The Americans with Disabilities Act reached the Supreme Court within a decade of its signing, and the resulting decisions did more than resolve individual disputes. They narrowed who counted as disabled until Congress overrode the definition, split the question of state immunity along title and constitutional lines, and turned a nondiscrimination rule into the legal foundation for moving hundreds of thousands of people out of institutions. A reader who follows these cases can trace that arc from restriction to reset, and can see why a single misread holding in this cluster would stand out to every disability lawyer in the country. The test of this article is whether that reader, finishing the last section, can trace the narrowing and the override, explain the immunity split and the constitutional reasoning behind it, and state the integration holding that reshaped long-term services. Each line of decisions below is written to make one of those tasks possible.

The statute at the center of this litigation arrived with unusual bipartisan force. The Americans with Disabilities Act of 1990, Public Law 101-336, was signed on July 26, 1990 by President George H. W. Bush after passing the House of Representatives 377 to 28 on July 12, 1990 and the Senate 91 to 6 on July 13, 1990. It built on an earlier funding-condition statute, Section 504 of the Rehabilitation Act of 1973, which had barred disability discrimination by programs receiving federal money. The 1990 law went further, extending the prohibition to private employers, state and local governments, and places of public accommodation regardless of federal funding. That breadth is what brought the law to the Court so quickly, because nearly every sector of American life had to work out what the new duties meant. The earlier statute did not disappear; it continued to operate alongside the new law, and its funding-condition structure would later supply an alternate route to the courthouse when the new statute’s own abrogation of state immunity faltered. Understanding the two statutes together is essential, because several of the decisions below turn on the interaction between them. Section 504’s model, nondiscrimination as a condition of federal money, differs structurally from the 1990 law’s model, nondiscrimination as a freestanding legal duty, and the Supreme Court’s treatment of each reflects that difference.

The statute organizes its duties into titles, and the titles matter enormously in the cases that follow. Title I governs employment. Title II governs public entities, meaning state and local governments and their programs, services, and activities. Title III governs public accommodations operated by private entities. Readers who need the full map of these duties can consult the guide to the statute’s three titles. Nearly every holding in this article turns on which title was at issue, because the Court read the definition of disability, the scope of congressional power, and the reach of the substantive duties differently depending on the title before it. The definition trilogy and the 2002 manual-tasks decision all arose under Title I, the employment title, which is why the narrowing of who counted as disabled hit workers first. The sovereign immunity decisions split precisely along the Title I and Title II boundary. The integration mandate is a Title II doctrine through and through. Two of the scope decisions construe Title III. A reader who keeps the titles straight will find the cluster far less confusing than it first appears, because the title often predicts the outcome better than the facts do.

Four lines of decisions make up the cluster, and each line answers a different question about the statute’s reach. The definition line asks who is protected: a trilogy of 1999 rulings plus a 2002 decision progressively narrowed who counted as disabled, culminating in a catch-22 that Congress addressed in 2008. These cases are of historical rather than operative significance, but they must be understood to make sense of the amendment, because the amendment’s findings name them and reject them. The sovereign immunity line asks who can be sued: the Court held in 2001 that the employment title does not validly abrogate state sovereign immunity because Congress had not documented a pattern of unconstitutional state discrimination sufficient to justify it, then held in 2004 that the public entity title does validly abrogate as applied to access to the courts because the underlying right is fundamental, and then held in 2006 that abrogation is valid wherever the conduct itself violates the Constitution. This is the most doctrinally interesting material in the cluster, and it produces the immunity map that organizes the entire field. The integration line asks where people with disabilities must be served: the 1999 decision holding that unjustified segregation in institutions is itself discrimination under the public entity title, subject to a reasonable modification limit and the resources of the state, is the single most consequential disability decision in American law. The accommodation and scope line asks what the duties require in practice: decisions holding that a seniority system ordinarily defeats a competing accommodation request, that an employer may refuse a placement posing a direct threat to the worker, that state prisons fall within the public entity title, that asymptomatic infection can be a disability, and that a professional sport must consider whether a requested modification would fundamentally alter the competition.

Reading this article: three axes

This article organizes fourteen decisions along three axes, and keeping the axes separate is the key to using the material correctly. The first axis is the title construed. Title I decisions govern employment, Title II decisions govern public entities, and Title III decisions govern public accommodations, and a holding under one title does not automatically transfer to another. The second axis is the question answered. Definition decisions determine who is protected. Immunity decisions determine whether states can be sued for damages. Duties decisions determine what the law requires of covered entities. A reader who confuses the axes will misapply the holdings, for example by treating a definition case as though it limited the integration duty, or by treating an immunity case as though it excused discrimination. The third axis is operative status. Some holdings remain operative. Some are historical, rejected by Congress in 2008. One survives in part. The doctrine ledger records each decision against all three axes, and the body text marks the status of every holding as it is discussed. Readers in a hurry can consult the ledger first and then read the sections relevant to their question. Readers who want the full picture should follow the article’s order, which moves from the threshold questions of coverage, through the remedial question of immunity, to the substantive questions of duty, mirroring the order in which a litigator encounters them. One final discipline governs the entire enterprise. Confirm every citation, every year, and every holding against the opinions themselves, and mark clearly which decisions were superseded by the 2008 amendments before treating any of them as authority. The definition cases are the trap: their reasoning is elegant, their majorities are lucid, and their holdings are no longer law. The immunity cases are the maze: three answers to one question, each correct within its domain, each misleading outside it. The duties cases are the reward: durable holdings that decided cases steadily through the article’s horizon. Hold those three characterizations in mind, and the cluster resolves from a jumble of citations into the coherent, if complicated, body of law this article describes.

How the definition of disability was narrowed and then reset

The 1990 statute defined disability to include a physical or mental impairment that substantially limits one or more major life activities, a record of such an impairment, or being regarded as having such an impairment. The definition looked capacious on paper. It was also structural: disability status was the front door of every claim under the statute, the element that determined whether the anti-discrimination and accommodation duties applied at all. A broad definition meant a broad statute; a narrow definition meant a narrow one, regardless of how generously the duties themselves were construed. This structural role is why the definition cases matter out of proportion to their doctrinal subtlety: they did not interpret a peripheral term but set the size of the protected class, and everything downstream, from accommodation disputes to immunity fights, presupposed that threshold. The threshold character also explains the litigation pattern of the era, in which definition disputes consumed the attention that might otherwise have gone to discrimination and accommodation. Over four decisions, the Court read it narrowly enough that Congress concluded the judiciary had departed from legislative intent and rewrote the definition by statute.

The definition cases arrived first and cut deepest. In 1999 the Court issued three opinions that together redefined disability for purposes of the employment title, and in 2002 it added a fourth that tightened the central statutory phrase further. The trilogy of 1999 took up plaintiffs whose impairments were controlled, corrected, or compensated: twin sisters with severe myopia who functioned normally with glasses, a mechanic with high blood pressure controlled by medication, and a truck driver with monocular vision whose brain had adapted to the loss of sight in one eye. The opinions held, in substance, that the statute asks how a person actually functions with whatever measures mitigate the impairment, not how the person would function without them. That single interpretive move excluded from coverage millions of people Congress had meant to protect, as the dissents warned at the time, and it created the catch-22 that the 2008 amendments would later address: the better a person managed an impairment, the less protection the law afforded. The 2002 decision then constricted the phrase “substantially limits” itself, demanding proof that an impairment prevents or severely restricts activities of central importance to most people’s daily lives, with an impact expected to be permanent or long term. Together these four opinions formed the most restrictive reading of a civil rights definition the Court had ever given, and for nearly a decade they functioned as a gate that kept cases from ever reaching the merits. They stand as historical rather than operative, because the ADA Amendments Act of 2008, Public Law 110-325, signed September 25, 2008 by President George W. Bush and effective January 1, 2009, rejected them by name. The amendment’s findings single out the 1999 trio and the 2002 decision as the errors to be corrected. Yet the override does not make the cases ignorable. Conduct before the effective date is still judged under the old definitions, the amendment is prospective only, and the opinions remain the essential background for understanding why the amended definition reads the way it does. A lawyer who cannot explain the trilogy cannot explain the amendment.

The three prongs and the regarded-as fallback

The statutory definition had three doors, and the Court’s narrowing shut them in sequence. The first door, actual disability, required an impairment that substantially limits a major life activity. Sutton’s mitigating-measures rule and Toyota’s strict construction of substantially limits together made this door narrow: the impairment had to be assessed after treatment, and even then it had to prevent or severely restrict activities of central importance to most people’s daily lives with permanent or long-term impact. The second door, a record of impairment, helped people whose conditions had resolved or whose history of impairment triggered discrimination, but it offered nothing to those whose conditions were ongoing and managed. The third door, being regarded as having an impairment, was the natural fallback for plaintiffs who failed the first door: even if not actually disabled under the narrowed reading, a worker could argue the employer perceived a substantially limiting impairment and acted on that perception.

Sutton and Murphy closed the fallback in the cases before them. In Sutton, the petitioners argued the employer regarded them as disabled, but the Court held the employer had not regarded them as substantially limited in the relevant sense; acting on an unmitigated condition did not equal regarding the worker as substantially limited. In Murphy, the employer’s reliance on Department of Transportation certification standards defeated the regarded-as claim, because the decision rested on regulatory compliance rather than on a judgment about the worker’s capacities. The pattern across both decisions was that the regarded-as prong required proof about the employer’s perception of substantial limitation, not merely proof that the employer knew about the impairment or applied a rule connected to it. For plaintiffs knocked out by the mitigating-measures rule, this meant there was often no second chance: the same facts that defeated the actual-disability claim tended to defeat the regarded-as claim as well, because employers who acted on neutral rules or regulatory standards were not regarding anyone as anything.

The closing of all three doors is what made the definition era so consequential and the congressional response so sweeping. Congress in 2008 did not merely tinker with one prong; its findings rejected the entire line of decisions, restoring a broad construction across the definition. The regarded-as prong’s history is a reminder that multi-prong definitions can fail together when the same interpretive approach governs each prong, and it explains why the amendments addressed the definition as a whole rather than case by case. The structural lesson extends beyond disability law: whenever a statute’s protections depend on a threshold definition, the interpretation of that definition determines the statute’s practical scope more than any other doctrinal question, and litigants will fight the threshold battle first because winning it ends the case.

Sutton v. United Air Lines, Inc. (1999)

Sutton v. United Air Lines, Inc., 527 U.S. 471 (1999), decided 7 to 2 with Justice O’Connor writing for the majority, established the mitigating-measures rule. The petitioners were twin sisters with severe myopia, each with uncorrected visual acuity of 20/200 or worse and corrected acuity of 20/20 with glasses or contact lenses. They applied for positions as commercial airline pilots, and the airline rejected them because its minimum vision requirement demanded uncorrected visual acuity of 20/100 or better. The sisters sued under the employment title, arguing that they were disabled within the meaning of the statute or, alternatively, that the airline regarded them as disabled. The Court held that whether an impairment substantially limits a major life activity must be determined with reference to any mitigating measures the individual uses, such as medication or corrective devices. A person whose impairment is controlled by treatment is assessed in the treated state, not in the untreated state. The petitioners in Sutton also failed the “regarded as” prong, because the airline’s vision requirement showed only that the carrier regarded them as unable to satisfy the visual standard for the single job of global airline pilot, not as substantially limited in the major life activity of working across a broad class of jobs. The logic of the majority rested on the statutory text: the definition speaks in the present tense of substantial limitation, and a limitation that has been corrected by mitigating measures is not, in the majority’s reading, a present substantial limitation.

The consequences of that textual choice were sweeping. Anyone whose condition was well managed by medication, devices, or other measures would be measured at their best rather than at their most impaired, and would therefore struggle to qualify as disabled even though the underlying condition persisted and required ongoing management. The decision functioned as a threshold gatekeeper: before any question of discrimination could be reached, a plaintiff had to show that the impairment, as mitigated, still substantially limited a major life activity. Plaintiffs who could not cross that threshold lost without any inquiry into whether the employer’s conduct was discriminatory, which meant the merits of many cases were never examined.

Justice Stevens dissented, joined by Justice Breyer, arguing that the majority had excluded from protection millions of people Congress meant to cover. The dissent’s logic ran in the opposite direction from the majority’s textualism: Congress had written a broad remedial statute aimed at the very population whose conditions were routinely managed by treatment, and reading those people out of the definition defeated the statute’s purpose. On the dissent’s view, the fact that an impairment responds to mitigating measures is precisely what makes the affected population large, and a definition that shrinks as treatment improves punishes people for managing their health. Justice Breyer also dissented separately on narrower grounds. The disagreement was therefore not only about grammar but about what kind of statute Congress had enacted: a narrow one aimed at a small class of severely limited individuals, or a broad one aimed at the much larger population that experiences discrimination on the basis of impairment.

Congress later superseded this holding in the 2008 amendments, whose findings explicitly rejected Sutton and its companion cases, so the mitigating-measures rule holds historical rather than operative significance after the amendments. The rejection was pointed: Congress named the decisions and repudiated their approach, an unusually direct override. For litigants whose claims arose before January 1, 2009, however, Sutton still governs, because the amendments apply only prospectively. The decision therefore remains essential knowledge for anyone working with pre-2009 conduct, and it remains the starting point for understanding why the definition had to be rewritten. The interpretive divide in Sutton also set the terms for the rest of the definition era: the majority’s textualism, reading “substantially limits” in the present tense as corrected, against the dissent’s purposivism, reading the statute as a broad remedial measure aimed at the population Congress described. That divide reappeared in every subsequent definition case, and Congress ultimately sided with the dissenters.

Murphy v. United Parcel Service, Inc. (1999)

The second member of the 1999 trilogy, Murphy v. United Parcel Service, Inc., 527 U.S. 516 (1999), applied the mitigating-measures rule to hypertension. The petitioner was a mechanic whose hypertension was controlled by medication, allowing him to function without significant restriction in daily life. The employer dismissed him when he failed to meet the Department of Transportation’s health certification requirements for driving commercial vehicles. The 7 to 2 majority, again written by Justice O’Connor, held that the petitioner’s high blood pressure had to be assessed in its medicated state, and that in that state he was not disabled within the meaning of the statute. The condition was real and ongoing, requiring daily medication to control, but under Sutton’s rule the controlled state was the legally relevant one, and in that state the limitation was not substantial. Murphy thus demonstrated that the mitigating-measures doctrine was not limited to unusual facts: it reached one of the most common chronic conditions in the workforce.

The Court further held that the employer did not regard him as disabled merely because it regarded him as unable to satisfy the federal certification standard for a particular driving job, because the certification decision turned on regulatory requirements rather than on a judgment about his capacity. This second holding mattered because the “regarded as” prong was the natural fallback for plaintiffs knocked out by the mitigating-measures rule: even if not actually disabled in the mitigated state, a worker might argue the employer treated the underlying condition as disqualifying. Murphy closed that fallback where the employer’s decision rested on compliance with federal safety standards rather than on stereotypes about the impairment. The distinction would echo in Albertsons, where reliance on federal safety regulation also proved decisive.

Justice Stevens dissented, joined by Justice Breyer, on reasoning consistent with his Sutton dissent: the majority’s approach excluded from the statute’s protection people whose conditions were managed but genuine, draining the statute’s coverage by treating successful management of an impairment as proof that no impairment worthy of protection existed. Murphy showed how the Sutton rule operated in practice: a common, treatable condition left the plaintiff outside the statute’s protection even though the underlying health condition was genuine and ongoing. Like Sutton, Murphy was superseded by the 2008 amendments and survives only as background for understanding why Congress acted. The pair of decisions, issued together, signaled that the mitigating-measures principle would be applied broadly rather than confined, which is part of why the legislative response eventually named the whole line. Murphy’s treatment of the Department of Transportation certification also foreshadowed Albertsons: in both cases, the employer’s reliance on federal safety standards supplied an independent basis for the decision that the disability analysis could not displace. The through line is that regulatory compliance and mitigating measures together left plaintiffs with little room, a combination Congress found inconsistent with the statute’s design.

Albertsons, Inc. v. Kirkingburg (1999)

The third member of the 1999 trilogy, Albertsons, Inc. v. Kirkingburg, 527 U.S. 555 (1999), was unanimous at 9 to 0, with Justice Souter writing for the Court, and differed from its companions in two respects that matter. First, the unanimity means the interpretive approach commanded agreement even from justices who had dissented in Sutton and Murphy. Second, the impairment was monocular vision, and the petitioner’s body had itself adapted: although he was blind in one eye, his brain had subconsciously compensated, giving him functional vision the Court described as effectively normal for most purposes. The employer, a grocery chain, had refused to retain the petitioner as a truck driver after a vision test revealed that he did not meet the Department of Transportation’s vision standard. The decision held that monocular vision is not a disability per se. Whether such a condition substantially limits a major life activity depends on an individualized assessment, and that assessment must account for mitigating measures, a category the opinion expressly extended to include the body’s own natural adaptations, not only devices and medications. The individualized-assessment requirement cut the other way, however, insisting that courts look at the actual person rather than applying categorical rules about conditions.

The Court also held that an employer may rely on binding federal safety regulations when making employment decisions, so compliance with a federal safety standard could justify an adverse action even where the statute might otherwise apply. Where a federal regulation sets a safety qualification, an employer’s adherence to that regulation is a legitimate basis for its decision, and the disability statute does not require the employer to second-guess the federal standard. This holding gave employers in regulated industries, particularly transportation, a clear defense rooted in federal law rather than in the employer’s own judgment.

The mitigating-measures portion of Albertsons fell with Sutton under the 2008 amendments, whose findings rejected Sutton and its companion cases. The safety-regulation holding, however, was not a mitigating-measures holding, and it survived. Albertsons therefore stands in a split posture: partly superseded, partly operative. The distinction is worth keeping precise, because it illustrates how a congressional override can be surgical: Congress repudiated the definition analysis while leaving the federal-regulation defense intact. A litigant invoking Albertsons for the safety-regulation point cites good law; a litigant invoking it for the mitigating-measures point cites history. The unanimity of the decision also matters for its afterlife: a 9 to 0 ruling carries institutional weight that survives partial supersession, and the surviving portion protects employers who follow binding federal safety standards rather than substituting their own judgments.

The federal safety thread

A subordinate theme running through Murphy and Albertsons deserves separate notice because it survived the 2008 amendments intact while the definition holdings around it fell. In both cases, the employer’s adverse action rested in part on a federal safety regulation: the Department of Transportation’s health certification for commercial drivers in Murphy, and its vision standard in Albertsons. The opinions treated compliance with binding federal safety requirements as a legitimate, nondiscriminatory reason for the employment decision, independent of the disability analysis. That treatment reflects a structural principle rather than a quirk of the two cases. The disability statute does not require employers to violate other federal law, and where Congress has entrusted safety judgments to a specialized agency, courts will not use the antidiscrimination law to second-guess those judgments through the back door of the definition. The 2008 amendments left this principle undisturbed because it was never part of the definition problem. The amendments rejected the idea that mitigating measures defeat coverage; they did not disturb the idea that federal safety regulations supply legitimate qualification standards. Under the amended statute, the safety-regulation issue is analyzed through the statute’s qualification-standards and direct-threat provisions rather than through the definition, which means the employer must still show that the standard is job-related and consistent with business necessity or that the worker poses a direct threat. The thread thus illustrates how a single opinion can contain both a rejected interpretive move and a durable substantive principle, and why Albertsons must be read prong by prong rather than as a unit.

Toyota Motor Manufacturing, Kentucky, Inc. v. Williams (2002)

Toyota Motor Mfg., Ky. v. Williams, 534 U.S. 184 (2002), decided 9 to 0 with Justice O’Connor writing for a unanimous Court, completed the narrowing from a different direction. The respondent was an assembly-line worker with carpal tunnel syndrome and related impairments affecting her hands, arms, and shoulders. She claimed that her impairments substantially limited the major life activity of performing manual tasks. Where Sutton had asked in what state to measure the impairment, Toyota asked how demanding the phrase “substantially limits” should be. The answer was strict: to be substantially limited in performing manual tasks, an individual must have an impairment that prevents or severely restricts the individual from doing activities that are of central importance to most people’s daily lives. The Court added that the impact of the impairment must also be permanent or long term. The test thus had two demanding components: the activities had to be of central importance to most people’s daily lives, not merely important to the particular plaintiff’s job, and the restriction had to be severe or preventive rather than moderate. The respondent’s evidence, which focused on her inability to perform a narrow set of job-related manual tasks, did not satisfy that standard.

This reading raised the bar for every plaintiff, including those whose impairments were not mitigated at all. A worker whose condition was uncontrolled but whose limitations, while real, did not prevent or severely restrict activities of central importance to most people’s daily lives would fail the test. Combined with Sutton, the effect was a pincer: the mitigating-measures rule removed people whose conditions were controlled, and Toyota’s strict standard removed many whose conditions were not. The unanimous vote underscored that the narrowing commanded broad agreement on the Court, which made the eventual legislative override all the more striking: Congress was not correcting a narrow majority but repudiating a consensus.

Congress later rejected this demanding standard in the findings of the 2008 amendments, and Toyota holds historical rather than operative significance following the amendments. The findings named the problem directly, characterizing the Court’s standard as unduly demanding and inconsistent with congressional intent. As with Sutton, the prospective-only character of the amendments means Toyota still governs pre-2009 conduct. The decision’s importance for a reader coming to these cases after the amendments is therefore diagnostic: it shows how far the definition had been narrowed, and it explains the specific language Congress chose when it rewrote the standard. The two components of Toyota’s test deserve separate attention because they did different work. The central-importance requirement directed courts away from the plaintiff’s own work life toward activities of most people’s daily lives, which excluded impairments whose effects were concentrated in the workplace. The permanent-or-long-term requirement excluded episodic or short-term conditions. Together they ensured that only severe, enduring, broadly-felt limitations qualified, a standard Congress found unduly demanding.

The trilogy’s internal logic

The three 1999 decisions fit together as rule, application, and extension. Sutton stated the rule: disability is assessed with reference to mitigating measures. Murphy applied the rule to a common chronic condition, showing that hypertension in its medicated state fell outside the definition and that reliance on federal certification standards defeated the regarded-as fallback. Albertsons extended the rule in two directions at once: it brought the body’s own natural adaptations within the mitigating-measures analysis, broadening the rule’s reach, while adding the federal-safety-regulation holding that gave employers an independent defense. The trilogy thus moved from principle to practice to boundary, each decision building on the last. The unanimity of Albertsons, following the 7 to 2 divisions in Sutton and Murphy, suggested the principle had consolidated; Congress’s later rejection of the whole line showed that consolidation on the Court did not equal acceptance by the legislature. The trilogy’s internal logic is also what made the 2008 findings name Sutton and its companion cases together: the three decisions were understood, by the Court and by Congress alike, as a single doctrinal project.

Why did the mitigating-measures rule create a catch-22 for plaintiffs?

Because the rule measured disability after treatment, people who managed their conditions fell outside the statute, while Toyota’s strict standard then excluded many whose conditions remained uncontrolled. Each doctrine removed a different group, so the combination covered fewer people than either rule alone, and Congress later rejected both.

The catch-22 stated precisely

The catch-22 that Congress addressed in 2008 can be stated as a dilemma facing any plaintiff with a manageable impairment. Under Sutton, the impairment had to be assessed as mitigated: medication, devices, and even the body’s own adaptations counted against the plaintiff. A person who diligently managed a condition was therefore measured at the person’s best, and the better the management, the weaker the disability claim. Under Toyota, the impairment as measured then had to prevent or severely restrict activities of central importance to most people’s daily lives, with permanent or long-term impact. A person whose condition was managed well enough to function would rarely meet that severity threshold, while a person whose condition was unmanaged would face the mitigating-measures question in reverse: the failure to mitigate could itself be held against the plaintiff, and the severity inquiry remained demanding regardless.

The result was a coverage band narrower than either doctrine produced alone. Sutton removed the well-managed from the protected class; Toyota removed many of the rest by raising the severity bar. The plaintiffs left inside the band were those whose impairments were both unmitigated and severe in the specific sense Toyota required, a far smaller population than the 1990 Congress had described. The regarded-as fallback, which might have caught those excluded from the actual-disability door, was itself narrowed in Sutton and Murphy, as discussed above. Congress’s findings in 2008 identified this dynamic as a departure from legislative intent and rejected both doctrines together, which is why the override addressed the definition as a whole rather than adjusting a single rule.

The litigation mechanics of the definition era

The definition trilogy and Toyota did more than construe words; they restructured Title I litigation around a threshold inquiry that often ended cases before the merits were reached. Because disability status was a gateway element of the claim, a plaintiff who failed the Sutton or Toyota analysis lost without any judicial examination of whether the employer’s conduct was discriminatory. Discovery into motive, comparator evidence, and accommodation feasibility became irrelevant once the threshold fell. For defendants, the definition was therefore the most valuable defense in the statute: it avoided the fact-intensive inquiries that make employment cases expensive and unpredictable, replacing them with a legal question about mitigating measures and severity that could often be resolved early.

This procedural dynamic explains why the definition era generated such pressure for legislative correction. When the threshold excludes a large share of plaintiffs, the statute’s substantive protections never get applied, and the pattern of outcomes reflects the gatekeeper rather than the merits. Employers learned to lead with the definition defense, and the reported decisions from the era show courts working through mitigating measures and substantial limitation with far more attention than they gave to discrimination or accommodation. The Stevens dissents identified the human cost: millions of people Congress meant to protect were losing at the threshold. The 2008 amendments, by rejecting the trilogy’s approach across the definition, moved the litigation focus back toward the merits, where Congress had intended it to be. Understanding the mechanics matters because the superseded cases still govern pre-2009 conduct, and in that closed universe the threshold still operates exactly as Sutton and Toyota described.

The congressional override of 2008

Congress answered with the ADA Amendments Act of 2008, Public Law 110-325, signed on September 25, 2008 by President George W. Bush and effective January 1, 2009. The legislative path was unusually consensual: the Senate passed the bill, S. 3406, by unanimous consent on September 11, 2008, and the House passed it by voice vote on September 17, 2008. Because neither chamber used a roll-call vote, no roll-call margins exist for this enactment, and none should be cited. The findings of the amendments explicitly rejected the Supreme Court’s Sutton line, naming Sutton and its companion cases, and rejected the demanding standard of Toyota. The full legislative history is set out in the ADA Amendments Act of 2008 guide.

The override is remarkable for its directness. Congress did not merely adjust the statutory language and leave the cases to be distinguished; it named the decisions in its findings and declared their approach inconsistent with congressional intent. The findings rejected the mitigating-measures analysis of Sutton and its companion cases, restoring the principle that disability is assessed without regard to the ameliorative effects of measures such as medication or devices, and they rejected Toyota’s demanding construction of substantially limits. The breadth of the legislative consensus, unanimous consent in the Senate and a voice vote in the House, reflected a judgment shared across the political spectrum that the Court had narrowed the protected class beyond what the 1990 Congress intended. The trilogy had made definition litigation expensive and unpredictable, and both sides of the employment relationship preferred a broad, clear threshold that would let cases be decided on the merits, on qualification and accommodation, rather than on whether the plaintiff counted as disabled at all. The coalition behind the bill united disability advocates with the business community, a breadth of support that explains how quickly the bill moved.

Two features of the override matter for reading the earlier cases correctly. First, the amendments are prospective only: they govern conduct occurring on or after January 1, 2009, so disputes arising before that date are still analyzed under the old, narrower readings. A claim based on a 2005 termination is a Sutton and Toyota case; a claim based on a 2010 termination is not. This temporal boundary is the reason the superseded decisions cannot simply be forgotten: they remain the governing law for a closed but real universe of disputes. Second, the override addressed the definition, not the structure of the statute’s duties, so the accommodation, immunity, and integration decisions discussed below were unaffected by it. Garrett, Lane, Georgia, Olmstead, Barnett, Echazabal, Yeskey, Bragdon, and PGA Tour all survived the amendments intact, because none of them turned on the definition analysis Congress repudiated. Overrides of this directness are rare: Congress named the decisions, described their errors, and rewrote the standard, all without a recorded vote against the measure.

The 2008 findings as a reply to the Court

The findings of the ADA Amendments Act repay close reading because they function as a point-by-point reply to the definition era. On Sutton and its companion cases, the findings rejected the mitigating-measures analysis root and branch: the determination whether an impairment substantially limits a major life activity must be made without regard to the ameliorative effects of mitigating measures. The phrase and its companion cases swept in Murphy and Albertsons alongside Sutton, repudiating the entire 1999 trilogy’s approach to measurement. On Toyota, the findings rejected the demanding standard that had required prevention or severe restriction of activities of central importance to most people’s daily lives, directing instead that substantially limits be construed in favor of broad coverage. The two rejections correspond exactly to the two narrowing moves the Court had made: Sutton narrowed the state in which the impairment is measured, and Toyota narrowed the severity required once measured.

The findings also addressed the interpretive posture the Court had adopted. Where the Court had construed the definition strictly, emphasizing the statute’s limits, Congress directed that the definition be construed broadly, emphasizing the statute’s remedial purpose. This was a deliberate reversal of interpretive direction, not merely a technical correction. And by making the amendments prospective only, Congress preserved the old readings for pre-2009 conduct while ensuring that no future litigant would face the catch-22 the trilogy and Toyota had created. The reply was thus complete: it identified the decisions, rejected their reasoning, reversed their interpretive posture, and set the temporal boundary for the transition.

Why the definition era still matters to practice

A reader might ask why an article should devote thousands of words to holdings Congress has erased. The answer is practical rather than antiquarian, and it has three parts. First, the amendments are prospective only, governing conduct on or after January 1, 2009. Claims arising from earlier events are still decided under the trilogy and Toyota, which means a lawyer evaluating an older claim, or defending one, must be able to apply the superseded framework accurately. The date-of-conduct question comes before the merits question, and getting it wrong means applying the wrong definition to the entire case. Second, the superseded opinions explain the amended text. The amended definition’s key phrases, the disregard of mitigating measures, the instruction to construe substantial limitation broadly, and the expansion of the regarded-as prong, were all written as direct responses to specific sentences in Sutton, Murphy, Albertsons, and Toyota. A practitioner who knows only the amended rule knows what the law says; a practitioner who knows the holdings it replaced knows why it says that, which matters whenever a novel fact pattern tests the edge of the new language. Third, the dissents in the trilogy did not lose in the long run. Justice Stevens’s warning that the majority had excluded millions Congress meant to protect became, nearly a decade later, the central finding of the amending Congress. Reading the dissents is therefore reading an early draft of the override, and the advocates who eventually persuaded Congress built their case on arguments the dissenters had already made.

The trilogy also teaches a lesson about interpretive method that outlives its holdings. The majorities read the definition through a textualist lens, emphasizing that the statute’s terms, impairment, substantially limits, major life activities, must be given content restrictive enough to distinguish the protected class from the general population. The dissents read the same terms through a purposivist lens, emphasizing the findings Congress had written into the statute about the breadth of the protected population. Neither side disputed that the words needed interpretation. They disputed which interpretive defaults should govern when the words run out. That dispute recurs across the disability field and across statutory interpretation generally, and the trilogy remains one of the clearest exhibits of the two approaches in direct collision. The unanimity of Albertsons sharpens the point. Even justices sympathetic to broad coverage accepted the mitigating-measures principle when the impairment was monocular vision and the adaptation was the body’s own, which suggests the principle had intuitive force beyond ideology. Congress rejected it anyway, and did so with the support of a coalition that included the business community, which indicates that the objection was not simply that the principle was wrong in the abstract but that its application had proved unworkable and unjust in practice.

There is a verification discipline that follows from all of this, and it belongs in any article on this cluster because the most common error in the field is also the easiest to commit. Before citing any definition holding, confirm the date of the conduct at issue. If the conduct postdates January 1, 2009, the trilogy and Toyota are background, not authority. If it predates that date, they are the governing law, and the amended regulations are irrelevant. Before citing Albertsons, identify which prong is needed: the mitigating-measures analysis is superseded, while the safety-regulation analysis survives. Before describing any of the four as governing law, stop, because none of them is, in whole, and one of them is only in part. These checks take seconds and prevent the kind of error that the brief for this article warns would be visible to every disability lawyer in the country. The doctrine ledger at the end encodes the same discipline in tabular form, so that the operative status of each decision can be confirmed without re-reading the analysis.

Can a state be sued for damages

The sovereign immunity line begins not with a disability case but with the doctrinal test that controls all of them. In City of Boerne v. Flores, 521 U.S. 507 (1997), Justice Kennedy writing for the majority, the Court held that Congress’s power under Section 5 of the Fourteenth Amendment is remedial rather than substantive. Congress may enact legislation enforcing the amendment’s guarantees, but it may not redefine them. For Section 5 legislation to be valid, there must be congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end. That phrase, congruence and proportionality, became the test against which every abrogation of state sovereign immunity would be measured. The test has two sides: Congress must identify a pattern of unconstitutional conduct by states that warrants a federal remedy, and the remedy it chooses must be proportionate to that pattern rather than a sweeping redefinition of constitutional rights. Boerne itself was not a disability case; it concerned the Religious Freedom Restoration Act, a federal statute that Congress had enacted to displace the Court’s own earlier interpretation of the Free Exercise Clause. Its migration into the disability cluster illustrates how constitutional doctrine travels: a test fashioned for one statute’s Section 5 problem became the controlling framework for another’s, because the underlying question, when Congress can subject states to private damages suits, is the same regardless of the substantive right at stake. Every disability immunity decision below is, in this sense, a Boerne application. In practice the Court first identifies the constitutional right at issue and the level of judicial scrutiny that right commands, because the strength of the underlying right determines how much legislative record is needed, and then examines whether Congress documented a pattern of unconstitutional state conduct sufficient to justify the chosen remedy.

The Eleventh Amendment background gives the test its bite. States enjoy sovereign immunity from private suits for damages in federal court unless Congress has validly abrogated that immunity or the state has waived it. Abrogation requires both a clear statement of congressional intent and a valid exercise of constitutional power, and after Boerne the power question turns on congruence and proportionality. The disability statute contained a clear abrogation provision, so the fight was entirely about power: had Congress, in subjecting states to damages suits, acted within its Section 5 authority? The three disability decisions answer that question differently for different titles and different rights, which is why the test must be kept in view throughout. The clear-statement half of the analysis, though undisputed here, is worth remembering because it is what separates abrogation from waiver: the disability statute speaks clearly enough to abrogate if the power exists, while the 2000d-7 funding condition speaks clearly enough to extract a waiver regardless of the power question.

Why did Garrett and Lane reach opposite results on state immunity?

Because the Court applied the same congruence-and-proportionality test to different rights. In Garrett the employment context offered no documented pattern of unconstitutional state discrimination, while in Lane the fundamental right of access to courts supplied the constitutional weight that made Title II’s remedy congruent and proportional as applied.

Board of Trustees of the University of Alabama v. Garrett (2001)

Board of Trustees of the University of Alabama v. Garrett, 531 U.S. 356 (2001), decided 5 to 4 with Chief Justice Rehnquist writing for the majority, applied the Boerne test to Title I of the disability statute and found the abrogation invalid. The plaintiffs were state employees who alleged disability discrimination in employment by state agencies and sought money damages. The decision was the first to test the statute’s abrogation provision against the congruence-and-proportionality standard, and its negative answer set the terms for everything that followed: every subsequent immunity dispute would be argued in the shadow of Garrett’s insistence on a documented pattern of unconstitutional state conduct. The majority concluded that Congress had not documented a pattern of unconstitutional discrimination by states as employers against people with disabilities sufficient to justify subjecting states to private suits for money damages under the employment title. The legislative record, in the majority’s assessment, contained evidence of discrimination by private actors and by society generally, but little showing that states themselves had engaged in unconstitutional employment discrimination against disabled workers. Without that predicate pattern, the damages remedy could not be congruent and proportional to any constitutional injury.

The constitutional baseline mattered enormously. Disability classifications receive only rational-basis review under the Equal Protection Clause, so most state employment decisions affecting disabled workers were not unconstitutional to begin with. A state that declines to hire a worker for a rational reason has not violated the Constitution, even if the same decision would violate the statute. Because the underlying constitutional violations were rare, a broad statutory damages remedy looked disproportionate: Congress was creating liability far beyond the constitutional floor, which is exactly what Boerne forbids. The majority thus treated the level of constitutional scrutiny as a key input to the congruence analysis, a move that would prove decisive again in Lane.

Chief Justice Rehnquist wrote for the majority, joined by Justices O’Connor, Scalia, Kennedy, and Thomas. Justice Kennedy filed a concurring opinion, joined by Justice O’Connor, emphasizing the limited nature of the holding. Justice Breyer dissented, joined by Justices Stevens, Souter, and Ginsburg, arguing that the legislative record did show sufficient state discrimination and that Congress was entitled to deference in assessing the pattern. The dissent would have credited Congress’s judgment about the record rather than conducting an independent audit of it. The disagreement over the pattern requirement reveals how much work that requirement does in the Boerne framework: the majority treated the absence of a documented pattern of unconstitutional state employment discrimination as fatal, while the dissent treated Congress’s broader findings about disability discrimination as sufficient. The pattern must be of unconstitutional conduct by states, not of discrimination generally, and that qualifier is what made the legislative record inadequate in the majority’s eyes. The holding is limited in an important way: it addresses money damages only, and it addresses Title I only. It does not answer whether the same reasoning would apply to other titles or to other forms of relief, and the Court would soon show that the answer differs. The damages-only scope is worth underscoring because it is the most common overreading of Garrett: the decision does not hold that states are immune from Title I suits altogether, but that Congress did not validly abrogate immunity for private damages actions. Other forms of relief and other enforcement mechanisms were left for other cases, and the Section 504 funding waiver discussed below provides an entirely separate route to damages against fund-receiving states.

Why employment was the hardest case for Congress

The employment title was always going to be the most difficult abrogation case Congress could bring, and understanding why clarifies the entire immunity map. The difficulty has two sources, one doctrinal and one evidentiary, and they compound each other. The doctrinal source is the level of scrutiny. Disability classifications receive rational-basis review under the Equal Protection Clause, which means states may draw disability-based distinctions in employment so long as they are rationally related to a legitimate interest. For Congress to justify abrogating immunity, it had to show not merely that states discriminated but that they did so unconstitutionally, meaning irrationally, and the rational-basis standard makes that showing demanding by design. The evidentiary source is the nature of the record. Employment discrimination is diffuse, individualized, and often documented through private-sector experience, while the abrogation inquiry demands evidence of unconstitutional conduct by states in their capacity as employers specifically. Congress had assembled extensive material on disability discrimination in the American workplace, but the subset concerning state employers acting unconstitutionally was, in the majority’s assessment, too thin to carry the remedy. The combination is what doomed the abrogation claim: the hardest standard of review multiplied by the hardest record to build. Lane succeeded where Garrett failed because it changed both variables at once, substituting a fundamental right for rational-basis review and a concentrated, well-documented courthouse record for a diffuse employment one. The lesson generalizes. Abrogation analysis is always a function of the right and the record together, and the employment title presented Congress’s weakest combination of the two.

The legislative record problem in Garrett

Garrett deserves slower treatment than it usually receives, because the opinion’s logic is frequently caricatured and its limits frequently ignored. The caricature runs as follows: the Court decided that disability discrimination by states does not matter, or that Congress lacks power to protect disabled workers. The opinion says neither. What it says is narrower and more technical. Congress may abrogate state sovereign immunity when it acts to enforce the Fourteenth Amendment, but the enforcement must be congruent and proportional to a documented pattern of unconstitutional conduct by the states. The legislative record behind the employment title documented extensive disability discrimination in the economy at large, but very little of it concerned states acting as employers violating the Constitution. The distinction between unconstitutional conduct and merely wrongful conduct does the decisive work here, and it is the distinction most commentary elides. The Fourteenth Amendment, as the Court had long interpreted it, prohibits only irrational disability classifications, since disability is not a suspect classification triggering heightened scrutiny. Much of the discrimination Congress documented, real and blameworthy as it was, involved private employers and local governments, or involved state conduct that a court applying rational-basis review would have upheld. For abrogation purposes, only the subset of the record showing unconstitutional state action counts, and the majority found that subset too thin to support subjecting states to damages liability.

The dissent’s answer, presented with the equal care the subject requires, is that the majority graded Congress’s homework with impossible strictness. Justice Breyer catalogued the testimony, reports, and findings Congress had assembled, including evidence of state employers’ discrimination, and argued that the record, viewed as a whole rather than sifted for only the most pristine examples of unconstitutional state action, amply supported the remedy. The disagreement is methodological. The majority demanded a record of constitutional violations by states qua states, discounted evidence about private and local actors, and treated anecdote as insufficient. The dissent treated the record holistically, credited Congress’s institutional competence to assess the national problem, and warned that the majority’s approach would make the enforcement power a dead letter whenever Congress legislated against a diffuse social problem rather than a concentrated pattern of state defiance. Both positions are internally coherent, and the choice between them reflects deeper views about the judicial role in reviewing congressional fact-finding that divide the justices across many fields beyond disability.

Two limits on the holding require emphasis because practitioners and commentators regularly overread them. First, Garrett is a damages case. It holds that Congress did not validly abrogate immunity from suits for money damages under the employment title. It does not hold that the employment title is unconstitutional as applied to the states, and it does not bar suits seeking prospective relief against state officers for ongoing violations. A state agency that discriminates in employment remains subject to the statute’s prohibitions; what changes is the remedy available in federal court, not the underlying legal duty. Second, the holding is title-specific. It says nothing about the public entity title, whose abrogation analysis Lane would address three years later under a different constitutional calculus. The temptation to read Garrett as a general statement about the disability statute and the states should be resisted. It is a statement about one title, one remedy, and one legislative record, and its reasoning affirmatively invites the different result Lane reached by identifying the variables, the strength of the underlying right and the specificity of the record, that would change the outcome.

The Kennedy concurrence in Garrett

Concurring opinions in five to four decisions repay attention because they often mark the boundary of the majority’s holding, showing where the fifth vote was willing to go and, by implication, where it was not. Justice Kennedy’s concurrence in Garrett, joined by Justice O’Connor, emphasized the limited nature of the holding, underscoring that the decision addressed only the validity of abrogation for money damages under the employment title and left the statute’s substantive prohibitions intact. The concurrence’s significance is partly temperamental. Justice Kennedy authored the Boerne test itself, and his separate writing in Garrett reads as an effort to prevent the majority opinion from being taken as a broader statement about congressional power or about disability rights than its reasoning supports. For lower courts, the concurrence functions as a caution against overreading: Garrett does not license dismissal of prospective-relief claims, does not question the statute’s application to private employers, and does not resolve the public entity title’s abrogation question. The pattern of a narrow concurrence cabining a five to four majority recurs in Lane, where Justices Souter and Ginsburg wrote separately to emphasize the application-specific character of the holding. Reading the concurrences alongside the majorities gives the truest picture of what each decision actually settled, which in this cluster is consistently less than the dissents feared and less than careless summaries claim.

Tennessee v. Lane (2004)

Tennessee v. Lane, 541 U.S. 509 (2004), also decided 5 to 4 but with Justice Stevens writing for the majority, reached the opposite result on abrogation by applying the same Boerne test to a different constitutional setting. The plaintiffs were individuals with disabilities who alleged that the state court system had denied them physical access to courthouses, including one plaintiff who had been required to crawl up stairs to reach a courtroom. They sued the state for damages under the public entity title. The decision is the immunity line’s hinge: it proved that Garrett’s negative answer was not a general hostility to abrogation but a specific failure of the Title I record, and it established the as-applied methodology that lets the same statute be validly abrogating in one category of cases and invalidly abrogating in another. The Court held that Title II validly abrogates state sovereign immunity as applied to the class of cases implicating the fundamental right of access to the courts. Where the underlying right is fundamental, the constitutional weight on Congress’s side of the scale is far greater: the Due Process Clause independently protects meaningful access to judicial proceedings, and the historical record of state exclusion of people with disabilities from courthouses and court services supported a congruent and proportional remedy. Congress had documented a pattern of states failing to provide accessible court facilities and proceedings, and the Title II remedy for that class of cases fit the documented injury.

The doctrinal structure of Lane deserves close attention because it is application-specific rather than blanket. The Court did not hold that Title II validly abrogates immunity in all its applications; it held that abrogation is valid for the class of cases implicating access to the courts. This as-applied approach meant that the congruence analysis could be conducted category by category, with the constitutional status of the underlying right doing the decisive work. For rights subject to heightened constitutional protection, the same statutory remedy that failed in Garrett could succeed, because the predicate constitutional violations were both more numerous and more serious. The methodology thus preserved Garrett while reaching the opposite outcome: the test stayed constant, and the inputs changed.

Justice Stevens wrote for the majority, joined by Justices O’Connor, Souter, Ginsburg, and Breyer. Justice Souter filed a concurring opinion, joined by Justice Ginsburg, and Justice Ginsburg filed a separate concurrence joined by Justices Souter and Breyer, offering additional support for the as-applied methodology that made the majority’s holding work. Chief Justice Rehnquist dissented, joined by Justices Kennedy and Thomas, and Justices Scalia and Thomas dissented separately, disputing both the reading of the legislative record and the propriety of the class-of-cases approach. The dissenters would have required a tighter fit between the documented violations and the remedy, and they questioned whether the record supported abrogation even for the access-to-courts category. The dissents objected that the class-of-cases framing proved too much: by defining the class at the level of court access, the majority effectively sustained a broad swath of the title’s applications on the strength of a record concentrated on courthouse exclusion, and the congruence between the documented injury and the remedy looked looser than Boerne should tolerate. The objection has force, and it identifies the pressure point in Lane’s logic. But the majority’s answer also has force: Congress legislates against problems, not against perfectly bounded categories, and the right of access to the courts is precisely the kind of fundamental interest for which prophylactic protection is most justified. Lane’s 5 to 4 division, with Justice O’Connor joining the majority this time, showed how narrowly the immunity question divided the Court and how much turned on the constitutional character of the right at stake. The shift from Garrett’s lineup is itself instructive: the same Justice who joined the majority in closing the Title I damages door joined the majority in opening the Title II access-to-courts door, because the Boerne inputs differed even though the Justice did not. Lane thus demonstrates that the immunity map is a product of doctrine rather than of fixed judicial attitudes toward the statute.

Why court access changed the calculus

Lane’s reasoning repays close attention because it shows the Boerne test operating as a sliding scale rather than a simple on-off switch. The critical variable is the constitutional right that the legislation enforces. In Garrett, the underlying equal protection claim for disability classifications received rational-basis review, the most deferential standard, which meant Congress needed an unusually strong record to justify abrogation. In Lane, the underlying right was access to the courts, a right the Court has long treated as fundamental, rooted in due process and in the structural role courts play in the constitutional order. Fundamental rights change the abrogation math in two ways. First, the judiciary guards them with heightened scrutiny, which means more state conduct counts as unconstitutional and the universe of violations Congress may remedy is larger. Second, Congress needs a less extensive legislative record to justify prophylactic legislation protecting such rights, because the risk of constitutional harm from state inaction is itself greater. The majority therefore did not need to find the kind of voluminous record of unconstitutional state employment discrimination that Garrett had demanded. It needed to show, and found, a documented pattern of states excluding people with disabilities from courthouses and court proceedings, a record the opinion described as well established.

The class-of-cases framing is the opinion’s most consequential methodological choice and the focus of the sharpest dissent. Rather than asking whether abrogation was valid for the public entity title in all its applications, the majority asked whether it was valid as applied to the class of cases implicating access to the courts. That framing allowed the Court to validate the remedy for a defined domain without deciding the title’s validity across every program and service states operate. The dissents attacked that framing, and the attack names the real vulnerability in Lane’s logic: a record concentrated on courthouse exclusion was asked to carry a remedy defined at the level of court access generally. The majority’s defense is that the congruence inquiry is pitched to the constitutional weight of the right at stake, and a fundamental right lowers what the record must show. The five to four division reflects a genuine disagreement about how tightly courts should police the fit between congressional findings and statutory remedies, not a disagreement about whether excluding disabled litigants from courthouses is wrong.

United States v. Georgia (2006)

United States v. Georgia, 546 U.S. 151 (2006), decided 9 to 0 with Justice Scalia writing for the Court, resolved the cleanest abrogation question and left the hardest one open. The case arose from a state prison inmate’s allegations of disability discrimination and related constitutional violations. The decision’s unanimity, following the 5 to 4 divisions of Garrett and Lane, marked the point on which the entire Court could agree: where Congress remedies conduct the Constitution itself forbids, the Section 5 power is at its strongest and the Boerne test is satisfied without controversy. The Court held that Title II validly abrogates state sovereign immunity insofar as the challenged conduct actually violates Section 1 of the Fourteenth Amendment itself. Where the plaintiff alleges conduct that is independently unconstitutional, no Boerne problem arises, because Congress is plainly enforcing the amendment rather than expanding it. A statute that provides a remedy for conduct the Constitution already forbids is the core case of valid Section 5 legislation, and congruence and proportionality are satisfied almost by definition when the remedy tracks the constitutional violation.

The Court ordered a claim-by-claim remand so that the lower courts could determine which allegations stated actual constitutional violations and which alleged only statutory violations. On remand, courts should determine claim by claim whether the alleged conduct violates the Fourteenth Amendment, and only where it does not should they reach the harder question of whether abrogation is valid for conduct that violates the statute alone. This procedural directive is as important as the substantive holding: it told lower courts to sort each claim into one of two buckets before deciding the immunity question. Claims in the constitutional bucket proceed against the state; claims in the statutory-only bucket face the unresolved abrogation question. The remand structure thus operationalized the immunity map, turning an abstract doctrinal distinction into a concrete pleading and proof exercise. The unanimity also carries a message about the limits of the earlier divisions. Justices who had dissented in Lane, skeptical of broad prophylactic abrogation, joined an opinion validating abrogation across a defined domain, because the domain was anchored in actual constitutional violations rather than in congressional predictions about where violations might occur.

Justice Stevens concurred, joined by Justice Ginsburg, underscoring the breadth of agreement on the core principle even as the Court deliberately reserved the question that would have divided it. The decision expressly left open whether abrogation is valid for Title II conduct that violates the statute but not the Constitution, so that question remained undecided after Georgia. The open question, whether Title II validly abrogates immunity for conduct that violates only the statute, remains the live frontier, and lower courts confronting it must apply Lane’s application-specific approach, asking which class of cases is at issue and which constitutional right stands behind the claim, without the benefit of a Supreme Court answer. The reservation is deliberate judicial restraint: the Court decided only what the case required and left the harder question for a case that squarely presents it. For litigators, the lesson is to plead constitutional violations wherever the facts support them, because doing so moves the case from the uncertain outer ring into Georgia’s safe harbor, and to develop the Lane-style record for the class of cases at issue as a fallback. Georgia therefore functions as the settled center of the immunity map, with the contested edges defined by Garrett on one side and the open question on the other.

The remand framework as a litigation procedure

Georgia’s most underappreciated contribution is procedural. Beyond its substantive holding that abrogation is valid where the conduct independently violates the Fourteenth Amendment, the opinion gave lower courts an order of operations for the cases that remain. First, determine whether the plaintiff’s allegations state a violation of section 1 of the Fourteenth Amendment on their own terms. If they do, the abrogation question is resolved, because Congress unquestionably has power to supply a remedy for conduct the Constitution directly forbids, and the court need not reach the harder prophylaxis question. Only if the alleged conduct does not independently violate the Constitution should the court proceed to ask whether abrogation is valid for conduct that violates the statute alone. That sequencing matters enormously in practice. It means many cases never reach the open question the Court reserved, because allegations of disability discrimination by state actors frequently include, or can be pleaded to include, independent constitutional claims such as due process or equal protection violations. The unanimous Court thus created a safe harbor defined by the Constitution itself, inside which abrogation is secure, and confined the uncertainty to the ring of prophylactic-only applications outside it.

Garrett and Lane side by side under the Boerne test

The clearest way to understand the immunity line is to run Garrett and Lane through the Boerne test input by input, because the test is identical in both cases and only the inputs differ. The first input is the constitutional character of the right at stake. In Garrett, the right was equal treatment in state employment for workers with disabilities, a classification that receives only rational-basis review under the Equal Protection Clause. Most adverse employment decisions are therefore constitutional, and the universe of unconstitutional conduct Congress could remedy was small. In Lane, the right was access to the courts, a fundamental right protected by the Due Process Clause. Excluding people with disabilities from judicial proceedings strikes at a core constitutional guarantee, so the universe of unconstitutional conduct was large and serious. The same statutory remedy looks disproportionate against a thin constitutional background and proportionate against a thick one.

The second input is the legislative record of unconstitutional state conduct. In Garrett, the majority found that Congress had documented discrimination in society generally and by private employers, but had not shown a pattern of unconstitutional discrimination by states acting as employers. The record spoke to the wrong defendant or the wrong constitutional standard. In Lane, the record contained documented instances of states excluding disabled people from courthouses and court proceedings, a pattern tied directly to the fundamental right at stake. The congruence inquiry asks whether the remedy matches a documented injury, and Lane’s record supplied the match that Garrett’s lacked.

The third input is the fit between the remedy and the injury. Title I’s damages remedy in Garrett swept across all state employment decisions, most of them constitutional, in service of a record that showed little unconstitutional state conduct. Title II’s remedy in Lane, as applied to the access-to-courts class of cases, tracked the documented exclusion from judicial proceedings. The as-applied framing did heavy lifting here: by limiting the holding to the class of cases implicating court access, the majority ensured the remedy’s scope matched the record’s scope. A blanket validation of Title II abrogation would have raised the same proportionality problem Garrett identified, which is why Lane’s narrowness is a feature of its reasoning rather than a cautious afterthought.

The fourth input is the scope of the resulting holding. Garrett closed the Title I damages door against states. Lane opened the Title II door for the access-to-courts class. Georgia then added the principle that the door is open wherever the conduct itself violates Section 1 of the Fourteenth Amendment, with a remand to sort claims accordingly. Taken together, the three decisions do not reflect shifting judicial attitudes toward disability rights; they reflect a single test producing different results as the constitutional inputs change. That is the doctrinal discipline the neutrality of this account requires: describe the test, identify the inputs, and let the outcomes follow.

The immunity map

The namable claim of this cluster is the immunity map, not the merits. For a state employee, the decisive question is rarely whether discrimination occurred but which title applies and whether the underlying right is one the Constitution independently protects. Identical facts produce different outcomes depending on the defendant and the theory, and the map explains why. A state worker denied a promotion because of a disability faces one legal regime under Title I and another under Title II; a person excluded from a courthouse faces a third. The merits of the discrimination claim may be identical in each scenario, yet the availability of damages against the state turns entirely on the doctrinal coordinates.

Under Garrett, a state employee seeking money damages for employment discrimination under Title I faces a closed door: Congress did not validly abrogate state sovereign immunity for that title, because it failed to document the pattern of unconstitutional state discrimination that congruence and proportionality demand. The employment context is constitutionally thin, because disability classifications receive only rational-basis review, and the legislative record did not supply the pattern of state constitutional violations needed to support the remedy. Under Lane, the same state employee alleging exclusion from the courts under Title II faces an open door, at least for the access-to-courts class of cases, because the fundamental right at stake supplies the constitutional foundation the employment context lacked. The Due Process Clause independently protects meaningful access to judicial proceedings, and the documented history of exclusion gave Congress the predicate its employment record did not.

Under Georgia, the door is open wherever the alleged conduct itself violates Section 1 of the Fourteenth Amendment, regardless of title, and the courts must sort claims one by one to identify which allegations meet that standard. This is the map’s most practically important feature: it converts the immunity question into a pleading exercise, in which the plaintiff’s lawyer must frame allegations to land in the constitutional bucket wherever possible. For Title II claims that allege only statutory violations without an independent constitutional violation, the door remains neither open nor closed: Georgia expressly left that question undecided, and it remains the largest blank area on the map.

For the immunity map, Lane contributes the principle that the title alone does not determine the outcome. Two plaintiffs suing the same state under the same title can face different abrogation results depending on the constitutional right their cases implicate. A claim about physical access to a courtroom invokes a fundamental right and falls within Lane’s validated domain. A claim about access to a state recreational program, implicating no fundamental right, falls outside it and must find another basis for abrogation or another route to relief. That variance is the point of the article’s central claim: the decisive question is which underlying right is at stake, and the map must be consulted case by case rather than memorized as a blanket rule.

This is why the abrogation reasoning must be described by its doctrinal test rather than by outcome preference. The test is always congruence and proportionality. What varies is the constitutional weight on each side: rational-basis employment classifications on one side, fundamental rights of court access and independently unconstitutional conduct on the other. A practitioner who grasps the map can predict outcomes; one who asks only whether the discrimination was real cannot. The map also explains the strategic importance of the Section 504 funding waiver discussed next: where the abrogation door is closed or uncertain, the waiver offers a separate entrance that does not depend on Section 5 at all.

What is the immunity map for a state employee?

For a state employee the map runs through three holdings: Garrett bars damages under the employment title, Lane permits damages under the public entity title for cases implicating court access, and Georgia permits damages wherever the conduct independently violates the Fourteenth Amendment, with the Section 504 waiver adding a fourth route for federally funded programs.

Pleading under the map

The immunity map is not an academic diagram. It dictates concrete choices in drafting complaints, selecting defendants, and choosing remedies, and the cost of ignoring it is dismissal of claims that describe real discrimination. The first choice is the title. A state employee with an employment grievance must understand that the employment title offers no damages against the state itself after Garrett, and should consider from the outset whether any aspect of the claim can be framed under the public entity title or the funding-condition statute instead. A person excluded from a state program or service starts from a stronger position under Lane and Georgia but must still identify the constitutional right behind the claim and be prepared to show why the case falls within a validated application. The second choice is the remedy. Where damages are barred, prospective relief against responsible officers for ongoing violations remains available, and the strategic question becomes whether an injunction, declaratory judgment, or policy change satisfies the client’s objectives. Lawyers who reflexively seek damages in every case will lose cases that a remedies-first analysis would have won.

The third choice is the parallel claim. The funding-condition statute’s express waiver means that a well-drafted complaint against a state defendant often includes a section 504 count alongside the statutory counts, routed through the federal funds the program receives. That count avoids the abrogation analysis entirely and can preserve a damages remedy where the direct statutory damages claim fails. The pleading must establish the funding nexus, which is usually straightforward for state agencies but must be alleged with care. The fourth choice concerns the constitutional claims themselves. Because Georgia’s safe harbor turns on whether the conduct independently violates the Fourteenth Amendment, complaints should plead viable due process and equal protection theories where the facts support them, rather than relying solely on the statutory violation. Each of these choices reflects the map’s central lesson: the outcome depends less on the moral clarity of the underlying facts than on the doctrinal route selected to reach a remedy. Two lawyers with identical facts can achieve opposite results depending on which title they invoke, which remedy they seek, and whether they plead the constitutional and funding-condition theories that the map makes available. That variance unsettles newcomers to the field, but it is the predictable product of a doctrine that tests congressional power rather than the wrongfulness of discrimination.

The two tracks of Title II after Lane and Georgia

Title II operates on two tracks after Lane and Georgia, and keeping them distinct prevents the most common analytical error in this area, which is treating Lane as a general validation of the title. The first track is the validated domain: cases implicating the fundamental right of access to the courts, where Lane holds abrogation valid, and cases where the alleged conduct independently violates the Fourteenth Amendment, where Georgia holds abrogation valid. Claims on this track proceed to the merits without constitutional detour, and the litigation focuses on the statutory questions of reasonable modification and fundamental alteration. The second track is the open domain: Title II-only conduct, where the plaintiff alleges a violation of the statute’s prophylactic requirements without an independent constitutional violation, in a class of cases the Court has not yet addressed. Claims on this track must survive the application-specific inquiry Lane modeled, with the outcome turning on the constitutional right behind the claim and the strength of the congressional record for that class of cases.

The practical consequence is that Title II litigation against states begins with a sorting exercise. Counsel must identify which track the claim occupies before briefing the merits, because a merits victory is worthless if the abrogation defense succeeds. The sorting turns on facts: does the complaint allege exclusion from a courthouse or court proceeding, or some other fundamental-rights context, or does it allege facts that independently state a due process or equal protection violation? If the answer to either question is yes, the claim sits on the validated track. If not, counsel must build the Lane-style case for the particular class of cases at issue, marshaling the congressional record for that domain and explaining why the remedy is congruent and proportional. The open domain is not a dead zone; it is a litigation frontier where careful lawyering can still win, but it demands constitutional briefing that the validated track does not. Until the Supreme Court resolves the reserved question, the two-track structure is the working reality, and any account of Title II that ignores it misdescribes the law. One further consideration completes the picture. The funding-condition route under section 504 runs parallel to both tracks and is unaffected by which one a claim occupies, because the waiver rests on the state’s acceptance of federal money rather than on abrogation. A plaintiff whose Title II damages claim sits on the uncertain open track can therefore pursue the section 504 count as a hedge, preserving a damages remedy even if the abrogation analysis fails. The complete map has three layers, the Title II tracks, the funding-condition bypass, and the availability of prospective relief, and competent counseling requires checking all three before advising a client that no remedy exists.

Abrogation and waiver: the two routes against states

A state defendant can be reached by two entirely different doctrinal routes, and confusing them is one of the costliest errors in this field. The confusion is understandable, because both routes end in the same place, a state answering a private damages suit, but they get there by different constitutional vehicles and are governed by different tests. Keeping the vehicles distinct is the first step in any state-defendant analysis: ask first whether the claim relies on Congress’s power to abrogate or on the state’s waiver through funding, because the answer determines which doctrine applies and which precedents control. The first route is abrogation: Congress uses its Section 5 power to strip the state’s Eleventh Amendment immunity by statute. Abrogation requires a clear statement, which the disability statute contains, and a valid exercise of Section 5 power, which is where Garrett, Lane, and Georgia do their work. The Boerne congruence-and-proportionality test governs, and the result is the immunity map: no valid abrogation for Title I damages, valid abrogation for Title II as applied to court access, valid abrogation wherever the conduct itself violates Section 1 of the Fourteenth Amendment, and an open question for Title II-only statutory violations. The route is powerful where it works and unavailable where it does not, and its availability turns on constitutional variables the plaintiff cannot control.

The second route is waiver: the state surrenders its immunity by accepting federal funds conditioned on the surrender. Section 1003 of Public Law 99-506, codified at 42 U.S.C. 2000d-7, states the condition explicitly for Section 504 of the Rehabilitation Act of 1973: a state shall not be immune under the Eleventh Amendment from suit in federal court for a violation of section 504. Waiver requires no Boerne analysis, because the state acts voluntarily rather than being stripped of immunity by congressional command. The doctrinal question is instead whether the funding condition is clear enough that the state’s acceptance counts as a knowing waiver, and the statute’s language satisfies that requirement directly. For plaintiffs, the practical consequence is that Section 504 claims against fund-receiving states avoid the entire immunity map: there is no need to sort by title or by underlying right, because the waiver does not depend on Section 5.

The two routes interact in litigation strategy. A plaintiff suing a state agency that receives federal funds will typically plead both the disability statute and Section 504, using the waiver route as insurance against the abrogation route’s gaps. Where the disability statute’s abrogation fails under Garrett, the Section 504 waiver may still support damages; where the conduct falls in Georgia’s open-question area, the waiver removes the uncertainty. Understanding both routes, and keeping their doctrinal bases distinct, is what separates a competent state-defendant practice from a confused one. The statutory basis for the waiver route is Section 1003 of Public Law 99-506, enacted October 21, 1986 and codified at 42 U.S.C. 2000d-7. Because Section 504 applies to programs receiving federal financial assistance, a state that accepts federal funds waives its immunity to private suits under that statute as a condition of the funding. The full context of that earlier statute is set out in the guide to Section 504 of the Rehabilitation Act of 1973. In practice, the waiver is often the more reliable route against state defendants precisely because it avoids the title-by-title and right-by-right sorting that the immunity map demands.

One later development on remedies deserves mention with explicit dates, because it postdates this article’s reference date of August 15, 2015 and therefore sits outside the article’s horizon. On April 28, 2022, the Supreme Court decided Cummings v. Premier Rehab Keller, 596 U.S. 212 (2022), 6 to 3 with Chief Justice Roberts writing for the majority, holding that emotional-distress damages are not recoverable under the Spending Clause anti-discrimination statutes, including Section 504. The reasoning reaches the funding-condition framework itself, limiting the remedies available through the waiver route even as the waiver of immunity remains intact. It is noted here only to complete the record; it was not part of the law as it stood within this article’s 2015 horizon and should not be read as qualifying the holdings described above.

The integration mandate

The third line of the cluster contains a single decision, but it is the line that reshaped American disability policy most profoundly. Olmstead v. L.C., 527 U.S. 581 (1999), decided 6 to 3 with Justice Ginsburg writing for the majority, is the single most consequential disability decision in American law. The plaintiffs were two women with mental disabilities who had been confined in a Georgia state psychiatric hospital. Their treatment professionals had determined that community-based care was appropriate for each of them, and each woman sought placement in a community program, but the state kept them institutionalized, citing the lack of available community placements and budgetary constraints. They sued under the public entity title. The Court held that unjustified institutional isolation of people with disabilities is itself a form of discrimination under Title II of the statute. Segregation in an institution, when community placement would be appropriate, counts as discrimination even without proof that nondisabled people were treated better. The conceptual move was significant: discrimination was not limited to differential treatment of similarly situated people but encompassed the unjustified segregation of disabled people in settings that separated them from community life. The majority’s reasoning proceeded in three steps. First, institutional placement of persons who can handle and benefit from community settings perpetuates unwarranted assumptions that such persons are incapable or unworthy of participating in community life. Second, confinement in an institution severely diminishes the everyday life activities of individuals, including family relations, social contacts, work options, economic independence, educational advancement, and cultural enrichment. Third, differential treatment in the form of unnecessary segregation is itself a form of discrimination the statute prohibits, even absent any showing that persons without disabilities received better treatment in a comparable setting. That third step is the doctrinal innovation.

The decision supplied the legal basis for the long shift of long-term services and supports away from institutional settings and toward community settings, a transformation that reshaped how states deliver care under the Medicaid service structure. The transformation was structural rather than merely doctrinal: once unjustified segregation counted as discrimination, states had a legal reason to reorient their service systems around community placement, and the reorientation touched funding, administration, and professional practice. The decision thus operated as both a liability rule and a planning signal, telling states not only what they could be sued for but how they should organize their programs going forward.

The holding is qualified, and the qualifications are as important as the rule. Community placement counts as a reasonable modification only when three conditions are met, and each condition must be stated precisely because practitioners apply them as a checklist. First, the state’s treatment professionals must determine that community placement is appropriate for the individual, which lodges the clinical judgment with professionals rather than with courts or with the individual alone. Second, the affected individual must not oppose the transfer to community placement, which preserves the autonomy of people who prefer institutional settings or who are not ready for community life. Third, the placement must be reasonably accommodated taking into account the resources available to the state and the needs of others with disabilities, which prevents the mandate from becoming an unlimited claim on state budgets. The state may also raise the fundamental-alteration defense, arguing that the requested modification would fundamentally alter the nature of its services or programs. Together, these limits mean the decision establishes a strong presumption in favor of community placement rather than an absolute command. A state that demonstrates that immediate community placement for the plaintiffs would inequitably burden its ability to serve others, for example by forcing cuts to services for persons on waiting lists, can defeat the claim.

Justice Thomas dissented, joined by Chief Justice Rehnquist and Justice Scalia, arguing that providing services cannot be discrimination absent differential treatment. On the dissent’s view, the state that confines a disabled person to an institution while providing appropriate care has not discriminated, because discrimination requires treating similarly situated people differently, and the institutionalized person has no nondisabled comparator who is treated better. The majority’s response, implicit in the holding, was that segregation itself is the harm the statute addresses: the injury lies in unjustified isolation from community life, not in a disparity between how disabled and nondisabled people are served. The disagreement thus reflects two competing conceptions of what discrimination means, and the majority’s broader conception is what gave the decision its transformative power.

Olmstead remains operative law, unaffected by the 2008 amendments, which addressed the definition of disability rather than the integration duty. The decision’s limits deserve equal emphasis, because the most common error in describing Olmstead is to present it as requiring deinstitutionalization without limits. It does not. The professional-judgment condition, the no-opposition condition, the resource constraint, and the fundamental-alteration defense together ensure that the mandate operates within the realities of state administration. A reader who understands both the rule and its limits understands the decision correctly. The decision’s relationship to the services system is what gave it practical force: by making unjustified segregation actionable as discrimination, Olmstead gave plaintiffs and advocates a legal tool for challenging institutional placements, and states responded by restructuring how long-term services are delivered. The three conditions channeled that restructuring, ensuring that community placement followed professional judgment and individual preference rather than judicial command. The fundamental-alteration defense, in turn, gave states a doctrinal vocabulary for the resource constraints they faced, keeping the mandate within administrable bounds.

How the three conditions allocate decisions

The Olmstead conditions can be read as an allocation of decision-making authority, and that reading clarifies why the mandate works in practice. The professional-judgment condition allocates the clinical question to treatment professionals: courts do not decide whether community placement is appropriate; the people who treat the individual do. This deference keeps the judiciary out of medical judgments while giving the mandate an objective trigger. The no-opposition condition allocates the autonomy question to the individual: even when professionals approve community placement, the person may decline it, and the mandate does not override that choice. This preserves the liberty of people who prefer institutional settings, a protection the dissent’s concerns about coercion implicitly demanded. The resource condition allocates the feasibility question to the state’s administration: placement must be reasonably accommodated within available resources and the needs of others, which means the mandate is implemented through budgets and waiting lists rather than by judicial decree. The fundamental-alteration defense then provides the doctrinal backstop for cases where even reasonable accommodation would change the nature of the program. Each condition thus assigns a question to the institution best suited to answer it, and the mandate’s durability comes from that division of labor as much as from its moral force.

Did Olmstead require states to close every institution?

No. The decision requires community placement only when treatment professionals approve it, the individual does not oppose it, and the placement can be reasonably accommodated within state resources and the needs of others. States may also invoke the fundamental-alteration defense, so the ruling sets a qualified standard rather than an unconditional closure order.

The three prongs as litigation issues

Each element of the Olmstead test generates its own body of disputes, and understanding the test as three separate litigation issues rather than a single rule clarifies where cases are actually decided. The first prong, the judgment of the state’s treatment professionals that community placement is appropriate, assigns decisive weight to clinical opinion. This deference to professionals has two edges. It protects states from being ordered to place individuals in community settings that clinicians consider unsafe or inappropriate, which answers the concern that the integration mandate might harm the very people it aims to help. But it also creates a potential bottleneck, because a state whose professionals systematically disfavor community placement can insulate itself from liability through the very judgments the test privileges. Litigation over the first prong therefore often becomes a battle of experts and institutional practices, examining whether professional determinations reflect genuine clinical judgment or administrative convenience. Plaintiffs challenge assessments as cursory or biased; states defend them as the product of expertise courts should not second-guess.

The second prong, the individual’s non-opposition to community placement, embodies a principle of autonomy that is easy to overlook in discussions focused on systemic reform. The statute does not authorize forced deinstitutionalization. A person who prefers institutional care, whether from fear of the unknown, attachment to a community of fellow residents, or a considered judgment about personal needs, cannot be moved against that preference under the integration mandate. This limit answers one of the persistent anxieties about the decision: that it treats institutions as categorically illegitimate and their residents as victims lacking agency. The opinion treats residents as decision-makers whose preferences constrain the remedy. In practice, the second prong requires careful attention to how preference is ascertained, particularly for individuals with cognitive disabilities whose capacity to express informed choice may be questioned. The inquiry must distinguish genuine opposition from the effects of institutionalization itself, which can suppress the ability to imagine alternatives.

The third prong, reasonable accommodation in light of resources and the needs of others, is where the fiscal reality of public administration enters the doctrine and where most contested cases are resolved. The opinion directs courts to consider the state’s overall resources and its obligations to the entire population it serves, not merely the cost of placing the individual plaintiffs. A state may show that immediate relief would be inequitable given its responsibility to others, for example where granting the plaintiffs’ request would require cutting services for persons on waiting lists or diverting funds from equally urgent needs. This is the doctrinal home of waiting-list litigation, in which plaintiffs argue that the state’s list moves too slowly to satisfy the reasonable-modification duty while states argue that the pace reflects genuine resource constraints. The fundamental-alteration defense operates here as well: where community placement on the requested terms would require restructuring the service system so profoundly as to change its nature, the state may prevail. Courts applying the third prong thus engage in a form of administrative review, assessing the reasonableness of the state’s allocation decisions without substituting their own budgetary judgment. The resulting case law is fact-intensive and jurisdiction-specific, which is why integration litigation looks so different from state to state despite resting on a single Supreme Court holding.

The dissent’s continuing influence

Justice Thomas’s Olmstead dissent lost the case but supplied a framework that organized skepticism about the integration mandate through the horizon, and it deserves treatment as a live intellectual position rather than a historical curiosity. The dissent’s core claim is definitional: discrimination means differential treatment, and providing a service in one setting rather than another is not differential treatment. On this view, the majority confused a desirable policy, community integration, with a legal wrong, discrimination, and in doing so authorized courts to supervise the administration of state service systems under the guise of enforcing an antidiscrimination norm. The concern is institutional as well as conceptual. If unjustified segregation counts as discrimination without any comparator showing worse treatment, then every allocative decision a state makes about where and how to deliver services becomes potential litigation, and judges become administrators of programs they lack the expertise and democratic accountability to run. That separation-of-powers anxiety gives the dissent its enduring appeal for critics of structural-reform litigation generally.

The majority’s response, which governed from 1999 forward, is that segregation is itself the differential treatment because it subjects persons with disabilities to a form of separation that others do not experience. The response draws on the deeper history of American antidiscrimination law, in which separate was held inherently unequal, and applies that insight to disability: unnecessary institutionalization is not a neutral service-delivery choice but a practice that reflects and reinforces the assumption that persons with disabilities belong apart. The philosophical divide between the two positions maps onto a broader debate about the nature of equality. The dissent embodies a formal conception, equality as identical treatment, under which the law’s job is to ensure that the same rules apply to everyone. The majority embodies a substantive conception, equality as genuine participation, under which the law sometimes requires different arrangements to achieve inclusion. Neither conception is self-evidently correct as an abstract matter, and reasonable people divide on which better captures the moral ideal. But as a matter of positive law, the majority’s conception won, and the integration mandate has operated on that basis since the decision. The dissent’s value for readers is diagnostic: whenever a court or commentator resists an integration claim on the ground that the state treated everyone the same, the ghost of the dissent is present, and the answer must come from the majority’s reasoning about what segregation itself communicates and inflicts.

Accommodation and the outer boundaries of coverage

The remaining decisions define the limits of the accommodation duty and the reach of the statute’s titles. Each one answers a question that recurs in practice, and each remains operative. Together they show the Court working out what the statute’s substantive commands require when they collide with competing interests: the reliance interests embodied in seniority systems, the safety interests embodied in the direct-threat defense, the administrative realities of prisons, the medical realities of asymptomatic infection, and the competitive realities of professional sports. None of these decisions was affected by the 2008 amendments, because none turned on the definition analysis Congress repudiated. They are the part of the cluster that survived the reset untouched, and they structured the application of the statute through the article’s 2015 horizon. Their endurance is itself evidence for the counter-reading developed below: the Court’s restrictive energy was spent on gateways, while the duties themselves received careful, often generous construction. They share a method: close attention to the statutory text, deference where the text delegates to regulation, and a recurring concern with the boundary between reasonable adjustment and fundamental change.

US Airways, Inc. v. Barnett (2002)

US Airways v. Barnett, 535 U.S. 391 (2002), decided 5 to 4 with Justice Breyer writing for the majority, addressed the conflict between a requested accommodation and an established seniority system. The respondent was a cargo handler who injured his back, transferred to a mailroom position as an accommodation, and then faced displacement from that position when a more senior employee sought it under the employer’s seniority system. He proposed as an accommodation that he be allowed to remain in the mailroom job as an exception to seniority. The Court held that when an accommodation would conflict with a seniority system, the accommodation is ordinarily not reasonable, because overriding seniority undermines the expectations of other employees and the consistent administration of the workplace. Seniority systems serve interests of fairness and predictability: workers order their lives around the expectation that seniority will govern assignments, transfers, and layoffs, and employers rely on the system to make personnel decisions without case-by-case favoritism. An accommodation that displaces a more senior worker disrupts those settled expectations in a way the statute’s reasonableness requirement does not demand.

The word ordinarily carries the weight of the decision: the employee may still show special circumstances that make the accommodation reasonable despite the seniority system, such as evidence that the system is riddled with exceptions or is routinely disregarded in practice. Where the employer itself does not treat the seniority system as binding, the expectations interest evaporates, and the accommodation may be reasonable after all. The burden of showing special circumstances rests on the employee, which preserves the employer’s ability to rely on a genuine, consistently applied system while preventing reliance on seniority as a pretext. Justice Scalia dissented, joined by Justice Thomas, arguing for a categorical rule that departures from neutral seniority systems can never be required, and Justice Souter dissented, joined by Justice Ginsburg, arguing for a more fact-intensive inquiry that would give less automatic weight to the seniority system. The two dissents pulled in opposite directions, which is worth noting: one would presumably have given seniority even more weight, the other less, and the majority’s ordinarily-unreasonable rule with its special-circumstances safety valve sat between them. The decision was not affected by the 2008 amendments. For the wider employment docket in which accommodation disputes arise, see the broader employment case law companion.

When may an employer lawfully choose seniority over accommodation?

Ordinarily, when the requested accommodation conflicts with an established seniority system, the Supreme Court treats the request as unreasonable because overriding seniority undermines employee expectations and consistent administration. The employee may still prevail by showing special circumstances, such as a system riddled with exceptions or one that is routinely disregarded in practice.

Special circumstances after Barnett

The special-circumstances exception is where Barnett litigation actually happens, because the presumption for seniority systems is strong enough that few plaintiffs can win without it. The majority offered two illustrative examples: cases where the employer retains discretion to make exceptions to the seniority system, and cases where the system is already so riddled with exceptions that one more would not matter. Both examples share a common logic. The presumption rests on the reliance interests that a consistently applied seniority system creates, the expectation among workers that the rules will be followed fairly and uniformly. Where the employer itself does not treat the system as binding, or where the system functions more as a guideline than a rule, those reliance interests are diminished and the accommodation claim regains its footing. The burden of showing special circumstances rests on the employee, which means the discovery phase of accommodation litigation often focuses on the employer’s actual practices: how often exceptions have been granted, to whom, for what reasons, and under what procedures. An employer whose written policy proclaims absolute seniority but whose managers routinely grant exceptions will struggle to sustain the presumption. Conversely, an employer with a documented history of consistent application holds a formidable advantage that only genuinely exceptional facts can overcome. The doctrine thus rewards institutional honesty and punishes the gap between paper rules and lived practice, a theme that recurs across employment discrimination law.

Chevron U.S.A., Inc. v. Echazabal (2002)

Chevron USA v. Echazabal, 536 U.S. 73 (2002), decided 9 to 0 with Justice Souter writing for the Court, tested the boundary of the direct-threat defense. The respondent had liver damage that would be aggravated by exposure to toxins in the employer’s refinery, and the employer refused to hire him for a position involving such exposure, citing the risk to his own health. The Equal Employment Opportunity Commission had issued a regulation permitting the direct-threat defense to be invoked where the threat is to the worker’s own health or safety, not only to others. The respondent argued that the statute authorized the defense only for threats to others and that the regulation exceeded the agency’s authority. The Court upheld the regulation as a permissible interpretation of the statute, holding that an employer may refuse to place a worker in a position that would pose a direct threat to the worker’s own health or safety. The reasoning emphasized that the statute’s structure contemplates qualification standards related to the safe performance of the job and that nothing in the text confines the threat concept to third parties. The unanimity of the decision reflected the Court’s deference to the agency’s reasonable interpretation of ambiguous statutory language: where Congress had not clearly confined the defense, the Commission’s broader reading was entitled to stand.

The result is that an employer may lawfully decline a placement that would endanger the applicant, even when the applicant is willing to assume the risk. The worker’s consent does not override the employer’s judgment, because the defense protects the employer’s interest in avoiding workplace injuries and the associated consequences, not merely the worker’s autonomy. The decision thus defines one of the outer boundaries of the accommodation duty: the duty to accommodate does not extend to placements that would threaten the health or safety of the person being accommodated. That deference principle matters beyond the threat-to-self question, because much of the statute’s employment machinery runs through agency guidance. The decision remains operative and was untouched by the 2008 amendments.

Deference and the paternalism objection

Echazabal is worth pausing over because it resolves a tension that runs through the entire statute: the line between protecting workers and patronizing them. The respondent’s position had genuine moral force. He wanted the refinery job, he understood the risk to his liver, and he argued that the law should not allow an employer, or an agency, to override his autonomous choice about what risks to accept. That argument draws on a powerful anti-paternalist tradition in disability advocacy itself, which has long insisted that persons with disabilities are entitled to make their own decisions, including risky ones, rather than being sheltered by well-meaning others. The unanimous Court rejected the argument, but the rejection is narrower than it appears. The opinion does not hold that employers may generally substitute their judgment for workers’ about acceptable risk. It holds that the agency’s regulation, permitting the direct-threat defense for threats to the worker’s own health or safety, is a permissible reading of a statute that nowhere confines the threat concept to third parties. The deference to the agency does the work: the Court sustained the regulation as a reasonable elaboration of ambiguous text, not as a freestanding judicial endorsement of paternalism. The practical consequence is that the direct-threat inquiry remains medical and individualized, focused on the actual risk the particular worker faces in the particular position, rather than on generalizations about what persons with particular conditions can handle. An employer invoking the defense must show a significant risk of substantial harm, supported by objective evidence, not merely a sincere concern. The paternalism objection thus survives as a constraint on how the defense is proved, even though it failed as an attack on the regulation’s validity.

Two 2002 decisions, two kinds of limits

Barnett and Echazabal were decided the same year, both construe Title I, and both mark outer limits of the accommodation duty, but they limit it for different reasons and in different directions. Barnett limits accommodation to protect the settled expectations of other workers: the seniority system embodies reliance interests that the statute’s reasonableness requirement does not lightly override. The limit is relational, balancing the disabled worker’s needs against the expectations of the workforce as a whole. Echazabal limits accommodation to protect safety: the direct-threat defense permits the employer to refuse a placement that would endanger the worker, even with the worker’s consent. The limit is protective, prioritizing the avoidance of injury over the worker’s preference.

The two decisions also differ in their doctrinal sources. Barnett is a judicial construction of the statutory term reasonable, with the Court itself fashioning the ordinarily-unreasonable rule and its special-circumstances exception. Echazabal is an exercise in agency deference, with the Court upholding the Commission’s reasonable interpretation of the direct-threat defense. One shows the Court making accommodation law directly; the other shows the Court letting the agency make it. Together they define the employment duty’s boundaries more completely than either does alone: accommodation yields to genuine seniority systems and to genuine safety threats, but in each case the employer must show the system or the threat is real rather than asserted. The employer’s burden of establishing genuineness is the common thread: a paper seniority system that is routinely disregarded will not defeat accommodation under Barnett’s exception, just as a speculative safety concern will not satisfy Echazabal’s direct-threat standard. Neither decision was affected by the 2008 amendments, and both remained operative guides to the accommodation analysis through the horizon.

Pennsylvania Department of Corrections v. Yeskey (1998)

Pennsylvania Dep’t of Corrections v. Yeskey, 524 U.S. 206 (1998), decided 9 to 0 with Justice Scalia writing for the Court, answered a threshold scope question for the public entity title with characteristic brevity. The respondent was a state prisoner who alleged that he had been denied admission to a motivational boot camp program because of his hypertension, and the state argued that the public entity title does not apply to state prisons at all. The opinion held that state prisons fall squarely within the public entity title, reasoning that the statutory phrase covering services, programs, or activities of a public entity is naturally read to include the administration of prisons, and that the text contains no exception for correctional facilities. The opinion’s reasoning was aggressively textual: Congress had defined public entity broadly, had listed no prison exception, and the Court would not invent one. The unanimity of the decision, with an opinion by one of the Court’s most committed textualists, gave the coverage holding particular force: even Justices skeptical of broad readings of the statute could not find a prison exemption in the text.

The Court expressly left open the constitutional question of whether Congress had validly abrogated state sovereign immunity under Section 5, a question the later Garrett, Lane, and Georgia decisions would take up. Yeskey thus decided coverage while reserving remedy: prisons are covered by Title II as a matter of statutory interpretation, but whether an inmate can recover damages from the state remained for another day. That reservation is what connects Yeskey to the immunity line, and it illustrates the cluster’s division of labor between what the statute means and whom it can be enforced against. Yeskey remains operative law on the coverage question. The reservation of the immunity question is what makes Yeskey a hinge between the coverage and immunity lines: it established that the statute reaches prisons while leaving for Garrett, Lane, and Georgia the question of what remedies inmates can obtain against the state. A reader who conflates coverage with remedy will misread Yeskey; the decision answers the first and expressly defers the second.

The correctional docket after Yeskey

Yeskey’s brevity conceals its consequences. By confirming that the public entity title reaches prisons, the decision opened a docket that grew steadily after 1998, covering the full range of accommodations that incarceration implicates. Mobility access within facilities, provision of assistive devices, access to prison programs such as education and work assignments, communication accommodations for deaf prisoners, and the administration of medical care have all been litigated under the title. The opinion’s reservation of the constitutional question also proved prescient: the section 5 issue the Court left open in 1998 became the central battleground of Garrett, Lane, and Georgia, and Georgia itself arose from a prisoner’s claims. The through-line from Yeskey to Georgia illustrates how a narrow scope holding can generate the constitutional litigation that defines a field. For correctional administrators, the practical message of Yeskey is that the statute applies with full force inside the walls, and that program eligibility criteria, from boot camps to work release, must be examined for disability discrimination like any other public program. For litigators, the decision means the threshold coverage fight is over before it begins, and cases turn on reasonable modification and fundamental alteration rather than on whether the statute applies at all.

Bragdon v. Abbott (1998)

Bragdon v. Abbott, 524 U.S. 624 (1998), decided 5 to 4 with Justice Kennedy writing for the majority, is the cluster’s principal decision on what counts as an impairment and what it means for an impairment to substantially limit a major life activity, and unlike the definition trilogy it remains operative law. The respondent was a woman with asymptomatic HIV infection who was denied dental treatment in the dentist’s office because of her infection. She sued under the public accommodations title. The majority held that asymptomatic HIV infection is a physical impairment from the moment of infection, rejecting the argument that the impairment arises only when symptoms appear, and further held that the infection substantially limits the major life activity of reproduction. The reasoning on reproduction proceeded from the medical evidence that HIV infection creates a significant risk of transmission to a partner and to a child, which the majority concluded substantially limits the major life activities associated with having children. The decision’s temporal holding, that impairment dates from infection, is its most durable contribution, because it established that the absence of symptoms does not mean the absence of disability.

Chief Justice Rehnquist dissented in part, joined by Justices Scalia and Thomas, and Justice O’Connor dissented in part separately, disputing both steps: whether an asymptomatic condition without present functional limitation qualifies as an impairment, and whether reproduction was the proper major life activity for the analysis. The partial dissents reflected disagreement about how to characterize the limitation rather than about the outcome’s framework. Bragdon was decided before the definition trilogy and was not superseded by the 2008 amendments; it remains operative law, and its reasoning on asymptomatic conditions anchors coverage for infections that have not yet produced symptoms. The decision’s method is as significant as its result: the majority treated medical evidence as the measure of impairment, deferring to clinical understanding of what the virus does in the body rather than to lay intuitions about what disability looks like. That evidentiary posture matters because asymptomatic conditions are precisely the cases where lay skepticism runs highest. Bragdon’s Title III setting also matters: as a public accommodations case, it showed that the definition analysis applied with full force outside employment, and its survival of the 2008 amendments confirms that the override targeted the trilogy’s reasoning rather than the definition project as a whole.

Asymptomatic coverage beyond HIV

Although Bragdon arose from HIV infection, its reasoning extends to the broader category of asymptomatic and episodic conditions, and that extension is the decision’s most practically important legacy. The holding that an impairment exists from the moment of infection, without symptoms, supports coverage for conditions such as epilepsy in remission, multiple sclerosis between flare-ups, and controlled diabetes, where the underlying medical condition persists even when its functional effects are managed or intermittent. The identification of reproduction as a major life activity likewise reaches beyond HIV, offering a doctrinal anchor for claims involving conditions that affect fertility, pregnancy, or childbearing. The dissents’ resistance to the impairment finding illustrates the competing intuition, that protection should track visible functional limitation, and the majority’s rejection of that intuition is what gives the opinion its continuing force for conditions the justices never considered.

PGA Tour, Inc. v. Martin (2001)

PGA Tour v. Martin, 532 U.S. 661 (2001), decided 7 to 2 with Justice Stevens writing for the majority, brought the statute to the fairway and produced the cluster’s most publicly discussed opinion. The respondent was Casey Martin, a professional golfer with Klippel-Trenaunay-Weber syndrome, a degenerative circulatory disorder that made walking a golf course painful and medically risky. He requested permission to use a golf cart during the PGA Tour’s qualifying school and tournaments, as a modification of the Tour’s walking rule. The Tour refused, arguing that it was not a public accommodation covered by the public accommodations title and that, in any event, waiving the walking rule would fundamentally alter the nature of the competition. The majority held first that the PGA Tour’s tournaments are public accommodations within the meaning of the title, rejecting the argument that the Tour’s selective membership took it outside the statute’s reach. It then held that permitting Martin to use a cart was a reasonable modification that would not fundamentally alter the game. The opinion examined the nature of championship golf and concluded that its essence is shot-making, the skill of hitting the ball from tee to green, and that walking between shots, while part of the traditional experience, is not essential to the competition in a way that makes the cart an unfair advantage. The majority emphasized the individualized inquiry the statute requires: the question is not whether carts would fundamentally alter golf in the abstract but whether this modification for this competitor would do so, and the record showed that Martin’s impairment made walking more rather than less burdensome for him.

Justice Scalia dissented, joined by Justice Thomas, arguing that the majority had misunderstood both the coverage question and the nature of the game. The dissent’s view was that the rules of a private competition define the activity, and that courts are ill suited to decide which rules are essential and which are peripheral. The dissent’s warning about courts as sports commissioners has become one of the opinion’s most quoted passages, and it deserves presentation alongside the majority’s reasoning because it captures a genuine concern about the limits of judicial competence in specialized domains. The disagreement reflected a broader divide about the judicial role in reviewing private institutions’ judgments about their own activities. The decision remains operative and supplies the framework for modification requests in competitive settings: the question is whether the change would fundamentally alter the activity, judged by what is essential to it. That framework has structured every subsequent dispute about modifications in sports and other competitive endeavors. The two-step structure of the opinion, coverage first and modification second, is itself a useful template: many Title III disputes turn on the threshold question of whether the defendant is a public accommodation at all, and PGA Tour shows that the answer can be yes even for an elite professional competition.

What did PGA Tour v. Martin decide about golf carts?

The Court held that the PGA Tour is a Title III public accommodation, that walking is not essential to championship golf, and that letting Casey Martin use a cart was a reasonable modification. The requested change did not fundamentally alter the competition because the essence of golf is shot-making.

The individualized inquiry as a method

Martin’s most transferable contribution is methodological: the fundamental-alteration question must be answered for this plaintiff and this modification, not for the activity in the abstract. The Tour argued from the general proposition that walking is part of championship golf. The majority answered with the particular facts of Casey Martin’s disorder, which made walking more burdensome for him than for his competitors, so that the cart leveled rather than tilted the field. That move from the abstract to the individual recurs across the statute. Albertsons demanded an individualized assessment of monocular vision rather than a per se rule. Barnett’s special-circumstances exception turns on the particular employer’s actual practices. Olmstead’s three-part test centers on the individual’s clinical profile and expressed preference. The pattern reveals a unifying principle beneath the cluster’s surface diversity: the statute resists categorical answers and requires decision-makers to examine the person before the court. That principle cuts against both plaintiffs and defendants who prefer bright lines. It denies plaintiffs the comfort of automatic coverage and denies defendants the comfort of automatic defenses, replacing both with a fact-intensive inquiry that rewards careful evidentiary development. For litigators, the lesson is to build the record around the individual: the specific impairment, the specific modification, the specific activity, and the specific effect of the change on what the activity essentially requires.

Title by title: where each decision sits

Mapping the fourteen decisions onto the statute’s titles reveals the cluster’s architecture. Title I, the employment title, is where the Court did its narrowing. All four definition decisions, Sutton, Murphy, Albertsons, and Toyota, arose under Title I, which is why the shrinking of who counted as disabled hit workers first and hardest. Garrett, the immunity decision that closed the damages door against states, is also a Title I case. Barnett and Echazabal, the two accommodation-boundary decisions, construe Title I as well. The employment title thus carries the full range of the Court’s restrictive holdings: the narrow definition, the failed abrogation, and the seniority and direct-threat limits on accommodation. A Title I plaintiff faced the Court at its most demanding. The concentration is not accidental: employment was the title where the statute’s duties imposed the most direct costs on defendants, and where the Court most carefully policed the boundaries of congressional power and statutory coverage.

Title II, the public entity title, is where the Court split and where it expanded. Lane and Georgia, the two abrogation decisions that came out the other way from Garrett, are Title II cases, and their as-applied and conduct-based holdings give Title II plaintiffs routes that Title I plaintiffs lack. Olmstead, the integration mandate, is a Title II decision, and its holding that unjustified segregation is discrimination ranks among the most expansive readings of any civil rights statute. Yeskey, holding that state prisons are covered public entities, is a Title II coverage decision. The public entity title thus combines the immunity map’s open doors with the cluster’s broadest substantive holding. A Title II plaintiff faced a Court that was simultaneously policing congressional power and enlarging the meaning of discrimination.

Title III, the public accommodations title, appears in two decisions, and both came out in favor of coverage. Bragdon held that asymptomatic HIV infection is a disability from the moment of infection, and PGA Tour held that a professional golf tour is a public accommodation that must grant a reasonable modification. Neither decision narrowed anything; both extended the statute’s reach into domains, infectious disease and professional sports, where coverage had been uncertain. The remaining entries sit outside the title structure: Boerne supplies the Section 5 test that governs the immunity analysis, the 2000d-7 funding waiver operates through the spending power rather than any ADA title, and Cummings, decided in 2022, addresses remedies under the Spending Clause statutes from outside this article’s horizon.

The title map explains why the namable claim of this cluster is the immunity map rather than any single merits holding. The decisive variables across the decisions are institutional: which title is invoked, whether the defendant is a state, and whether the underlying right is one the Constitution independently protects. Lawyers who internalize the title map can place any new dispute on it; those who memorize holdings without titles cannot. The map also reveals a cross-title pattern worth stating plainly: where the Court confronted the statute’s duties directly, in Olmstead, Yeskey, Bragdon, PGA Tour, Barnett, and Echazabal, it generally gave them broad scope or careful limits; where it confronted the statute’s threshold and enforcement architecture, in the definition trilogy, Toyota, and Garrett, it read narrowly. The duties were treated generously and the gateways strictly, a division that persisted across all three titles. The persistence of the division across titles is what makes it a pattern rather than a coincidence: Title I, Title II, and Title III each show the same shape, with threshold and enforcement questions answered narrowly and substantive duties answered broadly. The shape reflects the Court’s institutional role as the expositor of both statutory meaning and constitutional limits, policing the boundaries of congressional power while giving effect to the duties Congress clearly imposed. A reader who carries this shape in mind will find each new disability decision easier to place, because the first question to ask is always whether the case concerns a gateway or a duty.

A Court that read narrowly in some places and broadly in others

A common framing describes the Court of this era as uniformly hostile to the disability statute, and the immunity and definition decisions lend that framing surface plausibility. The record as a whole does not support it. The same Court that decided Garrett and the definition trilogy also decided the integration case, the prison case, and the infection case in favor of coverage, and in each of those the justices read the substantive duties broadly. The better description of the pattern is narrow readings of the definition of disability and of congressional enforcement power alongside broad readings of what the statute requires once a plaintiff is inside its protection. The definition cases constricted the gate. The immunity cases constricted the remedy against one class of defendants. But the cases on unjustified segregation, on prisons as public entities, on asymptomatic infection as a disability, and on the scope of public accommodations all expanded the law’s reach. Both halves of that pattern deserve equal weight, and the dissents in the restrictive cases deserve the same careful attention as the majorities, because the dissents in the definition trilogy supplied much of the intellectual groundwork for the 2008 override.

The complication this cluster demands is the framing of the Court as uniformly hostile to the disability statute. That framing does not survive the cases, and it does not survive the vote counts. The same Court that narrowed the definition in Sutton, Murphy, and Toyota, and that closed the damages door for Title I employment claims in Garrett, also decided Olmstead, Yeskey, and Bragdon in favor of coverage. The voting alignments make the point: Albertsons and Toyota were unanimous, Yeskey and Echazabal were unanimous, and Georgia was unanimous. A uniformly hostile Court does not produce that many 9 to 0 decisions expanding or preserving the statute’s reach. Hostility would have been simpler and more consistent; what the record shows instead is a Court applying different interpretive approaches to different questions, with results that vary accordingly.

Olmstead read Title II to prohibit unjustified segregation, the most expansive substantive holding in the field, and it did so over a dissent that would have required differential treatment. Yeskey read the public entity title to reach state prisons, where the statutory text offered no exemption, in an opinion by Justice Scalia, the author of some of the cluster’s narrowest holdings. Bragdon read the definition to cover asymptomatic infection from the moment of infection, a generous temporal holding that predated the trilogy and survived the amendments. PGA Tour read Title III to reach professional sports and required the Tour to justify its refusal to modify. Even Barnett, often cited as an employer-friendly decision, preserved the special-circumstances exception that keeps the seniority rule from becoming absolute.

The pattern is better described as narrow readings of the definition and of congressional power alongside broad readings of the substantive duties. The Court policed the boundaries of who could sue and when states could be made to pay, while giving the anti-discrimination and accommodation commands themselves a wide berth. The definition cases asked who counts, and the Court answered narrowly until Congress intervened. The immunity cases asked when Congress can subject states to damages, and the Court answered with a demanding test. But the cases asking what discrimination means, what accommodation requires, and how far the titles reach were answered generously. Majority and dissent deserve equal care in each of these decisions, because the dissents often identified the costs of the majority’s approach: Stevens in Sutton on the millions excluded, Breyer in Garrett on deference to Congress, Thomas in Olmstead on the meaning of discrimination, Scalia in PGA Tour on the judicial role. The cluster reads as a debate, not a campaign. The debate’s structure is what makes the cluster useful as precedent: because the Court reasoned narrowly where it narrowed and broadly where it broadened, later courts can apply each holding within its rationale rather than treating the cluster as a single attitude. The definition holdings were overridden because their rationale departed from legislative intent; the immunity holdings stand because their rationale was constitutional rather than statutory; the integration and accommodation holdings endure because their rationale gave the statute’s duties their full scope.

The dissents that history vindicated, and those it did not

Every major decision in this cluster divided the Court, and the dissents deserve the equal care that neutrality requires, because they map the roads not taken and occasionally predict the road Congress would later build. The division was not a simple liberal-conservative split: Albertsons, Toyota, Yeskey, Echazabal, and Georgia were unanimous, and the divided decisions cut across the usual alignments. Stevens wrote the Lane majority and dissented in Sutton and Garrett; Scalia wrote Yeskey and Georgia and dissented in Lane, Barnett, and PGA Tour; O’Connor wrote the Sutton, Murphy, and Toyota majorities and joined both sides of the immunity split. The alignments followed the doctrinal questions rather than fixed blocs, which is why the cluster rewards issue-by-issue analysis over ideological generalization. Justice Stevens’s dissents in Sutton and Murphy, the latter joined by Justice Breyer, argued that the majority had excluded from the statute’s protection millions of people Congress meant to cover. The 2008 amendments vindicated that argument in the most direct way available: Congress named Sutton and its companion cases in its findings and repudiated their approach. A dissent that becomes the legislative majority’s stated rationale is the rare dissent that wins after losing, and Stevens’s definition dissents are the clearest example in the cluster.

The vindication was selective, which is itself instructive. Justice Breyer’s dissent in Garrett, joined by Justices Stevens, Souter, and Ginsburg, argued that Congress was entitled to deference in assessing the legislative record of state discrimination and that the majority had conducted too searching an independent audit. No congressional override followed; Garrett remains operative law, and the Title I damages door against states remains closed. The contrast with Sutton shows that legislative correction is not automatic: Congress intervened where it perceived a definitional error affecting millions of workers, but left the federalism-sensitive immunity holding alone. The difference in treatment reflects the different constitutional stakes: rewriting a statutory definition is an ordinary legislative act, while overriding a Section 5 immunity holding would require confronting the Court’s constitutional reasoning directly. A reader who assumes every questionable decision gets overridden will misread the field.

Other dissents illuminate the majority’s choices without having been vindicated. Justice Thomas’s Olmstead dissent, joined by Chief Justice Rehnquist and Justice Scalia, offered the differential-treatment conception of discrimination that the majority rejected; the integration mandate’s transformative power comes precisely from the majority’s refusal of that narrower view. Justice Scalia’s PGA Tour dissent, joined by Justice Thomas, warned against courts deciding which rules of a private competition are essential; the majority’s essence-of-the-game analysis became the governing framework anyway. And Justice Scalia’s authorship of the unanimous Yeskey opinion, holding that prisons are covered, sits alongside his dissents as a reminder that the Justices followed the text where they found it clear and divided where they did not. The dissents, taken together, keep the majorities honest: each majority holding in this article was contested, reasoned, and decided over articulated disagreement. The equal care this account gives to majority and dissent reflects the reality that the dissents often supplied the vocabulary for later developments. Stevens’s millions-excluded argument became Congress’s stated rationale; Breyer’s deference argument in Garrett remains the principal critique of the majority’s record-auditing approach; Thomas’s differential-treatment argument in Olmstead remains the clearest statement of the integration mandate’s conceptual alternative. A reader who knows only the holdings knows the law; a reader who knows the dissents knows why the law is contested. The dissents also serve a forward-looking function: they preserve, in the reports, the arguments that a future Congress or a future Court might adopt. Stevens’s Sutton dissent did exactly that work, waiting nine years for legislative vindication. Breyer’s Garrett dissent waits still, a standing invitation to revisit the pattern requirement’s application. The cluster’s dissents are thus not merely the losing side’s consolation but part of the law’s unfinished business.

The votes tell a story

The vote lineups across the fourteen decisions trace the fault lines of the era with unusual clarity. The definition trilogy split seven to two, seven to two, and nine to zero, with the unanimity of Albertsons showing how broadly the mitigating-measures principle had been accepted before Congress rejected it. Toyota was nine to zero, the restrictive era’s high-water mark commanding the full Court. The immunity cases split five to four, five to four, and nine to zero: division where prophylaxis was at issue, unanimity where the conduct independently violated the Constitution. Olmstead split six to three, the integration mandate carried by a coalition that crossed the usual alignments. The duties cases scattered across the full range, nine to zero for the prison scope question and the threat-to-self regulation, five to four for seniority and asymptomatic infection, seven to two for the golf cart. Two patterns emerge from the numbers. First, unanimity clustered around textually clear questions, whether prisons are public entities, whether the statute permits the threat-to-self regulation, whether conduct violating the Fourteenth Amendment supports abrogation, while division clustered around open-ended judgments, how demanding the definition should be, how much prophylaxis the enforcement power allows, how to weigh seniority against accommodation. Second, the lineups do not map neatly onto a simple pro- versus anti-disability divide. The justices who formed the restrictive majorities in the definition and immunity cases supplied votes for the broad holdings in Olmstead, Yeskey, and Georgia. The numbers thus confirm the article’s counter-reading: the pattern is differentiated rather than hostile, with the Court narrowing thresholds and broadening duties, and the vote counts show that differentiation operating justice by justice rather than bloc by bloc.

What operative status means for practice

The doctrine ledger below assigns each decision a current operative status, and the statuses require careful handling because they differ in kind. Four decisions are superseded in whole or in substantial part: Sutton, Murphy, and Toyota were rejected by the 2008 amendments, and Albertsons was rejected as to its mitigating-measures analysis. These decisions hold historical rather than operative significance, which means they must never be cited as authority for post-2008 conduct. The recurring error the brief warns against, treating the pre-2009 definition cases as authority, is exactly this mistake. But historical does not mean irrelevant: the amendments are prospective only, so claims arising from conduct before January 1, 2009 are still analyzed under Sutton and Toyota. A practitioner with a pre-2009 claim works in the old world, and must know its rules. The ledger’s value lies in making these distinctions visible at a glance, so that no reader mistakes a superseded holding for a live one or overlooks the surviving portions of partially superseded decisions.

The remaining decisions are operative within this article’s horizon, but several carry limits that are easy to overread. Garrett is operative only for money damages under Title I; it says nothing about injunctive relief and nothing about other titles. Lane is operative only as applied to the access-to-courts class of cases; it is not a blanket validation of Title II abrogation. Georgia is operative for its narrow holding that abrogation is valid where the conduct itself violates Section 1 of the Fourteenth Amendment, and it expressly leaves the Title II-only question open. Olmstead is operative subject to its three conditions and the fundamental-alteration defense; it does not order unlimited deinstitutionalization. Barnett is operative subject to the special-circumstances exception. Echazabal, Yeskey, Bragdon, and PGA Tour are operative without the qualifications that complicate the others, though each is confined to its holding: the threat-to-self regulation, prison coverage, asymptomatic infection from the moment of infection, and the fundamental-alteration framework for competitive modifications. Cummings, decided April 28, 2022, sits outside the horizon entirely: it may be noted with explicit dates to complete the record on remedies, but it must never be described as within-horizon law. The verification discipline for this cluster is therefore exacting: confirm every citation, year, and holding, and mark clearly which decisions were superseded before anything is published.

Two congressional judgments: 1990 and 2008

The cluster is bookended by two legislative judgments, eighteen years apart, that frame everything the Court did in between. The first is the 1990 enactment itself: Public Law 101-336, signed July 26, 1990 by President George H. W. Bush, passed the House 377 to 28 on July 12, 1990 and the Senate 91 to 6 on July 13, 1990. Those margins reflect a bipartisan consensus that disability discrimination required a comprehensive federal response extending beyond the funding-condition approach of Section 504. The 1990 Congress wrote a broad definition, organized duties into three titles, and included a clear abrogation of state immunity, expecting the courts to give each provision full effect. The expectation was not unreasonable: the text was ambitious, the majorities were large, and the remedial purpose was stated plainly. What the 1990 Congress did not anticipate was how the courts would construe the definition it had written, and the eighteen years that followed tested each of its expectations in turn.

The second judgment is the 2008 override: Public Law 110-325, signed September 25, 2008 by President George W. Bush, effective January 1, 2009, passed by unanimous consent in the Senate on September 11, 2008 and by voice vote in the House on September 17, 2008. The eighteen-year gap between the two judgments is the space in which the definition era happened. The 2008 Congress looked at what the Court had done with the 1990 definition and concluded, again with bipartisan consensus, that the judiciary had narrowed the protected class beyond what the 1990 Congress intended. The findings named Sutton and its companion cases and rejected Toyota’s demanding standard, restoring the broad construction the 1990 Congress had written.

The pairing matters because it shows legislative intent operating at two moments rather than one. The 1990 intent set the statute’s design; the Court’s readings departed from it; the 2008 intent corrected the departure. A reader who knows only the 1990 passage understands the statute’s ambition but not why the definition had to be rewritten. A reader who knows only the 2008 override understands the correction but not what was corrected. The cluster requires both, and the series thesis, litigation as the stage where a statute’s reach is priced and where a congressional override reset the field once already, is simply the relationship between these two judgments stated as a principle. The naming of Sutton and its companion cases in the 2008 findings deserves a final note: by identifying the rejected decisions explicitly, Congress denied them even the afterlife of distinguished-but-unrepudiated precedent. The trilogy cannot be revived by analogy or extended to new facts, because the legislature has marked the entire line as error. That is the strongest form of override, and it is why the definition era is closed history rather than dormant doctrine.

The statute’s reach, priced in court

Litigation is the stage where a statute’s reach is priced, and this cluster shows the pricing in unusual clarity. The pricing metaphor deserves a final unpacking, because it captures what courts do to statutes over time. A statute’s text sets an initial price, the cost of compliance and the value of its protections, but each decision reprices it: Sutton and Toyota raised the price of proving disability until few could pay it; Garrett priced state-entity employment liability at zero for damages; Lane and Georgia repriced it upward for court access and unconstitutional conduct; Olmstead repriced segregation itself as a liability. Congress, observing the repricing, reset the definition’s price by statute in 2008. The metaphor holds because prices, like doctrines, are set by institutions interacting over time rather than by any single actor’s decree. The definition cases priced the threshold of coverage so high that Congress reset the field once already, rewriting the definition in 2008 and rejecting the Court’s readings by name. That reset is the series thesis made concrete: the courts price the statute’s reach through interpretation, and when the price departs too far from what the legislature intended, the legislature can reprice it by amendment. The 2008 override stands as the clearest example of that dynamic, with findings that name the rejected decisions and a legislative consensus broad enough to pass both chambers without a recorded vote.

The immunity cases priced the remedy, sorting plaintiffs by title and by constitutional right until the map above emerged. The price of damages against a state varies with the doctrinal coordinates: closed for Title I employment claims, open for Title II access-to-courts claims, open wherever the conduct is independently unconstitutional, and uncertain for Title II-only statutory claims. The variation is the point: unlike the definition cases, where Congress found the price uniformly too high and reset it, the immunity pricing was left in place, which means the map describes the law as it stands within this article’s horizon. The Section 504 funding waiver priced an alternative, offering a spending-power route that avoids the Section 5 analysis entirely. The integration and accommodation cases priced the duties themselves, and there the price came out in favor of breadth: segregation counts as discrimination, prisons are covered, asymptomatic conditions count from infection, and even professional sports must justify a refusal to modify. The breadth of the duties pricing is what makes the gateway pricing so striking by contrast: the same Court that read the definition narrowly read the duties generously, which suggests the narrowing reflected specific interpretive and constitutional judgments rather than a general reluctance to enforce the statute. Readers who want to keep these holdings organized across the cluster may use a legislation study notebook to track each decision’s title, holding, and operative status. The through line is that Congress and the Court share authorship of the statute’s meaning, and when the Court’s reading departed too far from legislative intent, Congress demonstrated that the field could be reset. Shared authorship is the cluster’s deepest lesson: neither the judiciary nor the legislature writes disability law alone, and each institution’s work is intelligible only in light of the other’s. The reset did not end the conversation: the immunity map’s open question, the integration mandate’s application, and the accommodation duty’s boundaries were worked out in the lower courts within the frameworks these fourteen decisions established. What the cluster teaches, above all, is that a statute’s reach is never fixed by its text alone. It is priced in litigation, title by title and doctrine by doctrine, and the price can be renegotiated when the legislature decides the courts have set it wrong.

What the cluster teaches about reading statutes

Stepping back, the fourteen decisions form a natural experiment in statutory interpretation, because the same Court, often the same justices, read the same statute narrowly in some cases and broadly in others. The variable is not ideology in any simple sense. Justices who voted to narrow the definition in Sutton voted to broaden the duty in Olmstead; the author of the unanimous scope holding in Yeskey dissented from the broad reading of public accommodations in Martin. What explains the pattern is the interaction between interpretive method and the question presented. On threshold questions, who counts as disabled, can Congress subject states to damages, the Court applied constraining frameworks: a textualism that demanded restrictive content for open-ended terms, and a federalism-inflected enforcement-power doctrine that demanded documented constitutional violations before abrogation. On substantive questions, what does the duty require, whom does the title reach, the Court applied the statute’s own logic: an antidiscrimination command that reaches segregation as such, a public-entity definition without a prison exception, a reasonable-modification duty bounded by fundamental alteration rather than by judicial timidity.

The dissents play a distinctive role in this experiment. In the definition cases, the dissents lost the battle but won the war, supplying the reasoning Congress adopted in 2008. In the immunity cases, the dissents lost and stayed lost, but they clarified the stakes of the majorities’ holdings by forcing the majorities to defend the limits they imposed. In Olmstead, the dissent supplied the most rigorous statement of the formal-equality position, which structured resistance to integration claims through the horizon. A reader who studies only majorities will understand what the law is; a reader who studies the dissents with equal care, as this article has attempted, will understand why the law took the shape it did and where its pressure points lie. That is the deeper reason the brief for this article insists on presenting majority and dissent with equal care. The holdings are the answers, but the disagreements are the map of the terrain on which future cases will be fought.

How the lines fit together

The cluster’s architecture rewards one final restatement. The definition line narrowed the protected class until Congress intervened in 2008, and its holdings are historical except for the surviving safety-regulation prong of Albertsons. The immunity line produced a map rather than a rule, with Garrett closing the employment title to damages against states, Lane opening the public entity title for cases implicating court access, and Georgia opening it wherever the conduct independently violates the Fourteenth Amendment, plus the funding-condition waiver as a parallel route. The integration line produced the most consequential substantive holding in the field, qualified by a three-part test and a fundamental-alteration defense that keep it from becoming an unlimited deinstitutionalization command. The duties line produced a set of durable, fact-sensitive holdings on seniority, direct threat, prisons, asymptomatic conditions, and public accommodations. The series thesis runs through all four lines. Litigation is the stage where a statute’s reach is priced, and the price the Court set in the definition cases proved high enough that Congress reset the field once already. Whether Congress will need to reset it again depends on how the remaining open questions, particularly the Title II-only abrogation question Georgia left open, are answered in the lower courts. The definition reset of 2008 shows what a congressional correction looks like when it comes: findings that name the errors, operative provisions that reverse each doctrinal move, and a coalition broad enough to move the bill by unanimous consent and voice vote. No comparable correction has been attempted for the immunity map, and none may be needed if the lower courts fill in the open domain sensibly under Lane’s application-specific approach. But the possibility remains, and it is the reason practitioners should watch the Title II-only cases as the most likely source of the next major development in this field. The through-line of the entire cluster is that statutory meaning is not fixed at enactment but priced through litigation, and the price can be renegotiated, by courts refining doctrine case by case or by Congress resetting the field in a single statute. A reader who has followed the full arc should be able to do three things. First, trace how the Court narrowed the definition of disability through the 1999 trilogy and the 2002 Toyota decision, explain the catch-22 that narrowing produced, and describe how the 2008 amendments overrode it while leaving the safety-regulation prong of Albertsons intact. Second, explain the split result on state sovereign immunity: why the employment title does not validly abrogate for damages, why the public entity title does as applied to court access, and why abrogation holds wherever the conduct independently violates the Fourteenth Amendment, with the Title II-only question reserved. Third, state the Olmstead holding in its qualified form, unjustified segregation as discrimination subject to the three-part test and the fundamental-alteration defense, and identify it as the legal foundation for the shift of long-term services toward community settings. Those three competencies are the working knowledge this cluster demands.

The doctrine ledger

The table below collects the fourteen decisions in a single view: each decision with its year, the title construed, the holding in one sentence, whether Congress later overrode it, and its current operative status. The overridden column and the status column must be read together, because override determines status: the definition decisions are historical precisely because the 2008 amendments rejected them, while the immunity, integration, and accommodation decisions are operative precisely because no override touched them. Albertsons is the instructive mixed case, partly superseded and partly operative, and Boerne is the structural entry, a test rather than a disability holding. Readers should treat the one-sentence holdings as summaries only; the full holdings are stated in the sections above, with majority and dissent given equal care. The ledger is the article’s findable artifact: a practitioner who needs the year, the title, or the operative status of any decision in the cluster can locate it here without working through the full analysis, and a reader verifying a citation can confirm each element against the table before relying on it.

decision	year	title construed	holding in one sentence	overridden by Congress	current operative status
Sutton v. United Air Lines, Inc.	1999	Title I	Disability is assessed with reference to mitigating measures, and the petitioners also failed the regarded-as prong.	Yes, by the ADA Amendments Act of 2008	Historical only, superseded
Murphy v. United Parcel Service, Inc.	1999	Title I	Hypertension is assessed in its medicated state, and failure to meet DOT certification does not show the employer regarded the worker as disabled.	Yes, by the ADA Amendments Act of 2008	Historical only, superseded
Albertsons, Inc. v. Kirkingburg	1999	Title I	Monocular vision is not per se a disability and is assessed individually including natural adaptations, and employers may rely on binding federal safety regulations.	In part, mitigating-measures prong rejected by the 2008 amendments	Partly operative, safety-regulation holding survives
Toyota Motor Mfg., Ky. v. Williams	2002	Title I	Substantially limits is strictly construed to require prevention or severe restriction of activities of central importance to most people’s daily lives, with permanent or long-term impact.	Yes, by the ADA Amendments Act of 2008	Historical only, superseded
City of Boerne v. Flores	1997	None, Section 5 test	Section 5 power is remedial only, and enforcement legislation must satisfy congruence and proportionality.	No	Operative doctrinal test
Board of Trustees of the Univ. of Alabama v. Garrett	2001	Title I	Title I does not validly abrogate state sovereign immunity for money damages because Congress showed no pattern of unconstitutional state employment discrimination.	No	Operative, limited to damages under Title I
Tennessee v. Lane	2004	Title II	Title II validly abrogates immunity as applied to the class of cases implicating the fundamental right of access to the courts.	No	Operative, application-specific
United States v. Georgia	2006	Title II	Title II validly abrogates immunity where the conduct actually violates Section 1 of the Fourteenth Amendment, with claim-by-claim remand and the Title II-only question left open.	No	Operative, narrow holding
Olmstead v. L.C.	1999	Title II	Unjustified institutional isolation is discrimination, subject to a three-part reasonable-modification test and the fundamental-alteration defense.	No	Operative law
US Airways, Inc. v. Barnett	2002	Title I	A seniority-system conflict ordinarily makes an accommodation unreasonable, subject to a special-circumstances showing by the employee.	No	Operative law
Chevron U.S.A., Inc. v. Echazabal	2002	Title I	The EEOC threat-to-self regulation is valid, so an employer may refuse a placement posing a direct threat to the worker’s own health or safety.	No	Operative law
Pennsylvania Dep’t of Corrections v. Yeskey	1998	Title II	State prisons are public entities under Title II, with the Section 5 constitutional question left open.	No	Operative on coverage
Bragdon v. Abbott	1998	Title III	Asymptomatic HIV is a physical impairment substantially limiting reproduction, with the impairment dating from infection.	No	Operative law
PGA Tour, Inc. v. Martin	2001	Title III	The PGA Tour is a public accommodation and a golf cart is a reasonable modification that does not fundamentally alter golf, whose essence is shot-making.	No	Operative law

Frequently Asked Questions

Q: What did Olmstead decide under the Americans with Disabilities Act?

Olmstead v. L.C., 527 U.S. 581 (1999), held that unjustified institutional isolation of persons with disabilities is a form of discrimination under Title II. The six to three majority, written by Justice Ginsburg, reasoned that unnecessary segregation perpetuates unwarranted assumptions about the capabilities of people with disabilities and severely diminishes everyday life activities, including family relations, social contacts, and work options. The holding is qualified by a three-part test: the state’s treatment professionals must consider community placement appropriate, the individual must not oppose it, and the placement must be reasonably accommodated in light of state resources and the needs of others. States may invoke the fundamental-alteration defense. The decision remains operative law and supplied the legal foundation for the long shift of long-term services toward community settings.

Q: Can state employees sue a state under the Americans with Disabilities Act?

It depends on the title invoked and the relief sought. Under Board of Trustees of the University of Alabama v. Garrett, 531 U.S. 356 (2001), Title I does not validly abrogate state sovereign immunity, so a state employee cannot recover money damages from a state employer under the employment title, though prospective relief against state officers remains available. Under Title II, Tennessee v. Lane, 541 U.S. 509 (2004), permits damages for cases implicating the fundamental right of access to the courts, and United States v. Georgia, 546 U.S. 151 (2006), permits damages wherever the alleged conduct independently violates the Fourteenth Amendment. A further route exists under section 504 of the Rehabilitation Act, where states accepting federal funds have waived immunity by statute. Careful pleading of the correct title is essential.

Q: What did PGA Tour v. Martin hold about the Americans with Disabilities Act?

PGA Tour, Inc. v. Martin, 532 U.S. 661 (2001), held that the PGA Tour’s tournaments are public accommodations under Title III and that permitting Casey Martin to use a golf cart was a reasonable modification that did not fundamentally alter the competition. The seven to two majority, written by Justice Stevens, concluded that the essence of championship golf is shot-making rather than walking between shots, and that the individualized inquiry the statute requires showed the cart gave Martin no unfair advantage given his degenerative circulatory disorder. Justice Scalia dissented, joined by Justice Thomas, warning against courts deciding what is essential to a sport. The holding was not disturbed by the 2008 amendments and remains operative law for the fundamental-alteration framework.

Q: Does the Americans with Disabilities Act cover prisons?

Yes. Pennsylvania Department of Corrections v. Yeskey, 524 U.S. 206 (1998), held unanimously that state prisons are public entities under Title II, so the statute’s protections extend to prison programs, services, and activities. Justice Scalia wrote for the nine to zero Court, reasoning that the statutory phrase covering services, programs, or activities of a public entity naturally includes prison administration and contains no correctional exception. The opinion expressly left open the constitutional question whether Congress had power under section 5 of the Fourteenth Amendment to apply the title against the states in this setting. The scope holding has never been disturbed and remains operative, supporting litigation over accommodations in correctional settings from mobility access to program eligibility.

Q: Does seniority beat accommodation under the Americans with Disabilities Act?

Ordinarily, yes. US Airways, Inc. v. Barnett, 535 U.S. 391 (2002), held that where a requested accommodation would conflict with the rules of a bona fide seniority system, that conflict ordinarily shows the accommodation is not reasonable, because seniority systems create expectations of fair and uniform treatment. The five to four majority, written by Justice Breyer, preserved an escape route: the employee may show special circumstances making the exception reasonable in the particular case, such as employer discretion to depart from the system or a system already full of exceptions. Justice Scalia would have barred such exceptions categorically, while Justice Souter favored a more fact-intensive approach. The holding was untouched by the 2008 amendments and remains operative.

Q: Is HIV a disability under the Americans with Disabilities Act?

Yes. Bragdon v. Abbott, 524 U.S. 624 (1998), held that asymptomatic HIV infection is a physical impairment from the moment of infection and that it substantially limits the major life activity of reproduction. The five to four majority, written by Justice Kennedy, rejected the argument that the impairment arises only when symptoms appear and relied on medical evidence that infection creates a significant risk of transmission to a partner and to a child. The dissenters disputed whether an asymptomatic condition qualifies as an impairment and questioned the choice of reproduction as the relevant major life activity. The holding was not disturbed by the 2008 amendments and remains operative, supporting coverage for asymptomatic conditions generally rather than only for conditions producing present functional incapacity.

Q: What is the integration mandate in Americans with Disabilities Act cases?

The integration mandate is the rule from Olmstead v. L.C., 527 U.S. 581 (1999), that unjustified segregation of persons with disabilities in institutions constitutes discrimination under Title II. It requires community placement as a reasonable modification when three conditions are met: the state’s treatment professionals determine that community placement is appropriate, the individual does not oppose it, and the placement can be reasonably accommodated given the state’s resources and the needs of others with disabilities. The mandate is not unlimited. States may invoke the fundamental-alteration defense by showing that immediate placement would inequitably disrupt their service systems. The rule remains operative law and underpins the decades-long rebalancing of long-term services from institutions toward community settings.

Q: Why did Tennessee v. Lane come out differently from Garrett under the Americans with Disabilities Act?

The difference lies in the constitutional right behind each claim and the legislative record Congress assembled. Garrett concerned Title I employment claims, where disability classifications receive only rational-basis review, so Congress needed a substantial record of unconstitutional state employment discrimination; the five to four majority found that record too thin to satisfy the congruence and proportionality test. Lane concerned Title II claims implicating the fundamental right of access to the courts, which commands heightened protection, so a less extensive record sufficed, and Congress had documented the exclusion of people with disabilities from courthouses. The stronger the underlying constitutional right, the less demanding the abrogation inquiry. Lane’s holding is application-specific rather than a blanket validation of Title II.

Q: What did Sutton v. United Air Lines decide about mitigating measures?

Sutton v. United Air Lines, Inc., 527 U.S. 471 (1999), held 7 to 2 that disability must be assessed with reference to mitigating measures such as medication or corrective devices. Justice O’Connor’s majority ruled that a person whose impairment is controlled by treatment is evaluated in the treated state, and that the petitioners also failed the regarded-as prong. Justice Stevens dissented, joined by Justice Breyer, arguing that the ruling excluded millions Congress meant to protect. The findings of the 2008 amendments explicitly rejected Sutton and its companion cases, so the mitigating-measures rule holds historical rather than operative significance after the amendments.

Q: How did Toyota Motor Manufacturing v. Williams interpret “substantially limits”?

Toyota Motor Mfg., Ky. v. Williams, 534 U.S. 184 (2002), construed the phrase strictly in a unanimous opinion by Justice O’Connor. To be substantially limited in performing manual tasks, an individual must have an impairment that prevents or severely restricts performance of activities of central importance to most people’s daily lives, and the impact must be permanent or long term. This demanding standard raised the bar for all plaintiffs, including those with unmitigated impairments. Congress rejected the standard in the findings of the 2008 amendments, so Toyota’s interpretation holds historical rather than operative significance after the amendments.

Q: What did Albertsons v. Kirkingburg add to the 1999 definition trilogy?

Albertsons, Inc. v. Kirkingburg, 527 U.S. 555 (1999), decided 9 to 0 with Justice Souter writing, added two elements. First, monocular vision is not a disability per se; the assessment is individualized and must account for mitigating measures, including the body’s own natural adaptations. Second, an employer may rely on binding federal safety regulations in employment decisions. The first element fell with Sutton under the 2008 amendments, whose findings rejected Sutton and its companion cases. The second element was not a mitigating-measures holding and survived, leaving Albertsons partly superseded and partly operative.

Q: What happened to the “regarded as” argument in Sutton and Murphy?

It failed. In Sutton, the twin-sister pilots argued in the alternative that United Air Lines regarded them as disabled because its vision requirement excluded them from pilot positions. The seven to two majority held that the requirement showed only that the airline regarded them as unable to satisfy the visual standard for the single job of global airline pilot, not as substantially limited in the major life activity of working across a broad class of jobs. In Murphy, the majority similarly held that the employer’s view that the mechanic could not meet the Department of Transportation’s health certification for commercial driving did not establish that the employer regarded him as disabled. Both rulings narrowed the “regarded as” prong sharply and were later displaced by the 2008 amendments.

Q: What doctrinal test did the Court apply in Board of Trustees v. Garrett?

The Court applied the congruence-and-proportionality test from City of Boerne v. Flores, 521 U.S. 507 (1997). Under Boerne, Congress’s Section 5 power is remedial, not substantive, so enforcement legislation must show congruence and proportionality between the injury addressed and the means chosen. In Garrett, 531 U.S. 356 (2001), the 5 to 4 majority led by Chief Justice Rehnquist found that Congress had not documented a pattern of unconstitutional state employment discrimination against people with disabilities. Because disability classifications receive only rational-basis review, the Title I damages remedy was disproportionate to the underlying constitutional violations, and the abrogation failed. The holding covers money damages under Title I only.

Q: What question about Title II did United States v. Georgia expressly leave open?

United States v. Georgia, 546 U.S. 151 (2006), held unanimously that Title II validly abrogates state sovereign immunity where the alleged conduct itself violates section 1 of the Fourteenth Amendment, and it ordered lower courts to decide the constitutional question claim by claim on remand. The opinion expressly left open whether abrogation is valid for Title II-only conduct, meaning conduct that violates the statute’s prophylactic requirements without independently violating the Constitution. Justice Scalia’s majority and Justice Stevens’s concurrence both emphasized that the open question was not prejudged. That reservation is the principal remaining uncertainty in the immunity map, and lower courts worked through it within the horizon using the application-specific approach of Tennessee v. Lane.

Q: What three conditions make community placement a reasonable modification under Olmstead?

Olmstead v. L.C., 527 U.S. 581 (1999), conditions the reasonable-modification duty on three findings. First, the state’s treatment professionals must determine that community placement is appropriate for the individual. Second, the individual must not oppose the transfer to a community setting. Third, the placement must be reasonably accommodated taking into account the resources available to the state and the needs of others with disabilities. States may additionally raise the fundamental-alteration defense. These conditions are why Olmstead sets a qualified standard rather than an unconditional deinstitutionalization order, and they remain operative law.

Q: What exception can overcome a seniority system under US Airways v. Barnett?

In US Airways v. Barnett, 535 U.S. 391 (2002), the majority held that a seniority-system conflict ordinarily makes an accommodation unreasonable, but the word ordinarily preserves an exception. The employee may show special circumstances demonstrating that the accommodation remains reasonable despite the system. Examples the opinion contemplates include a seniority system riddled with exceptions or one that management routinely disregards in practice, either of which undermines the expectations and consistency interests that justify deference to seniority. The exception keeps the rule from becoming absolute while leaving the burden on the employee to establish the special facts.

Q: What did Chevron v. Echazabal hold about the EEOC’s threat-to-self regulation?

Chevron USA v. Echazabal, 536 U.S. 73 (2002), unanimously upheld the Equal Employment Opportunity Commission regulation permitting employers to consider threats to the worker’s own health or safety. Justice Souter’s opinion held that the statute’s direct-threat defense is not confined to dangers the worker poses to others; the agency reasonably interpreted it to encompass danger to the worker himself. An employer may therefore lawfully refuse a placement that would endanger the applicant even when the applicant is willing to accept the risk. The decision remains operative and was untouched by the 2008 amendments.

Q: What is the fundamental-alteration defense in disability case law?

The fundamental-alteration defense allows a covered entity to resist a requested modification by showing that it would fundamentally alter the nature of the service, program, or activity. In Olmstead, the defense protects states from integration claims where immediate community placement would inequitably disrupt service systems given overall resources and the needs of others. In PGA Tour v. Martin, the defense framed the central dispute: the Tour argued that waiving its walking rule would fundamentally alter championship golf, and the majority rejected that showing after examining what the sport essentially tests. The defense appears across titles and functions as the outer boundary of the reasonable-modification duty. The defendant bears the burden of establishing it, and courts apply it through an individualized rather than abstract inquiry.

Q: What was Justice Thomas’s dissenting theory in Olmstead?

Justice Thomas, joined by Chief Justice Rehnquist and Justice Scalia, argued that discrimination requires differential treatment, meaning treating a person with a disability worse than others are treated in a comparable setting. On that view, Georgia provided institutional care to the plaintiffs rather than denying them a service available to others, so the dispute concerned the level or location of services rather than discrimination. The dissent treated the majority as having converted a funding and administration question into a civil rights violation. The majority answered that unnecessary segregation is itself the differential treatment, because it subjects persons with disabilities to a separation that persons without disabilities do not face. The disagreement reflects competing conceptions of equality that structured integration litigation through the horizon.

Q: How should readers treat Supreme Court holdings that Congress later overrode?

As historical authority that explains the amended law without governing it. The definition trilogy of 1999 and Toyota Motor Manufacturing v. Williams (2002) were expressly rejected by the findings of the ADA Amendments Act of 2008, so they do not state the rule for conduct occurring on or after the January 1, 2009 effective date. They remain relevant for three reasons: claims arising from earlier conduct are still analyzed under the old definitions because the amendments are prospective only; the opinions supply the background that makes the amendments intelligible; and the dissents supplied much of the reasoning Congress adopted. Albertsons illustrates the need for precision, since only its mitigating-measures prong was overridden while its safety-regulation holding survives. Always check the date of the conduct first.