constitutional theory: whatever the merits of more or less enforcement as policy, the decision belongs to Washington. That is the federalism lesson Arizona teaches, and it is why the decision constrains states even when their policy preferences align with stricter federal enforcement.

Arizona is the clearest modern illustration of the asymmetric doctrine. The federal government, acting under the plenary power, receives deference so broad that an entry suspension survives rational-basis review. A state, acting in the same field, faces preemption doctrines that invalidated three of four challenged provisions. The asymmetry is not a contradiction. It follows from the constitutional structure: the immigration power belongs to the national sovereign, so federal exercises of it receive the deference sovereignty commands, while state exercises of it intrude on a field Congress has occupied. Litigants understand this, which is why challenges to restrictive immigration policies are so often framed as challenges to state action even when the underlying policy preferences concern federal law.

Why did the status-inquiry provision survive when three others fell?

Because Section 2(B) added no state penalty and no independent arrest power; it required status checks during otherwise lawful stops. Narrowly construed, it could operate consistently with federal law, so the facial challenge failed, since facial invalidity requires unconstitutionality in every application. The Court left as-applied challenges open for how the provision worked in practice.

The decision also clarified the preemption taxonomy for the field. Field preemption applied where Congress had built a complete registration system. Obstacle preemption applied where the state provision frustrated a deliberate federal calibration, whether the employer-employee penalty choice or the allocation of arrest authority. What the Court did not do was hold that states may never touch immigration enforcement. Cooperative provisions, information sharing, and enforcement consistent with federal priorities survived. The line Arizona drew was between states that supplement the federal system and states that build a parallel one.

Entry, Deference, and the Later Plenary Power Cases

The two decisions that close the arc sit at opposite ends of the deference spectrum, and together they demonstrate how little the core doctrine has moved. One sustained an executive entry suspension with the broadest deference the modern Court has given. The other set aside an executive rescission on procedural grounds while expressly declining to rule on the underlying policy. Neither retreated from plenary power. Neither expanded it beyond its nineteenth century foundations.

Trump v. Hawaii, 585 U.S. 667 (2018), decided June 26, 2018, reviewed Presidential Proclamation 9645, which suspended entry of nationals from several countries following a worldwide review of information-sharing practices. The challengers argued that the proclamation exceeded the president’s statutory authority and violated the Establishment Clause. Chief Justice Roberts wrote for a 5 to 4 majority, joined by Justices Kennedy, Thomas, Alito, and Gorsuch. Justice Kennedy concurred separately. Justice Sotomayor dissented, joined by Justice Ginsburg.

The statutory holding was the heart of the decision. Section 212(f) of the Act, codified at 8 U.S.C. section 1182(f), provides that the president may suspend the entry of any aliens or class of aliens whose entry he finds would be detrimental to the interests of the United States. The majority held that Proclamation 9645 fell “squarely within” this delegation: the proclamation followed a multi-agency review, contained findings about deficient information sharing, included exceptions and waiver provisions, and was facially neutral as to religion. The breadth of the delegation was the point. Congress had given the president sweeping authority over entry suspension, and the Court would not narrow it by inference.

The statutory text that carried the decision is worth quoting in substance, because its breadth is the point. Section 212(f) provides that whenever the president finds that the entry of any aliens or of any class of aliens would be detrimental to the interests of the United States, he may by proclamation suspend their entry or impose restrictions he deems appropriate. The provision contains no list of qualifying reasons, no time limit, and no requirement beyond the presidential finding of detriment. Congress wrote it that way deliberately, as a standby authority for the executive to use when the ordinary admission machinery proved inadequate to a threat. The majority treated that breadth as the answer to the statutory challenge: a delegation this sweeping, exercised through documented findings and bounded by exceptions and waivers, was exactly what the text authorized. To narrow it by inference, the opinion suggested, would be to rewrite the delegation courts were bound to apply.

The waiver provisions mattered to the majority’s reasoning because they answered the charge that the suspension was a blanket nationality ban. The proclamation directed consular officers to consider waivers for foreign nationals who demonstrated undue hardship, whose entry was in the national interest, and who posed no security threat, with guidance on the waiver process issued to implement the provision. The majority treated the waiver system as evidence that the proclamation operated as a calibrated entry restriction rather than an absolute bar: nationals of the listed countries could still seek entry through individualized review, and the executive had built the mechanism for doing so into the order itself. The dissent questioned whether the waivers functioned meaningfully in practice, but the majority judged the order on its face, consistent with the deferential posture the plenary power doctrine prescribed. The facial approach was itself the point. In the admission context, the Court would assess what the political branches had authorized and documented, not how generously line officers exercised the discretion the order conferred.

The challengers’ second statutory argument concerned the interaction between the suspension authority and the nondiscrimination provision of the Act. Section 202(a)(1)(A), codified at 8 U.S.C. section 1152(a)(1)(A), provides that no person shall receive any preference or priority or be discriminated against in the issuance of an immigrant visa because of nationality. The challengers argued that this anti-discrimination rule limited the president’s suspension power, so that a nationality-based entry suspension violated the statute even if section 212(f) authorized suspensions generally. The majority rejected the argument on textual and structural grounds. The nondiscrimination provision governed the issuance of immigrant visas, a specific process within the admission system, while section 212(f) conferred a distinct and broader authority to suspend entry itself. The two provisions operated in different spheres, and the specific anti-discrimination rule for visa issuance did not impliedly repeal the general suspension authority. The dissent would have read the provisions together as expressing a congressional policy against nationality discrimination that the proclamation violated. The majority’s resolution preserved the full breadth of the suspension delegation, holding that Congress knew how to limit the president’s entry authority and had not done so in section 212(f).

The constitutional holding applied the most deferential review available. For the Establishment Clause claim, the majority applied rational-basis review, asking only whether the proclamation could reasonably be understood as the product of a legitimate government interest rather than religious animus. It held that the national security and information-sharing justifications satisfied that test. The dissent argued that the majority had ignored overwhelming evidence of anti-Muslim purpose in the president’s public statements and that a proper application of Establishment Clause principles required invalidating the proclamation. The disagreement was not about whether the plenary power doctrine existed; it was about how much work the doctrine could do when the challengers alleged an unconstitutional motive behind an otherwise authorized act. The majority’s answer, that the facial legitimacy of the order sufficed under rational-basis review, is the strongest modern statement of deference to the political branches on admission.

Rational basis review in the immigration context operates differently than in other fields, and the difference is the plenary power doctrine working through the tiers of scrutiny. In ordinary constitutional law, rational basis is already deferential; the government need only show a conceivable legitimate purpose and a rational connection. In immigration, the doctrine adds a further instruction: judges should not look behind the political branches’ stated justification to probe for hidden motives, because the admission judgment implicates foreign affairs and national security, areas where the judiciary lacks competence and the political branches possess unique information. The 2018 majority applied that instruction strictly, accepting the proclamation’s national security rationale without testing the challengers’ evidence of animus. The dissent argued that the ordinary tools for smoking out improper purpose, statements by officials, departures from normal procedure, historical background, should apply with full force, and that refusing to use them was abdication rather than deference. The dispute is really about whether immigration gets its own version of rational basis, more deferential than the standard model, and the majority’s answer was yes.

Department of Homeland Security v. Regents of the University of California, 591 U.S. 1 (2020), decided June 18, 2020, is the most misreported decision in the arc, and the brief for this article flags the misreporting as a recurring error worth correcting at length. The case concerned the rescission of Deferred Action for Childhood Arrivals, the 2012 executive policy under which certain persons brought to the country as children received forbearance from removal and eligibility for work authorization. In 2017 the Acting Secretary of Homeland Security issued a memorandum rescinding the policy, relying principally on a legal conclusion by the Attorney General that the policy was unlawful.

Chief Justice Roberts wrote for a 5 to 4 majority, joined in full by Justices Ginsburg, Breyer, and Kagan, and in part by Justice Sotomayor. Justices Thomas, joined by Alito and Gorsuch, and Justice Kavanaugh dissented in part. The equal protection challenge to the rescission failed by 8 to 1. The Administrative Procedure Act challenge succeeded. The Court held the rescission memoranda arbitrary and capricious on two specific grounds. First, the agency had failed to consider an obvious alternative: retaining the forbearance component, the decision not to pursue removal, while ending the benefits component, the work authorization and related eligibility. Forbearance and benefits were legally distinct, and an agency rescinding a policy must consider the significant alternatives within its discretion. Second, the agency had failed to address the reliance interests engendered by the policy: hundreds of thousands of recipients had structured their education, employment, and family lives around it, and the memoranda did not weigh those interests at all.

The opinion then said, in language worth quoting because it defines the decision’s limits: “We do not decide whether DACA or its rescission are sound policies.” And: the question was “not whether DHS may rescind DACA. All parties agree that it may.” The Court remanded so the agency could provide a reasoned explanation addressing the required considerations. The holding was procedural in the fullest sense. It decided nothing about whether deferred action was lawful, nothing about whether the rescission was substantively justified, and nothing about the plenary power doctrine’s application to the underlying policy. To describe the decision as upholding the policy’s legality, as much public commentary did, is to misstate what the Court decided. The decision upheld only the proposition that agencies must reason through the alternatives and reliance interests before acting, a requirement of administrative law that applies identically in every field.

The arbitrary-and-capricious standard that decided the case comes from the Administrative Procedure Act and the Court’s gloss on it, and its operation here illustrates why procedure proved dispositive. An agency action is arbitrary and capricious when the agency fails to consider an important aspect of the problem, a formulation associated with Motor Vehicle Manufacturers Association v. State Farm. The rescission memoranda failed that test twice. First, deferred action as practiced had two legally distinct components: forbearance, the discretionary decision not to pursue removal against a particular class, which flowed from enforcement discretion; and benefits, the work authorization and related eligibility that flowed from separate regulatory provisions. An agency ending the policy could have retained forbearance while ending benefits, or explained why the two stood or fell together. The memoranda did neither, treating the policy as an indivisible whole and thereby ignoring an obvious alternative within the agency’s discretion. Second, the memoranda did not address the reliance interests of the policy’s recipients, who had enrolled in schools, taken jobs, obtained mortgages, and raised families in reliance on the government’s forbearance. Reliance interests do not make an agency action unreviewable or irreversible; they require the agency to weigh the costs of disruption against the reasons for change, on the record, before acting.

The reliance analysis in particular has become a template for challenging policy reversals across the administrative state. The opinion treats reliance not as a trump that freezes policy forever but as a cost the agency must acknowledge and weigh. An agency may change course, even drastically, but it must show its work: identify who relied, assess how heavily, and explain why the change is justified despite the cost. That requirement disciplines reversals without forbidding them, and it explains the opinion’s careful statement that the Department could rescind the policy if it supplied adequate reasoning. The remand was an instruction to reason, not a command to retain. Later agencies have internalized the lesson, building reliance analyses into rescission memoranda as a matter of course, and later litigants have tested whether those analyses are adequate. The procedural turn thus generates its own litigation, review of the reasoning rather than review of the policy, and that litigation is where the modern field increasingly lives.

The Court’s express reservations defined what the decision left for another day. By stating that it did not decide whether DACA or its rescission were sound policies, and that all parties agreed DHS could rescind DACA, the majority insulated the holding from the substantive debate entirely. The remand instructed the agency to supply the missing reasoning, not to reach a particular result. A later, better-reasoned rescission that addressed the forbearance-benefits distinction and weighed reliance interests could survive the very review that invalidated this one. That is the nature of procedural review: it disciplines how the government decides, not what it decides. For the plenary power story, the case’s placement is therefore precise. It belongs in the arc not as a limit on the immigration power but as proof that the modern Court’s meaningful interventions in this field run through procedure, through the APA, through canons of interpretation, through federalism, while the substantive core of plenary power remains where the nineteenth century put it.

The equal protection ruling, rejecting the challenge by 8 to 1, deserves its own note because it shows the limits of motive-based claims in this field. The challengers had argued that the rescission was motivated by animus toward persons of Latino descent. Only Justice Sotomayor found the allegations sufficient to proceed; the other eight justices held that the pleaded facts did not support an inference of discriminatory purpose. The lopsided vote is significant. In a decision otherwise divided 5 to 4 along familiar lines, the near-unanimous rejection of the equal protection claim demonstrated that even justices sympathetic to the APA challenge would not constitutionalize the dispute. The message for future litigants was clear: procedural claims under the APA could succeed where substantive constitutional claims about motive would not, and the doctrine’s deference extended to protecting executive immigration decisions from purpose-based invalidation absent extraordinary proof.

The Complication: Why These Decisions Are Not a Break

The assumption this article must address is that the later decisions represent a break with the plenary power tradition, a judicial turn toward subjecting immigration to ordinary constitutional review. The record does not support it. Deference to the political branches on entry and admission has been the consistent frame for well over a century, and the meaningful developments of the last four decades are better described as procedural review and state preemption than as a retreat from plenary power.

Consider what the supposedly transformative decisions actually did. Plyler constrained states, not the federal government, and left Congress’s admission power untouched. Chadha constrained congressional procedure, not immigration substance, and its holding applied identically in every field. Cardoza-Fonseca enforced Congress’s textual choices against an agency’s narrowing construction, which is deference to the legislature, not defiance of it. St. Cyr applied clear-statement canons to habeas and retroactivity, expressly acknowledging that clear congressional language would have produced the opposite result. Zadvydas read a limit into a statute through constitutional avoidance rather than striking the statute down, and it distinguished rather than overruled Mezei. Demore sustained mandatory detention. Arizona preempted state law while reaffirming federal primacy. Trump v. Hawaii applied the doctrine at full strength. DHS v. Regents decided an APA case that could have arisen under any statute.

What a genuine break from the plenary power tradition would look like is worth stating, because it clarifies how far the later decisions are from one. A break would require the Court to hold that admission and expulsion decisions are subject to ordinary constitutional review, that Congress’s immigration classifications must satisfy the same scrutiny as its other classifications, or that the executive’s entry judgments are reviewable for substantive reasonableness rather than procedural regularity. No decision in the arc does any of those things. The 2018 decision applied rational basis, the most deferential review available, and declined to probe the stated justification. The 2020 decision reviewed procedure under a statute that applies to every agency and said nothing about the immigration power. The 2012 decision bound the states, not the political branches. The 2001 decisions read statutes. A break would be visible in the standard of review applied to a federal admission judgment, and the standard has not moved since 1889. Until it does, the doctrine stands, and the smart litigation targets the surroundings rather than the wall.

The temptation to read the later decisions as a break has understandable sources. The 2001 term alone produced two government losses in St. Cyr and Zadvydas, decided three days apart, and the political salience of the entry-suspension and deferred-action disputes made the later cases feel like turning points. Counting losses, however, is not the same as measuring doctrine. A government can lose individual cases for many reasons, defective reasoning in a memo, an overreading of ambiguous text, a state statute that overshoots the federal design, without the underlying allocation of power shifting at all. The question is never whether the government lost. It is whether the Court questioned the political branches’ authority to decide the underlying question. On that measure, the record is unbroken from 1889 forward.

The pattern is not erosion but channeling. The Court has not abandoned the proposition that the political branches hold plenary authority over admission and removal. It has specified, at the margins, what that authority does not include: it does not include rewriting clear statutory standards, it does not include implied repeals of habeas, it does not include retroactive destruction of plea reliance without clear statement, it does not include indefinite post-order detention untethered from foreseeable removal, it does not include state enforcement schemes that duplicate or contradict the federal design, and it does not include agency action that ignores obvious alternatives and reliance interests. Each of those limits is real, and each was hard-fought. None of them questions the core.

The interpretive presumptions deserve recognition as a body of law in their own right, because together they amount to the field’s real bill of rights. The presumption against retroactivity protects settled transactions, most powerfully the plea bargains entered in reliance on relief Congress later repeals. The habeas presumption protects access to a judge, refusing to read the writ’s repeal into ambiguous jurisdiction-stripping text. The constitutional avoidance canon protects against purposeless confinement, reading time limits into detention statutes rather than deciding the due process question. None of these presumptions questions the plenary power; each of them assumes it, and then asks what background law requires before the power is exercised in a particular way. The result is a layered system: plenary power at the foundation, presumptions in the middle, and individual outcomes determined by the interaction of the two. A reader who knows only the foundation will predict deference everywhere and be surprised by the limits; a reader who knows only the presumptions will predict limits everywhere and be surprised by the deference. The field requires both.

The voting pattern across the arc tells its own story about where the Court divides. The interpretation and detention cases of 2001, the mandatory detention case of 2003, the entry suspension case of 2018, and the deferred action case of 2020 were all decided five to four, while the 1889 foundation was unanimous and the 2012 preemption case was five to three with a recusal. The close divisions cluster around the questions the doctrine leaves open: how to read ambiguous statutes, how long detention may last, how much procedure the APA demands. The questions the doctrine settles, that exclusion is a sovereign act, that states are bound by preemption, that the Citizenship Clause means what it says, command broader agreement. The pattern confirms the central claim. The core is settled and the margins are contested, and the five to four decisions are the sound of the Court working out the margins.

Walk the arc era by era and the continuity becomes visible. The nineteenth century announced plenary power in its strongest form, sovereignty-sourced and judicially unreviewable. The mid-century extended it to executive action at the border without softening it. The 1980s constrained states and congressional procedure while leaving federal substance untouched. The turn-of-the-century decisions imposed interpretive discipline on agencies and temporal limits on detention while distinguishing, not overruling, the deference precedents. The 2012 federalism decision reaffirmed federal primacy even as it struck down state law. The 2018 entry decision applied the doctrine at full strength to a presidential proclamation. The 2020 decision policed agency reasoning without touching the underlying power. In no era did the Court hold that Congress lacked plenary authority over admission and removal, or that the executive’s admission judgments were subject to ordinary judicial review. The doctrine’s content changed at the margins in every generation. Its center never moved.

The complication thus resolves into a research strategy. The reader who wants to predict the next immigration decision should not ask whether the plenary power doctrine is rising or falling; it is doing neither. The reader should ask three questions instead. First, who is the defendant: the political branches, where deference runs high, or a state, where it does not. Second, what is the claim: a substantive attack on the admission judgment, which the doctrine deflects, or a procedural attack on how the judgment was reached, which administrative law welcomes. Third, what does the statute say: the Court reads immigration text with the full apparatus of modern interpretation, clear statement rules included, and agencies that conflate distinct standards or repeal relief without plain language will lose on the words. Those three questions price any immigration dispute more reliably than any narrative of doctrinal rise or decline.

A worked example will show the three-question strategy in action. Suppose the executive announces a new policy narrowing asylum eligibility, and advocates ask how to challenge it. The first question, who is the defendant, directs attention away from Congress, whose power to define eligibility categories is plenary, and toward the executive’s implementation, where procedure and interpretation apply. The second question, what is the claim, rules out a substantive attack on the policy’s wisdom, which the doctrine deflects, and points toward the statute’s text, did the agency read the eligibility provisions correctly, toward the APA, did the memorandum consider alternatives and reliance, and toward retroactivity, does the policy reach back to pending applications without a clear statement. The third question, what does the statute say, sends the litigant to the actual provisions, the well-founded fear standard, the withholding definitions, the jurisdiction channeling rules, to find the words the agency may have misread. That sequence, defendant, claim, text, prices the dispute before a complaint is drafted, and it works because the doctrine’s shape is stable enough to predict.

The persistence of the doctrine also explains a feature of the field that puzzles newcomers: why the most heated public controversies over immigration produce the least doctrinal movement. Public debate centers on substance, who should be admitted, who should be removed, how enforcement should prioritize. Doctrine centers on everything but substance when the defendant is federal. The Court’s answers to the public’s questions are therefore oblique: not whether the policy is wise but whether the procedure was adequate, not whether the classification is just but whether the statute was read correctly, not whether the state should help enforce but whether the Constitution permits it to. The obliqueness is not evasion; it is the doctrine working as designed, channeling disputes into the forms judges can resolve without second-guessing sovereign judgments. The litigant who understands the channeling wins the cases that can be won; the litigant who fights the channeling loses even sympathetic claims.

Litigation as the stage where a statute’s breadth is priced, in the field where the price is lowest. That is why challengers pay the most and recover the least when they attack the federal government’s admission and removal decisions directly, because the plenary power doctrine discounts their claims before the merits are reached. The rational litigant therefore aims elsewhere, at state statutes where preemption and equal protection apply in full, at agency interpretations where textual canons bite, at procedures where the APA demands reasoned explanation. The doctrine does not end litigation. It reroutes it.

The Asymmetric Doctrine

The asymmetric doctrine: plenary power constrains courts reviewing federal immigration decisions far more than it constrains courts reviewing state ones, which is why the same field produces sweeping deference to Congress and firm limits on states, and why litigants choose their target accordingly.

That single sentence organizes the entire arc. Chae Chan Ping and Fong Yue Ting built the federal side: sovereignty-sourced, judicially unreviewable in substance, capable of overriding treaties. Plyler and Arizona built the state side: equal protection and preemption as firm boundaries on state action in the same field. Chadha, Cardoza-Fonseca, St. Cyr, Zadvydas, Demore, Trump v. Hawaii, and DHS v. Regents filled in the interior, distinguishing substance from procedure, admission from deportation, exclusion from removal, federal action from state action, and merits from process. The doctrine that emerges is not a wall but a prism, bending identical constitutional arguments into different outcomes depending on which government acted and at which stage.

The prism metaphor carries a practical consequence that the brief for this article identifies as the reason litigants choose their targets. A challenger who attacks a federal admission decision head-on faces the full weight of plenary power and almost always loses. The same challenger, reframing the dispute as a challenge to a state enforcement measure, an agency’s departure from statutory text, or a procedure that skipped required reasoning, faces ordinary judicial review and can win. Nothing about the underlying policy preference changes between the two framings. What changes is the doctrinal shelf the case sits on, and the shelf determines the standard of review, which in this field determines the outcome more reliably than the merits. Understanding the asymmetry is therefore not an academic exercise. It is the single most useful thing a reader can know about immigration litigation, because it predicts where challenges succeed and where they fail before a brief is ever filed.

Where the Cluster Goes Next

A reader who has worked through the arc holds the key to the rest of the immigration cluster. The provisions most of the modern cases construed, the detention sections, the relief provisions, the removal machinery, belong to the 1996 amendments, and the companion guide to that statute supplies the text the decisions interpret. The asylum standard of Cardoza-Fonseca belongs to the 1980 humanitarian framework, and its companion guide supplies the legislative history the Court relied upon. The statutory framework underlying every decision here belongs to the 1965 Act as amended, and its companion guide supplies the architecture. The agency captions that change mid-arc belong to the 2002 reorganization, and its companion guide supplies the institutional map. Each specialist article owns its subject; this article owns the question of what courts have done with them all.

For the practitioner, the doctrine is less a philosophy than a map of where arguments can win. Challenges to the substance of federal admission and removal decisions are the hardest cases in public law, and the experienced litigator does not file them expecting deference to yield. Instead the litigator aims at the margins the Court has marked: the agency interpretation that departs from Congress’s text, where Cardoza-Fonseca supplies the precedent; the jurisdiction-stripping provision without a clear statement, where St. Cyr supplies the canon; the detention that has outlasted its justification, where Zadvydas supplies the framework and Kennedy’s Demore concurrence supplies the warning; the state statute that duplicates or contradicts the federal scheme, where Arizona supplies the preemption analysis; the agency action that skipped the required reasoning, where Regents supplies the APA template. Each of these is a real and winnable claim. None of them asks the Court to abandon plenary power. The most successful immigration litigators of the last four decades have been those who understood that the doctrine is not an obstacle to be denounced but a terrain to be navigated, and who chose their targets accordingly.

A final checklist will let the reader test mastery. For each decision in the ledger, the reader should be able to state the vote and the author, the provision or power at issue, the holding in one sentence, the strongest argument on each side, and the plenary power effect. The vote and author matter because they reveal the coalition: unanimous in 1889, closely divided in the modern cases, with concurrences doing decisive work. The provision matters because the field’s questions recur around the same statutory sections, and recognizing the section predicts the reasoning. The holding matters because it is what governs. The arguments matter because the dissent of one era may become the majority of the next. The effect matters because it maps the doctrine’s shape, expansion for political-branch admission decisions, constraint for states and for procedure and text, application where the Court reads without touching the power. A reader who can fill all five columns from memory, for all thirteen rows, has the field.

The deeper test is predictive. Presented with a new immigration dispute, the reader should be able to price it: identify the defendant and the deference level that follows, identify the claim type and whether the doctrine deflects or welcomes it, and read the statute for the words that will decide the case. The worked example in the complication section showed the method; the reader should now be able to run it on any fact pattern. That predictive capacity is the article’s real product. Information about past decisions depreciates as new ones arrive; a method for reading new decisions appreciates, because the doctrine’s stability makes the method durable. The wall stands, the gates are marked, and the reader now holds both the map and the compass.

The assessment this article offers is deliberately narrow, matching the One Test it was built to pass. The doctrine that makes immigration law constitutionally unlike every other field in this series is the plenary power doctrine, sourced in sovereignty rather than in an enumerated clause, announced in 1889 and 1893, extended to the executive at the border at mid-century, and never abandoned. It has been softened at the edges: by equal protection limits on states, by preemption of state enforcement schemes, by textual canons that bind agencies to Congress’s enacted standards, by clear-statement rules protecting habeas and reliance interests, by constitutional avoidance limiting post-order detention, and by administrative law requiring reasoned agency action. The softening is real and litigable, and it is where the field’s practitioners earn their fees. But the center holds. When the political branches decide who may enter and who must leave, courts still defer as they defer nowhere else, and every honest account of the field must begin from that fact rather than from the exceptions.

Return, finally, to the One Test. The reader who has followed the arc should be able to explain, in plain terms, the doctrine that sets immigration apart: that the power over admission and removal flows from national sovereignty rather than from an enumerated clause, that courts have treated its exercise by the political branches as largely immune from ordinary constitutional review since 1889, and that the immunity covers Congress when it legislates and the executive when it acts at the border. That reader should be able to trace the softening without mistaking it for abandonment: Plyler’s limit on states, Chadha’s limit on congressional procedure, Cardoza-Fonseca’s enforcement of Congress’s textual choices, St. Cyr’s protection of habeas and plea reliance, Zadvydas’s temporal limit on post-order detention, Arizona’s preemption of parallel state regimes, and Regents’ demand for reasoned agency action. And that reader should be able to state the holdings that govern: the ledger above supplies them in one-sentence form, with the vote, the provision, and the plenary power effect attached. Doctrine explained, softening traced, holdings stated. That is what this article promised, and the rest of the cluster builds on it.

Studying the Doctrine Further

A reader working through these opinions benefits from keeping a running ledger of holdings, vote counts, and doctrinal effects, and from testing recall against the questions below. The VaultBook legislation study notebook is built for exactly that kind of work: saving and annotating long guides, building case chronologies, and keeping citation lists in one place. The doctrine ledger above supplies the skeleton; the study notebook supplies the room to flesh it out.

The most effective way to study these opinions is to brief each one in the four-part structure this article uses: holding, reasoning, dissent, effect. Write the holding as a single sentence that a non-lawyer could understand. Write the reasoning as the chain of steps that gets from the provision to the result, noting which interpretive tools the majority used and in what order. Write the dissent as the strongest version of the losing argument, not a caricature, because the next case may adopt it. Write the effect as one of the ledger’s three categories, and be prepared to defend the choice. Ten minutes per case, thirteen cases, and the field’s architecture will be committed to memory more durably than any amount of rereading.

The questions in the FAQ section below are designed as a self-test keyed to that method. Each answer restates the holding and the reasoning in self-contained form, so a reader can cover the answers, attempt the questions, and check recall. The H3 questions in the body serve the same function at a higher level of generality, testing whether the reader has absorbed the doctrine’s structure rather than just its details. Used together, the ledger for the skeleton, the body for the reasoning, and the questions for testing, they form a complete study system for the field.

Three recurring errors deserve explicit warning, because they appear in almost every first attempt to master the field. The first is treating DHS v. Regents as a ruling that the deferred action policy was lawful or that its rescission was substantively impermissible. The opinion decided neither, and the distinction between procedural invalidity and substantive protection is the most tested point in the modern arc. The second is treating St. Cyr as a holding that the 1996 legislation’s aggravated-felony expansion cannot apply retroactively. The opinion expressly acknowledged the opposite for the expansion itself, limiting its retroactivity holding to the repeal of discretionary relief as applied to plea bargains. The third is treating the plenary power doctrine as a grant of unlimited authority to do anything to anyone. The doctrine concerns admission and expulsion by the political branches; it does not authorize states to act, does not displace constitutional text, and does not excuse procedural failures. Each error comes from overreading a holding in a favored direction, and each is corrected by returning to the opinions’ actual words.

Frequently Asked Questions

Q: What is the plenary power doctrine in immigration law?

The plenary power doctrine is the Supreme Court’s holding that Congress and the executive possess essentially complete authority over immigration, subject to only minimal judicial review. It originates in Chae Chan Ping v. United States (1889) and Fong Yue Ting v. United States (1893), which located the power to exclude and expel aliens in national sovereignty rather than in any enumerated constitutional clause. Because the power flows from sovereignty itself, the Court has treated immigration decisions as political judgments committed to the political branches, not as ordinary exercises of regulatory authority subject to searching judicial scrutiny. The doctrine does not mean courts never intervene: later decisions imposed limits through equal protection, preemption, statutory interpretation canons, and administrative procedure. But those limits operate at the margins. On the core questions of who may enter and who must leave, the doctrine remains the reason immigration law is constitutionally unlike every other field.

Q: What did Plyler v. Doe decide about immigration law?

Plyler v. Doe, 457 U.S. 202 (1982), held 5 to 4 that Texas could not deny free public kindergarten through twelfth grade education to children present in the United States unlawfully. Justice Brennan’s majority opinion rested on equal protection: the children bore no responsibility for their parents’ decisions, education imposed unique lifelong consequences when denied, and Texas could not show that excluding these children served a substantial state interest. The decision is frequently misdescribed as limiting federal immigration power. It did nothing of the sort. It constrained a state, not Congress or the president, and its reasoning applied the equal protection and federalism principles that govern state action. Plyler illustrates the field’s asymmetry: courts defer sweepingly to federal immigration decisions while scrutinizing state ones, which is why the decision belongs to the state-side half of the doctrine.

Q: What did INS v. Chadha decide about immigration law?

INS v. Chadha, 462 U.S. 919 (1983), invalidated the legislative veto, the device by which Congress reserved the power to overturn executive actions by resolution of one or both Houses without presentment to the president. The case arose when the House of Representatives vetoed the Attorney General’s suspension of Jagdish Chadha’s deportation under section 244 of the Immigration and Nationality Act. By a 7 to 2 judgment, the Court held that a one-House resolution altering legal rights was legislation subject to the Constitution’s bicameralism and presentment requirements. Although the case began as an immigration dispute, its holding applied across the entire United States Code, invalidating similar veto provisions in nearly two hundred statutes. For immigration law specifically, Chadha confirmed that the plenary power doctrine constrains substance but not constitutional procedure: Congress holds vast power over admission and removal, but it must exercise legislative power through Article I procedures.

Q: What did Arizona v. United States hold about state immigration law?

Arizona v. United States, 567 U.S. 387 (2012), held that federal immigration law preempted most of Arizona’s S.B. 1070 enforcement statute while allowing one provision to survive a facial challenge. By 5 to 3, with Justice Kagan recused, the Court struck down three sections: the state alien-registration crime, preempted because Congress had occupied the registration field; the state crime for unauthorized work, preempted as an obstacle to Congress’s deliberate choice to penalize employers rather than employees; and the warrantless-arrest authorization, preempted for giving state officers broader arrest power than federal officers held. Section 2(B), requiring status inquiries during lawful stops, survived the facial challenge because, narrowly construed, it could be applied consistently with federal law; all eight participating justices agreed the facial challenge failed, while as-applied challenges remained open. The decision drew the modern line between state cooperation with federal enforcement and parallel state regimes.

Q: Can immigrants be detained indefinitely under immigration law?

The answer depends on the stage of the case. After a final order of removal, indefinite detention is not permitted: Zadvydas v. Davis (2001) read section 241(a)(6) of the Immigration and Nationality Act to authorize detention only for a period reasonably necessary to effect removal, with six months as the presumptively reasonable period, after which an alien showing no significant likelihood of removal in the reasonably foreseeable future must be released unless the government rebuts the showing. During removal proceedings, however, Demore v. Kim (2003) sustained mandatory detention of criminal aliens under section 236(c), emphasizing its limited duration and its purposes of preventing flight and protecting the public, with Justice Kennedy’s deciding concurrence warning that prolonged detention would raise different questions. At the border, in the exclusion context, Shaughnessy v. Mezei (1953) permitted extended detention incident to an unreviewable exclusion decision. Duration, stage, and context decide each case.

Q: Is birthright citizenship in immigration law or the Constitution?

Birthright citizenship rests in the Constitution, not in immigration statutes. United States v. Wong Kim Ark, 169 U.S. 649 (1898), held 6 to 2 that the Fourteenth Amendment’s Citizenship Clause, which provides that all persons born or naturalized in the United States and subject to its jurisdiction are citizens, conferred citizenship on a man born in San Francisco to Chinese parents who were themselves ineligible for naturalization. Justice Gray’s majority opinion traced the clause to the English common-law rule of birth within the sovereign’s dominion, with narrow exceptions for children of diplomats, hostile occupying forces, and at that time certain tribal relations. Because the right flows from the Constitution itself, Congress cannot legislate it away through the immigration code; a statute purporting to deny citizenship to persons the clause covers would be unconstitutional. The decision is therefore a boundary marker for the plenary power doctrine: the immigration power, however broad, stops where the Constitution’s own grants of citizenship begin.

Q: What did Trump v. Hawaii decide about immigration law?

Trump v. Hawaii, 585 U.S. 667 (2018), upheld Presidential Proclamation 9645, which suspended entry of nationals from several countries after a worldwide review of information-sharing practices. By 5 to 4, the Court held that the proclamation fell squarely within section 212(f) of the Immigration and Nationality Act, which authorizes the president to suspend entry of any class of aliens whose entry he finds detrimental to United States interests. The majority emphasized the breadth of that delegation and the proclamation’s findings, exceptions, and waiver provisions. On the Establishment Clause challenge, the majority applied rational-basis review and held that the national security and information-sharing justifications supplied a legitimate purpose. Justice Sotomayor’s dissent argued that the majority had disregarded evidence of anti-Muslim animus. The decision is the strongest modern application of the plenary power doctrine to executive action on admission, confirming that courts review entry suspensions with extreme deference.

Q: Why do courts defer to Congress on immigration law?

Courts defer because the Supreme Court has located the immigration power in national sovereignty rather than in any enumerated clause, beginning with Chae Chan Ping v. United States (1889) and Fong Yue Ting v. United States (1893). The reasoning is that deciding who may enter and remain in the national community is an attribute of nationhood itself, tied to foreign relations, security, and self-definition, and therefore a political judgment committed to the political branches rather than a legal question for judicial balancing. The Court extended that deference to executive determinations at the border in the mid-century exclusion cases. Later decisions softened the doctrine at the edges through equal protection, preemption, interpretive canons, and administrative procedure, but none repudiated the core. The deference also reflects institutional competence: judges lack the tools to second-guess the foreign policy and security judgments embedded in admission decisions. That is why challenges to federal immigration decisions face the steepest climb in American public law.

Q: Why did Chae Chan Ping ground the exclusion power in sovereignty rather than in an enumerated clause?

Justice Field’s unanimous opinion in Chae Chan Ping v. United States, 130 U.S. 581 (1889), faced a genuine textual problem: no clause of the Constitution expressly grants Congress the power to exclude aliens. The Commerce Clause was an awkward fit, the Naturalization Clause governed citizenship rather than admission, and the war powers did not cover peacetime exclusion. Rather than stretch an enumerated grant beyond recognition, Field located the power in the sovereignty of the nation itself, reasoning that every independent nation possesses as an incident of sovereignty the authority to determine who may enter its territory. The choice had two lasting consequences. First, it freed the immigration power from the limits that attach to enumerated powers, producing the plenary power doctrine. Second, combined with the holding that treaties and statutes stand on equal footing, it meant a later exclusion statute could override an earlier treaty, which is how the Scott Act of 1888 defeated the Burlingame Treaty of 1868.

Q: How did Zadvydas and Demore reach different results on immigration detention?

Zadvydas v. Davis (2001) and Demore v. Kim (2003) addressed detention at different stages of the process, which explains their different outcomes. Zadvydas concerned post-order detention: the alien had already been ordered removed, and the question was how long the government could hold someone when removal was not foreseeable. Through constitutional avoidance, the Court read a reasonable-period limit into the statute, with six months presumptively reasonable. Demore concerned detention during proceedings: the alien’s removability was still being adjudicated, and the detention served the purposes of ensuring appearance and protecting the public for the limited duration of the case. The liberty interests differ accordingly. After a final order with no foreseeable removal, detention becomes untethered from any legitimate purpose. During proceedings, it remains tied to the adjudication. Justice Kennedy’s concurrence in Demore added the bridge between them: prolonged detention during proceedings would raise the constitutional questions Zadvydas answered for the post-order context.

Q: What did INS v. St. Cyr decide about habeas review and retroactive application of immigration law?

INS v. St. Cyr, 533 U.S. 289 (2001), decided 5 to 4, made two holdings about the 1996 amendments to the immigration laws. First, the Court held that the amendments had not repealed habeas corpus jurisdiction under 28 U.S.C. section 2241, because repealing the historic writ requires a clear statement from Congress, and the statutes lacked one; the Suspension Clause concerns behind any such repeal demanded explicit language. Second, the Court held that the repeal of section 212(c) discretionary relief could not apply retroactively to lawful permanent residents who had pleaded guilty before the repeal, because guilty pleas involve reliance: defendants surrender constitutional rights in exchange for the law’s benefits as they stood, and undoing that bargain attaches new consequences to completed conduct under the Landgraf framework. A precision matters here: the decision did not invalidate retroactive application of the expanded aggravated-felony definition, whose retroactivity section 321(b) of IIRIRA stated clearly. St. Cyr limited retroactivity where reliance was strongest, not everywhere.

Q: How did Cardoza-Fonseca change the standard for asylum?

INS v. Cardoza-Fonseca, 480 U.S. 421 (1987), restored the independent meaning of the asylum standard that the Board of Immigration Appeals had collapsed into the stricter withholding standard. The Refugee Act of 1980 created two protections: asylum under section 208, requiring a well-founded fear of persecution, and withholding of deportation under section 243(h), requiring a clear probability of persecution, meaning more likely than not. The Board had applied the clear-probability test to both. By a 6 to 3 judgment, the Court held that Congress chose different words deliberately and meant different things: well-founded fear, drawn from the United Nations refugee protocol the Act implemented, is the more generous standard, and even a modest probability of persecution can support it. The decision did not create a right to asylum; it clarified the evidentiary threshold, leaving the Attorney General’s discretion intact. Its doctrinal lesson is that plenary power deference runs to Congress’s enacted text, not to an agency’s narrowing rewrite of it.

Q: Did DHS v. Regents uphold the DACA policy on its merits?

No. Department of Homeland Security v. Regents of the University of California, 591 U.S. 1 (2020), decided nothing about whether Deferred Action for Childhood Arrivals was lawful or whether its rescission was substantively justified. The 5 to 4 majority held the rescission memoranda arbitrary and capricious under the Administrative Procedure Act on two procedural grounds: the agency failed to consider retaining forbearance from removal while ending benefits such as work authorization, and it failed to address the reliance interests of hundreds of thousands of recipients. The opinion expressly stated that the Court was not deciding whether DACA or its rescission were sound policies, and that the question was not whether DHS may rescind DACA, which all parties agreed it could. The case was remanded for a reasoned decision. Describing the ruling as a vindication of the policy’s legality is the most common error in public discussion of the decision; it was a ruling about how agencies must reason, applicable identically in every field of administrative law.

Q: What was the legislative veto that INS v. Chadha struck down?

The legislative veto was a device Congress wrote into hundreds of statutory provisions, reserving for one or both Houses the power to nullify executive actions by resolution without presentment to the president. In the immigration context, section 244(c)(2) of the Immigration and Nationality Act allowed either House to overturn the Attorney General’s suspension of an individual’s deportation, which the House did to Jagdish Chadha without debate or stated reasons. The Supreme Court held 7 to 2 in INS v. Chadha (1983) that such a veto was unconstitutional: a resolution with the purpose and effect of altering legal rights is legislation, and the Constitution prescribes a single exclusive legislative procedure, passage by both Houses and presentment to the president. Because the defect was procedural rather than substantive, the reasoning invalidated similar veto provisions across the United States Code, affecting nearly two hundred statutes. Congress retained every substantive immigration power; it lost only the shortcut procedure for controlling executive discretion.

Q: Why did INS v. Chadha invalidate the legislative veto across the entire Code?

The reasoning that killed the particular veto could not be confined to immigration because it rested on the Constitution’s lawmaking requirements, not on anything specific to deportation. The majority held that the one-House disapproval of a deportation suspension was legislation in substance, since it altered legal rights outside the legislative branch, and the Constitution permits legislation only through bicameral passage and presentment to the President. That rule applies to every exercise of legislative power, so any veto device sharing the same defect fell with it. Congress had written some version of the veto into roughly two hundred provisions across nearly two hundred statutes, covering war powers, trade, rulemaking, and energy, and the logic of the opinion left none of them standing. A hardship suspension case thus became the vehicle for a structural holding of government-wide scope.

Q: When are state immigration enforcement laws preempted by federal law?

State immigration enforcement laws are preempted when they intrude on a field Congress has occupied or stand as an obstacle to the federal scheme, as Arizona v. United States (2012) illustrates. Field preemption struck down Arizona’s alien-registration crime because Congress had constructed a comprehensive federal registration system leaving no room for parallel state penalties. Obstacle preemption struck down the state work crime because Congress deliberately chose to penalize employers rather than employees, and a state employee-side penalty frustrated that calibration. It also struck down the warrantless-arrest provision for giving state officers broader arrest authority than Congress gave federal officers. What survives is cooperation: provisions that assist federal enforcement without adding state penalties or independent arrest power, like the status-inquiry requirement the Court narrowly construed, can withstand facial challenge. The line is between supplementing the federal system and building a parallel one. States may help enforce federal choices; they may not rewrite them.

Q: What entry suspension authority did Trump v. Hawaii confirm?

Trump v. Hawaii, 585 U.S. 667 (2018), confirmed the breadth of the president’s delegated authority under section 212(f) of the Immigration and Nationality Act, codified at 8 U.S.C. section 1182(f). That provision authorizes the president to suspend the entry of any aliens or class of aliens whenever he finds their entry detrimental to United States interests. The 5 to 4 majority held that Presidential Proclamation 9645, which suspended entry of nationals from several countries following a multi-agency review of information-sharing deficiencies, fell squarely within the delegation, noting its findings, exceptions, and waiver provisions. The Court declined to narrow the broadly worded grant by inference. On the Establishment Clause claim, the majority applied rational-basis review, holding that the stated national security justifications supplied a legitimate purpose. The decision did not hold that the president’s entry authority is unlimited; it held that this exercise, grounded in an express and sweeping congressional delegation with documented findings, survived both statutory and constitutional challenge under the deferential standards the plenary power doctrine supplies.

Q: What is the difference between exclusion and deportation in Supreme Court doctrine?

Exclusion concerns aliens stopped at the border seeking initial admission; deportation, which the 1996 amendments renamed removal, concerns aliens already admitted or present who the government seeks to expel. The distinction carries constitutional weight. In the exclusion context, the plenary power doctrine runs at maximum: Shaughnessy v. Mezei (1953) held exclusion largely immune from judicial control, treating even a long-detained entrant as stopped at the threshold. In the deportation context, the alien’s presence within the territory triggers greater procedural protection, which is why Zadvydas v. Davis (2001) carefully distinguished Mezei when limiting post-order detention. The nineteenth century cases complicate the picture: Fong Yue Ting (1893) held expulsion as absolute as exclusion, yet later doctrine developed the entry fiction and the presence-based liberty interest into a genuine divide. Practically, the distinction determines which precedents govern: border cases invoke Mezei’s deference, while removal and detention cases invoke the calibrated limits of Zadvydas and Demore.

Q: How does the Court tell a statutory interpretation case from a constitutional one in this field?

The Court asks what the dispute is really about: the meaning of words Congress wrote, or the limits the Constitution imposes. Cardoza-Fonseca and St. Cyr are statutory cases: the first construed the asylum and withholding standards, the second applied the presumptions against retroactivity and in favor of habeas to ambiguous text. Chadha and Arizona are constitutional cases: the first enforced bicameralism and presentment, the second applied Supremacy Clause preemption. Zadvydas sits at the boundary, using constitutional avoidance to read a limit into a statute without deciding the constitutional question. The distinction matters because the plenary power doctrine deflects constitutional attacks on federal immigration judgments while leaving statutory attacks fully open. An agency that misreads its statute loses on the words; the political branches acting within clear statutory text almost never lose at all.

Q: Why do immigration litigants aim their challenges at states rather than at Congress?

Litigants aim at states because the plenary power doctrine makes challenges to federal action nearly unwinnable while leaving state action exposed to ordinary review. A constitutional attack on a federal exclusion or expulsion faces deference at its highest: the political branches decide admission as an incident of sovereignty, and judges stand back. The same attack on a state measure faces the full apparatus of constitutional law: equal protection, as in Plyler, and preemption, as in Arizona, apply to states without any plenary power discount. Procedure offers a third target, since the Administrative Procedure Act and the habeas and retroactivity presumptions bind the executive regardless of the field. Rational litigants therefore challenge state measures on the merits, federal measures on the procedure, and almost never challenge the federal power itself.