Congress passed a statute in 1964 that told hotels, restaurants, employers, unions and federal grant recipients what they could no longer do. It did not, and could not, tell the courts what the words meant when applied to a testing program at a power plant, a supervisor’s conduct on an oil rig, a paycheck issued nineteen years after the decision that set its amount, or a firing that the employer explained by pointing at an employee’s participation in a recreational softball league. The Civil Rights Act Supreme Court cases are where those questions were answered, and the answers moved in both directions. The Court upheld the act’s core against constitutional attack within six months of enactment, then spent five decades expanding some provisions, narrowing others, and prompting Congress to write two of its own constructions back out of the statute.
The constitutional question came first and was settled fastest. Congress rested the public accommodations title on its power to regulate interstate commerce rather than on the Fourteenth Amendment, a choice that struck many readers at the time as a lawyer’s evasion of the obvious constitutional home for a civil rights statute. It was not an evasion. It was a direct response to a holding the Court had issued eighty-one years earlier, and the two decisions handed down on December 14, 1964 vindicated it completely. Everything after that is statutory construction rather than constitutional law, which is why the later cases could be, and were, reversed by ordinary legislation.
This article walks the decisions in sequence, states each holding in one sentence before discussing what it did, and marks the ones Congress overrode. The organizing observation, and the thing that makes the sequence intelligible rather than a pile of case names, is what this series calls the three-round pattern: the meaning of the Civil Rights Act of 1964 has moved through a round of judicial expansion, a round of judicial narrowing, and a round of legislative restoration, and almost every confident public claim about what the act “means” is really a claim about which round the speaker stopped reading at. A person who stopped after Griggs v. Duke Power Co. describes a statute with a broad effects test and no damages. A person who stopped after 1989 describes a statute with a weak effects test and no damages. A person who stopped after 1991 describes a statute with a codified effects test, capped damages and jury trials. All three descriptions were accurate on the date the reader stopped, and only the third is accurate afterward.
The authority Congress claimed and why the choice determined the litigation
The Civil Rights Act of 1964, Public Law 88-352, was signed on July 2, 1964, appears at 78 Stat. 241, and is codified principally at 42 U.S.C. sections 2000a and following. Its eleven titles do not all rest on the same constitutional authority, and the differences among them explain why the litigation histories of Title II, Title VI and Title VII look nothing alike.
The problem Congress had to legislate around was the Civil Rights Cases, 109 U.S. 3 (1883), in which the Court held that the Fourteenth Amendment’s enforcement power reaches state action and not the conduct of private individuals, and struck down the public accommodations provisions of the Civil Rights Act of 1875 on that basis. That holding had never been overruled. Legislating against private discrimination in hotels and restaurants under Section 5 of the Fourteenth Amendment in 1964 would have meant asking the Court to overrule an eighty-year-old precedent as the first order of business, with the entire statute riding on the answer. The Kennedy administration’s Justice Department, and the House Judiciary Committee that produced H.R. 7152, chose a different route.
Title II therefore reaches establishments whose operations affect commerce or whose discrimination is supported by state action, and the statute defines the commerce connection establishment by establishment: an inn or hotel providing lodging to transient guests, a restaurant or lunch counter serving food a substantial portion of which has moved in commerce, a movie house or other place of exhibition presenting films or performances that move in commerce, and any establishment physically located within a covered establishment. The mechanism is textual rather than aspirational. Each category carries its own jurisdictional hook, which is why the earliest Title II cases are arguments about supply chains and out-of-state guests rather than arguments about equality.
Title VII took the same commerce route, defining a covered employer as a person engaged in an industry affecting commerce with the requisite number of employees. Title VI took an entirely different one. It conditions the receipt of federal financial assistance on non-discrimination, which is an exercise of the spending power, and its enforcement mechanism is fund termination rather than a suit for damages. That structural difference produced a separate line of cases about who may sue to enforce a spending condition, a question that has no analogue in Title VII litigation and that the Court eventually answered in a way that surprised a generation of civil rights practitioners.
Why did Congress use the commerce power instead of the Fourteenth Amendment?
Because the Civil Rights Cases of 1883 held that the Fourteenth Amendment reaches state action rather than private conduct. Resting Title II on the commerce power let Congress regulate private hotels and restaurants without asking the Court to overrule that precedent, and the Court upheld the choice in December 1964.
The tradeoff was accepted deliberately. Grounding a civil rights statute in commerce meant that the government’s opening argument in the first test case would be about the interstate movement of guests and groceries rather than about the equal protection of the laws, and the Justice Department knew how that would read. The alternative was a statute whose constitutionality depended on the Court reversing itself, at a moment when the administration needed the law to be operating, not litigating. The full account of how that choice was made in committee and defended on the floor belongs to the passage history of the act, which owns the legislative narrative; what matters here is that the constitutional theory in the enrolled text is what the Court was asked to review, and it held.
The first challenges: two decisions, one day, in under six months
The act was signed on July 2, 1964. The Supreme Court decided the two cases testing Title II on December 14, 1964. That interval is short by any measure and is itself part of the story, because the government and the establishments challenging the law both wanted speed, for opposite reasons.
Heart of Atlanta Motel v. United States, 379 U.S. 241 (1964), involved a large motel near two interstate highways in downtown Atlanta that solicited business from outside Georgia through national advertising and drew most of its guests from other states. The motel had refused to rent rooms to Black travelers and announced its intention to continue. It sued for a declaratory judgment that Title II exceeded Congress’s power. The holding, stated precisely: Congress may prohibit racial discrimination by an establishment serving interstate travelers under the Commerce Clause, because the burden such discrimination places on interstate travel is a permissible object of federal regulation. The practical consequence was that every hotel and motel of any size in the country was covered from that day forward, and the litigation risk to the statute’s most visible title was gone.
Katzenbach v. McClung, 379 U.S. 294 (1964), decided the same day, was the harder case and was chosen precisely because it was harder. Ollie’s Barbecue was a family restaurant in Birmingham, Alabama, eleven blocks from an interstate highway, serving a local clientele and catering almost entirely to neighborhood customers. It had no interstate travelers to speak of. The jurisdictional connection was that a substantial portion of the food it served had moved in interstate commerce before reaching its kitchen. The holding: Congress may reach a restaurant serving food that has moved in commerce, because Congress had a rational basis for finding that discrimination in such establishments, in the aggregate, burdens interstate commerce in food and in travel. The practical consequence was that the coverage question for restaurants turned on the supply chain rather than on the customer base, which made Title II’s restaurant coverage close to universal in practice.
Read together, the two decisions did something more than uphold a title. They confirmed that the aggregation principle from the Court’s New Deal commerce cases applied to a civil rights statute, and they did so unanimously in result. Justices Douglas and Goldberg wrote separately in both cases to say they would have preferred to rest the decision on the Fourteenth Amendment, a preference that has become a standard footnote in constitutional law casebooks but that did not command a Court. The majority opinions in both were written by Justice Clark.
Has any operative provision of the act been held unconstitutional?
No. The Court upheld Title II against constitutional challenge in 1964 and has never invalidated an operative provision of the act. What it has done is construe particular provisions narrowly, and Congress reversed several of those constructions by statute in 1972, 1978, 1991 and 2009.
The distinction matters more than it might appear, because the claim that the Court “struck down” part of the act circulates constantly and is often confused with what happened to a different civil rights statute. The Voting Rights Act of 1965 did lose the operative effect of one of its central mechanisms when the Court held its coverage formula unconstitutional, an outcome traced in detail in the Shelby County decision and what it left standing. Nothing comparable happened to Public Law 88-352. The 1964 act’s constitutional foundation was tested immediately, held, and has not been seriously threatened since. Its later contractions came from statutory interpretation, and statutory interpretation is reversible by the branch that wrote the statute.
Two smaller Title II decisions completed the early picture. Hamm v. City of Rock Hill, 379 U.S. 306 (1964), held that the act abated pending state prosecutions of sit-in demonstrators for conduct the statute had since made lawful, which cleared several hundred criminal cases arising from the protest wave that preceded enactment. Newman v. Piggie Park Enterprises, 390 U.S. 400 (1968), held that a prevailing plaintiff under Title II should ordinarily recover attorney’s fees unless special circumstances make an award unjust, and rejected the proprietor’s argument that serving Black customers violated his religious beliefs. The fee holding mattered structurally: Title II authorizes injunctive relief and not damages, so without a fee award a successful plaintiff wins an order and a legal bill. Daniel v. Paul, 395 U.S. 298 (1969), extended coverage to a private recreational facility that charged a membership fee, holding that the club label did not remove an otherwise covered place of entertainment from the statute’s reach.
Round one: the decisions that built Title VII, 1971 to 1987
Title VII took effect on July 2, 1965, one year after enactment, and its enforcement agency spent its first years without the power to sue anyone. The machinery of the charge process, the deferral to state agencies and the eventual grant of litigation authority are the subject of how Title VII and the EEOC operate in a real charge. The consequence for litigation history is that the doctrines were built in private suits, mostly brought by the NAACP Legal Defense and Educational Fund, and the fact patterns the Court saw were the ones private counsel selected.
The first of those, and the most consequential statutory construction in the act’s history, was Griggs v. Duke Power Co., 401 U.S. 424 (1971). Duke Power’s Dan River Steam Station had openly segregated its departments before the act took effect. After July 1965 it stopped assigning by race and instead required a high school diploma and passing scores on two general aptitude tests for transfer into the better-paid departments. Neither requirement had been shown to predict job performance, and both disqualified Black applicants at substantially higher rates than white applicants. The holding, stated precisely: Title VII prohibits employment practices that are neutral on their face and in intent but that operate to exclude protected groups, unless the practice is shown to be related to job performance, and the burden of showing that business necessity rests on the employer. The opinion was unanimous among the eight Justices participating, written by Chief Justice Burger.
The practical consequence was that a plaintiff no longer needed to prove that an employer wanted to discriminate. Proof that a selection device screened out a protected group at a disproportionate rate shifted the burden to the employer to justify the device. Personnel departments across the country revised or abandoned general intelligence testing, credential requirements that had never been validated, and physical standards that had been adopted without reference to the work. Griggs also gave the statute a theory of what discrimination is: a matter of consequences in operation rather than states of mind, at least for one class of claims. That theory has been contested ever since, inside and outside the Court, and the contest is the reason this line of cases moved twice.
Griggs did not stand alone for long. Albemarle Paper Co. v. Moody, 422 U.S. 405 (1975), held that back pay should ordinarily be awarded to a successful Title VII plaintiff, since the remedy serves the statute’s purposes of deterring violations and making victims whole, and it applied the Griggs validation requirement to a testing program the employer had attempted to justify after the fact. Washington v. Davis, 426 U.S. 229 (1976), is often miscited here: it held that the constitutional equal protection standard requires proof of discriminatory purpose and that the Griggs effects standard is a feature of Title VII rather than of the Fourteenth Amendment. The distinction has confused readers for decades. A statutory effects test and a constitutional intent test can coexist because Congress may prohibit more than the Constitution prohibits.
Connecticut v. Teal, 457 U.S. 440 (1982), closed a loophole that employers had begun to use. An employer argued that its overall promotion rate for Black candidates was favorable, so the disparate exclusion caused by one written test in a multi-step process should not create liability. The holding: Title VII protects individuals rather than groups, so a bottom-line result that appears balanced does not immunize a particular selection device that excludes on a prohibited basis. The practical consequence was that each step in a hiring or promotion sequence became independently reviewable, which is why compliance work in this area is conducted step by step rather than in the aggregate. Watson v. Fort Worth Bank and Trust, 487 U.S. 977 (1988), then held that subjective selection practices, including undocumented supervisory judgment, are subject to disparate impact analysis and not merely to intentional discrimination analysis, though the Court fractured on how the burdens should run and left that question for the following term.
The individual proof framework arrived two years after Griggs. McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973), addressed a case in which a mechanic laid off in a workforce reduction had participated in a protest against the employer and was later refused rehire. The holding: in an individual disparate treatment case without direct evidence, the plaintiff first establishes a prima facie case by showing membership in a protected class, qualification for an available position, rejection, and that the position remained open or was filled; the employer then articulates a legitimate non-discriminatory reason; and the plaintiff may then show that the stated reason is a pretext for discrimination.
The framework has structured tens of thousands of employment cases since, and two later decisions refined it. Texas Department of Community Affairs v. Burdine, 450 U.S. 248 (1981), held that the employer’s second-stage obligation is a burden of production and not a burden of persuasion, and that the burden of persuading the factfinder of intentional discrimination remains with the plaintiff throughout. St. Mary’s Honor Center v. Hicks, 509 U.S. 502 (1993), held that a factfinder’s rejection of the employer’s stated reason permits, but does not compel, a finding of discrimination, so disproving the explanation does not automatically win the case. Practitioners describe the combined effect as a framework that organizes proof at summary judgment without dictating outcomes at trial.
How are most Title VII cases actually decided?
It is the three-step burden-shifting method the Court adopted in 1973 for individual discrimination claims proved by circumstantial evidence. The plaintiff establishes a prima facie case, the employer articulates a legitimate reason, and the plaintiff shows pretext. Most Title VII cases are decided within this structure at summary judgment.
Two further strands of the expansion round deserve their own treatment because they answered questions the 1964 text left genuinely open. The first concerns sex discrimination, a category added to Title VII on the House floor and treated by many observers in the first years as a drafting accident. Phillips v. Martin Marietta Corp., 400 U.S. 542 (1971), decided that an employer who refused to hire women with preschool-age children while hiring men with preschool-age children had drawn a line on the basis of sex, rejecting the argument that a sex-plus classification falls outside the statute because it does not exclude all women. Dothard v. Rawlinson, 433 U.S. 321 (1977), applied Griggs to height and weight requirements for Alabama prison guards, found them unjustified, and then held that the bona fide occupational qualification defense permitted excluding women from contact positions in that particular maximum-security environment, giving the BFOQ its narrowest realistic illustration. Automobile Workers v. Johnson Controls, 499 U.S. 187 (1991), held that a battery manufacturer’s policy barring fertile women from jobs involving lead exposure was facially discriminatory and not saved by the BFOQ, because the defense reaches qualifications that concern an employee’s ability to perform the job rather than the employer’s assessment of risk to potential offspring.
General Electric Co. v. Gilbert, 429 U.S. 125 (1976), ran in the other direction, holding that an employee disability plan excluding pregnancy-related conditions was not sex discrimination under Title VII because the classification distinguished pregnant persons from non-pregnant persons rather than women from men. That construction lasted two years. The Pregnancy Discrimination Act of 1978 amended the statutory definition of discrimination because of sex to include pregnancy, childbirth and related medical conditions, and the amendment is the first clean instance of the legislative override cycle traced in the amendments that rebuilt this statute after 1964.
The second strand is harassment, a theory the 1964 text does not mention. Meritor Savings Bank v. Vinson, 477 U.S. 57 (1986), held that a hostile or abusive working environment based on sex violates Title VII even when the employee suffers no economic loss, and that the relevant inquiry is whether the conduct was unwelcome rather than whether participation was voluntary. Harris v. Forklift Systems, 510 U.S. 17 (1993), held that conduct must be severe or pervasive enough that a reasonable person would find the environment hostile and that the plaintiff so perceived it, but that no showing of psychological injury is required. Oncale v. Sundowner Offshore Services, 523 U.S. 75 (1998), held unanimously that harassment by members of the same sex is actionable, since the statutory question is whether the conduct constituted discrimination because of sex and not whether the parties belong to different sexes.
A final expansion-era question concerned voluntary race-conscious and sex-conscious personnel plans. United Steelworkers v. Weber, 443 U.S. 193 (1979), held that Title VII does not prohibit a private employer and union from adopting a voluntary, temporary training plan reserving a share of places for Black workers to correct a manifest imbalance in a traditionally segregated job category. Johnson v. Transportation Agency, Santa Clara County, 480 U.S. 616 (1987), extended the reasoning to a public employer’s consideration of sex as one factor in a promotion decision under a similar plan. Local 28 of the Sheet Metal Workers v. EEOC, 478 U.S. 421 (1986), held that a court may order race-conscious relief benefiting individuals who were not themselves identified victims where the violation has been egregious and persistent. Whether these holdings are consistent with the statute’s declaration in section 703(j) that nothing in the title requires preferential treatment to correct an imbalance is a genuine and continuing argument, and the operative text of that section is walked through in the title-by-title reading of the act, which owns the provisions questions.
Round two: the 1989 term and the contraction of Title VII
Between January and June of 1989 the Supreme Court decided five employment discrimination cases that narrowed doctrines the lower courts had been applying for years. No single one of them was a constitutional ruling, none touched the act’s validity, and all five were statutory constructions that Congress could revisit. Taken together they changed enough of the litigation landscape that a two-year legislative fight followed, ending in the most substantial rewrite Title VII has received.
Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989), is the central case. Salmon canneries in Alaska staffed unskilled cannery jobs largely with Alaska Native and Filipino workers and staffed skilled non-cannery jobs largely with white workers hired through separate channels, with segregated housing and dining following the job classifications. The plaintiffs relied on the workforce comparison to establish disparate impact. The holding had three parts. A statistical comparison must be between the racial composition of the qualified labor market and the composition of the at-issue jobs, rather than between two internal workforces, since a high proportion of minority workers in one job category does not itself show that another category is closed. A plaintiff must identify the specific employment practice alleged to cause the disparity rather than challenging the selection process as an undifferentiated whole. And the employer’s justification burden is a burden of production rather than persuasion, with the practice needing to serve legitimate employment goals rather than being shown essential to the business.
The practical consequence was that disparate impact litigation became substantially harder to bring and easier to defend. The specific-practice requirement in particular is difficult to satisfy when the challenged decision emerges from a diffuse process with no single documented gate. Whether Wards Cove restated Griggs or departed from it was disputed at the time, including inside the Court, and the dissenters said plainly that the Court had rewritten a settled framework. The important point for a reader tracing operative law is narrower and not disputed at all: the 1991 act displaced this allocation of burdens, and citing Wards Cove for the burden rules describes law that stopped operating in November 1991.
Price Waterhouse v. Hopkins, 490 U.S. 228 (1989), came out of a partnership decision at an accounting firm, where a candidate with a strong record was denied partnership after evaluations that criticized her interpersonal style in terms that referred to her being a woman, including advice about walking, talking and dressing more femininely. The Court divided, producing a plurality opinion by Justice Brennan and separate concurrences. The holding as it functioned in the lower courts: when a plaintiff shows that a protected characteristic played a motivating part in an employment decision, the employer may avoid liability by proving by a preponderance of the evidence that it would have made the same decision absent that consideration. The decision also established that sex stereotyping evidence is probative of discrimination because of sex, a proposition that carried substantial weight in later litigation.
The practical consequence cut two ways, which is why the case is claimed by both sides in argument. It recognized mixed-motive liability and made stereotyping evidence usable, and it simultaneously gave employers a complete defense on a preponderance showing. Congress adjusted the second half in 1991 without disturbing the first.
Three further 1989 decisions completed the set. Patterson v. McLean Credit Union, 491 U.S. 164 (1989), construed the Reconstruction-era statute now codified at 42 U.S.C. section 1981, holding that its guarantee of the same right to make and enforce contracts covers the formation of the contract and access to legal process but not racial harassment occurring after the employment relationship has begun. That is not a Title VII holding, and readers routinely file it as one; its significance here is that section 1981 carried uncapped damages and a longer limitations period, so narrowing it removed the main alternative to Title VII for race claims. Lorance v. AT&T Technologies, 490 U.S. 900 (1989), held that the charge-filing period for a challenge to a facially neutral seniority system adopted with discriminatory intent runs from the adoption of the system rather than from its later application to the employee, which for the plaintiffs in that case meant their claims had expired before the harm reached them. Martin v. Wilks, 490 U.S. 755 (1989), held that white firefighters in Birmingham who were not parties to earlier consent decrees could bring later challenges to promotions made under them, since a person is generally not bound by a judgment in litigation to which they were not a party.
Two years later, EEOC v. Arabian American Oil Co., 499 U.S. 244 (1991), held that Title VII did not apply to United States citizens employed by American employers outside the United States, applying the presumption against extraterritorial application to a statute that lacked a clear statement. The decision issued in March 1991, while the legislative response to the 1989 term was already moving, and it was folded into that response.
Which holdings from the 1989 term did the 1991 statute displace?
The Civil Rights Act of 1991 displaced the burden allocation in Wards Cove, adjusted the mixed-motive defense from Price Waterhouse, restored post-formation coverage under section 1981 after Patterson, changed the seniority-system clock after Lorance, limited collateral attacks after Martin v. Wilks, and extended Title VII abroad after Aramco.
Round three: Congress writes the constructions back out
The legislative sequence is the subject of its own article and is only summarized here, because the canonical treatment of the amendments belongs to the account of the 1991 act and everything that followed it. What a reader tracing case law needs is the fact of the override and its precise reach.
A bill responding to the 1989 decisions passed both chambers in 1990 and was vetoed, with the veto message resting on the argument that its burden-of-proof provisions would push employers toward hiring by the numbers to avoid litigation exposure. The Senate override attempt fell one vote short. A renegotiated version passed the following year and became the Civil Rights Act of 1991, Public Law 102-166, signed on November 21, 1991.
For the litigation history, four features of that statute matter. It codified disparate impact analysis in the text of Title VII, placing the burden of proving job relatedness and business necessity on the employer once a plaintiff demonstrates that a particular practice causes a disparate impact, and providing that where the elements of a decision process cannot be separated for analysis the process may be challenged as one practice. It provided that an unlawful practice is established when a protected characteristic was a motivating factor, while limiting a plaintiff who prevails on that showing to declaratory and injunctive relief plus fees, and not damages or reinstatement, if the employer proves it would have taken the same action anyway. It created compensatory and punitive damages for intentional discrimination, along with the right to a jury trial, with combined caps scaled by employer size that run from fifty thousand dollars for the smallest covered employers to three hundred thousand dollars for employers with more than five hundred employees. And it amended section 1981 to reach the performance and termination of contracts, undoing Patterson.
The damages provision is the least understood part of the override, so it is worth stating exactly. Before November 1991, a prevailing Title VII plaintiff could obtain injunctive relief, reinstatement, back pay and attorney’s fees, and no damages for emotional harm and no punitive damages, which is why race plaintiffs with a viable section 1981 theory had preferred that statute. After 1991 the statute carries damages and juries, subject to caps that have not been adjusted by statute since enactment. The uncapped element remains back pay, which is a make-whole remedy rather than damages.
The override cycle did not end there. Ledbetter v. Goodyear Tire and Rubber Co., 550 U.S. 618 (2007), held by a vote of five to four that the charge-filing period for a pay discrimination claim runs from the discriminatory pay decision itself, so a plaintiff who learned years later that her pay had been set below that of comparable men had filed too late, with the majority reasoning that each subsequent paycheck did not restart the clock absent a fresh discriminatory act. The dissent argued that pay disparities are typically concealed and accumulate in small increments, and it called on Congress to act. Congress did, in the Lilly Ledbetter Fair Pay Act of 2009, which provides that an unlawful practice occurs when a discriminatory compensation decision is adopted, when an individual becomes subject to it, and each time compensation is paid pursuant to it. The mechanics of that statute and its effect on the charge clock are covered in the pay-clock statute of 2009.
That sequence, running from a judicial construction to a public argument to an amended text, has now repeated four times: 1972 for enforcement authority and coverage, 1978 for pregnancy, 1991 for the 1989 term, and 2009 for the pay clock. It is the clearest working example in American law of a statute maintained by legislative override, and it is why any accurate answer to a question about what Title VII requires has to be dated.
The employer liability settlement of 1998 and what followed it
Meritor recognized hostile environment liability in 1986 but left open when an employer is answerable for a supervisor’s conduct. Lower courts split for a decade over whether agency principles, notice, or something else supplied the rule. The Court resolved it in two decisions issued on the same day in June 1998.
Burlington Industries v. Ellerth, 524 U.S. 742 (1998), and Faragher v. City of Boca Raton, 524 U.S. 775 (1998), adopted a single framework. An employer is vicariously liable for actionable harassment by a supervisor with immediate or successively higher authority over the employee. Where the harassment culminated in a tangible employment action, meaning a significant change in employment status such as discharge, demotion, or an undesirable reassignment, no defense is available. Where no tangible employment action occurred, the employer may raise an affirmative defense with two elements, both of which it must prove: that it exercised reasonable care to prevent and promptly correct harassing behavior, and that the employee unreasonably failed to take advantage of preventive or corrective opportunities the employer provided.
The practical consequence is visible in every workplace in the country and is rarely traced to its source. The written anti-harassment policy, the multiple reporting channels that let an employee bypass a direct supervisor, the mandatory training, the documented investigation file, and the prompt remedial action memorandum are all artifacts of the first element of that defense. The second element is why a policy that no one knows about is worth nothing to an employer, and why complaint procedures are publicized rather than filed. What began as a judicially crafted allocation of liability became a compliance architecture. Vance v. Ball State University, 570 U.S. 421 (2013), later narrowed who counts as a supervisor for this purpose, holding that the term covers a person empowered by the employer to take tangible employment actions against the victim rather than anyone who directs the victim’s daily work, with the consequence that harassment by a lead worker without hiring or firing authority is analyzed under the negligence standard applicable to coworker harassment.
Retaliation developed on a parallel track. Section 704(a) prohibits discrimination against a person who has opposed an unlawful practice or participated in an investigation or proceeding, and the question was how serious an employer’s response must be before it violates that section. Burlington Northern and Santa Fe Railway Co. v. White, 548 U.S. 53 (2006), held that the retaliation provision is not limited to actions affecting the terms and conditions of employment, and that the standard is whether the employer’s action might well have dissuaded a reasonable worker from making or supporting a charge. The practical consequence was that a lateral reassignment to a less desirable job, or a suspension later reversed with back pay, can support a retaliation claim even though it would not support a discrimination claim. University of Texas Southwestern Medical Center v. Nassar, 570 U.S. 338 (2013), then held that retaliation claims require proof that the protected activity was a but-for cause of the adverse action, rather than the motivating-factor standard the 1991 act created for status-based discrimination, because the 1991 amendment names the protected characteristics and does not name retaliation.
Ricci v. DeStefano, 557 U.S. 557 (2009), addressed the tension the two theories of liability create for an employer caught between them. New Haven had administered promotional examinations for firefighters, obtained results with a pronounced racial disparity, and discarded the results rather than certify them, citing exposure to disparate impact liability. Firefighters who had scored well sued, claiming that discarding the results because of their racial consequences was intentional discrimination. The holding: before an employer may take a race-conscious action to avoid disparate impact liability, it must have a strong basis in evidence to believe it would be subject to such liability if it failed to act, and the city did not meet that standard on this record. The practical consequence is a standard governing an employer’s discretion to abandon a completed selection process, and a reminder that the two Title VII theories can point in opposite directions on the same facts.
Two procedural decisions round out the modern picture. Fort Bend County v. Davis, decided in 2019, held that Title VII’s charge-filing requirement is a claim-processing rule rather than a jurisdictional one, so an employer that does not raise the defect in time forfeits it. And Green v. Brennan, decided in 2016, held that the filing period for a constructive discharge claim begins when the employee gives notice of resignation rather than at the employer’s last discriminatory act. Neither changed a substantive standard, and both changed the calendar in a meaningful number of cases.
Title VI’s separate path and the private right of action
Title VI has produced a smaller and stranger body of case law than Title VII, because its enforcement design is different. It prohibits discrimination on the ground of race, color or national origin under any program or activity receiving federal financial assistance, and it directs federal agencies to effectuate that prohibition by issuing rules, with termination of assistance as the ultimate sanction after a hearing. Nothing in the text says a private person may sue.
Lau v. Nichols, 414 U.S. 563 (1974), held that a school district receiving federal funds violated Title VI regulations when it provided instruction only in English to students who did not speak it, without remedial measures, because identical treatment does not constitute equal treatment where students cannot understand the instruction. The decision rested on the implementing regulations rather than on the statute alone, a distinction that most summaries drop and that turned out to matter enormously three decades later.
Regents of the University of California v. Bakke, 438 U.S. 265 (1978), produced no majority opinion, but a majority of Justices accepted the proposition that Title VI prohibits only the racial classifications that the Equal Protection Clause prohibits when the recipient is a public institution. The practical consequence is that Title VI litigation against public recipients tends to collapse into constitutional analysis, which is why Title VI arguments appear inside affirmative action cases rather than as an independent statutory theory.
Guardians Association v. Civil Service Commission of the City of New York, 463 U.S. 582 (1983), fractured badly, but the working proposition the lower courts drew from it was that the statute itself reaches intentional discrimination, while agency regulations validly reach practices with discriminatory effects, and that remedies differ depending on which is proved. That left an obvious question. If the regulations reach effects, may a private plaintiff sue to enforce the regulations?
Alexander v. Sandoval, 532 U.S. 275 (2001), answered no. A driver’s license applicant challenged Alabama’s policy of administering the examination only in English as a violation of Department of Justice regulations implementing Title VI. The holding, by a vote of five to four: private individuals may not sue to enforce disparate impact regulations promulgated under section 602 of Title VI, because the private right of action recognized under section 601 reaches only intentional discrimination and a regulation cannot create a right of action Congress did not create.
The practical consequence was immediate and large. Environmental justice claims, language access claims, school discipline claims and transportation siting claims that had been brought under Title VI regulations lost their private vehicle in a single decision. Enforcement of the effects regulations reverted to the funding agencies, which retain authority to investigate and to pursue administrative remedies. Plaintiffs’ counsel adapted by pleading intentional discrimination, by moving to state law, or by pressing administrative complaints. The decision is a clean illustration of the difference between a statutory duty and a means of enforcing it, and it is the strongest single argument for the proposition, developed at length in this series, that the enforcement architecture of a statute determines its operative reach more reliably than its substantive commands do.
One more piece of the spending-power story runs alongside. Grove City College v. Bell, 465 U.S. 555 (1984), construed the phrase program or activity in Title IX of the Education Amendments of 1972, a statute modeled on Title VI, to reach only the specific program receiving federal money rather than the whole institution. Because the same phrase appears in Title VI and in two other civil rights statutes, the reading threatened the coverage of all four. Congress responded with the Civil Rights Restoration Act of 1987, Public Law 100-259, enacted on March 22, 1988 over a presidential veto, which defines program or activity institution-wide across all four statutes. A reader assembling the override list should include it, and should note that it is the only entry that was produced by a Title IX case.
Bostock and the text-first turn
Bostock v. Clayton County, decided on June 15, 2020, consolidated three cases: a county employee fired after joining a gay recreational softball league, a skydiving instructor fired after telling a customer he was gay, and a funeral home employee fired after informing the employer she would live and work as a woman. Each employer conceded the reason for the termination and argued that Title VII does not cover it.
The holding, stated precisely: an employer who fires an individual merely for being homosexual or transgender discriminates against that individual because of sex, because it is impossible to discriminate on those grounds without taking the individual’s sex into account. The vote was six to three, with the majority opinion by Justice Gorsuch and dissents by Justice Alito, joined by Justice Thomas, and by Justice Kavanaugh.
The reasoning is worth stating carefully because it is routinely described incorrectly. The majority did not hold that Congress in 1964 intended to protect gay and transgender employees, and it said the opposite: the legislators who adopted the sex provision likely did not anticipate this application. It held that the statutory text prohibits discrimination against an individual because of that individual’s sex, that but-for causation is the standard the text sets, and that an employer who tolerates a female employee’s attraction to men but fires a male employee for the same attraction has treated the two differently in a way that depends on sex. The dissents argued that the ordinary public meaning of discrimination because of sex in 1964 did not encompass sexual orientation or gender identity, and that the majority had updated a statute under the label of textualism.
The practical consequence was to settle a question on which the federal courts of appeals had divided, and to do so as a matter of statutory construction rather than constitutional law. That distinction has two implications a careful reader should hold. Bostock did not amend Title VII; the text is what it was, and the decision states what that text has meant since 1964. And because the holding is statutory, it is subject to the same override mechanism as every other construction in this article, which is a structural observation rather than a prediction. The opinion also expressly declined to address bathrooms, locker rooms, dress codes, and the operation of the Religious Freedom Restoration Act of 1993 in Title VII cases, so claims that it resolved those questions describe something the decision did not do.
The construction ledger: every major holding and what happened to it
A reader who needs to check a claim quickly wants one place where the holdings sit next to their consequences and next to the fact of any legislative reversal. This is the construction ledger. The final column is the one that resolves most arguments, because a decision that Congress has overridden describes the law of its own period and not the operative law afterward, and a great deal of confident writing about this statute cites overridden holdings without noticing.
| Year | Title or statute construed | Holding in one sentence | Practical consequence | Overridden by Congress |
|---|---|---|---|---|
| 1964 | Title II | Congress may bar racial discrimination by an establishment serving interstate travelers under the commerce power (Heart of Atlanta Motel) | Hotels and motels covered nationwide; constitutional challenge closed | No |
| 1964 | Title II | Congress may reach a local restaurant serving food that moved in commerce (Katzenbach v. McClung) | Restaurant coverage turns on supply chain, not clientele | No |
| 1968 | Title II | A prevailing plaintiff ordinarily recovers attorney’s fees absent special circumstances (Newman v. Piggie Park) | Injunction-only title becomes economically viable to enforce | No |
| 1971 | Title VII | Facially neutral practices that exclude disproportionately violate the title absent job relatedness (Griggs v. Duke Power) | Unvalidated tests and credentials abandoned; effects theory established | No, later codified in 1991 |
| 1973 | Title VII | Circumstantial individual claims proceed through prima facie case, articulated reason, and pretext (McDonnell Douglas v. Green) | The organizing framework for most employment cases | No |
| 1975 | Title VII | Back pay is ordinarily awarded to a prevailing plaintiff (Albemarle Paper v. Moody) | Remedial expectation set; validation requirements tightened | No |
| 1976 | Title VII | Excluding pregnancy from a disability plan is not sex discrimination (General Electric v. Gilbert) | Pregnancy exclusions permitted for two years | Yes, Pregnancy Discrimination Act of 1978 |
| 1979 | Title VII | Voluntary plans correcting a manifest imbalance in a segregated job category are not barred (Steelworkers v. Weber) | Voluntary affirmative action plans given statutory room | No |
| 1982 | Title VII | A favorable bottom line does not immunize a discriminatory selection step (Connecticut v. Teal) | Each step of a selection process independently reviewable | No |
| 1984 | Title IX (shared text) | Program or activity reaches only the funded program, not the institution (Grove City College v. Bell) | Coverage of four civil rights statutes narrowed | Yes, Civil Rights Restoration Act of 1987 |
| 1986 | Title VII | A hostile environment based on sex violates the title without economic loss (Meritor Savings v. Vinson) | Harassment established as a Title VII theory | No |
| 1989 | Title VII | Plaintiff must identify a specific practice and employer bears only a production burden (Wards Cove v. Atonio) | Disparate impact claims substantially harder to sustain | Yes, Civil Rights Act of 1991 |
| 1989 | Title VII | Employer avoids liability by proving it would have decided the same way (Price Waterhouse v. Hopkins) | Mixed motive recognized; complete defense available | Partly, Civil Rights Act of 1991 |
| 1989 | Section 1981 | The contracts guarantee does not reach post-formation harassment (Patterson v. McLean) | Main uncapped alternative to Title VII narrowed | Yes, Civil Rights Act of 1991 |
| 1989 | Title VII | The clock for a seniority system runs from adoption, not application (Lorance v. AT&T) | Claims expired before injury reached employees | Yes, Civil Rights Act of 1991 |
| 1991 | Title VII | The title does not apply outside the United States (EEOC v. Arabian American Oil) | Overseas employees of American firms uncovered | Yes, Civil Rights Act of 1991 |
| 1998 | Title VII | Employers are vicariously liable for supervisor harassment, subject to a two-element defense (Faragher and Ellerth) | Modern policy, reporting and training architecture | No |
| 2001 | Title VI | No private right of action to enforce disparate impact regulations (Alexander v. Sandoval) | Effects claims moved to agency enforcement | No |
| 2006 | Title VII | Retaliation reaches actions that would dissuade a reasonable worker (Burlington Northern v. White) | Retaliation standard decoupled from terms of employment | No |
| 2007 | Title VII | The pay claim clock runs from the compensation decision (Ledbetter v. Goodyear) | Concealed pay disparities time-barred before discovery | Yes, Lilly Ledbetter Fair Pay Act of 2009 |
| 2009 | Title VII | Race-conscious rejection of test results requires a strong basis in evidence (Ricci v. DeStefano) | Limits on abandoning a completed selection process | No |
| 2013 | Title VII | Retaliation requires but-for causation (University of Texas Southwestern v. Nassar) | Retaliation and status claims proved to different standards | No |
| 2020 | Title VII | Firing someone for being gay or transgender is discrimination because of sex (Bostock v. Clayton County) | Circuit split resolved; coverage settled as a textual matter | No |
Six entries in the final column, spread across 1978, 1988, 1991 and 2009, are the reason this statute cannot be described accurately without a date attached. Two of them, Gilbert and Grove City, were overridden within four years. One, Ledbetter, took eighteen months. The 1989 cluster took two years and a failed veto override. No other major federal statute of the period has been corrected by Congress this often in response to judicial construction, and the pattern is the strongest available evidence for a claim this series makes repeatedly: a statute is not a fixed object but a machine that Congress, the agencies and the courts keep operating on.
What the Civil Rights Act Supreme Court cases add up to: four doctrinal families
The case list is long enough that a reader benefits from sorting it by what each decision was actually doing. Four families cover nearly all of it, and a claim about the act almost always belongs to one of them.
Coverage and constitutional authority
This family asks whether the statute reaches a given actor at all. Heart of Atlanta and McClung answered the constitutional version of the question for Title II and closed it. Daniel v. Paul answered a coverage version, holding that a nominal club structure does not remove a place of entertainment from the title. In Title VII the coverage questions were largely resolved by amendment rather than by decision, since the Equal Employment Opportunity Act of 1972 lowered the employee threshold and extended the title to state and local governments and to educational institutions. Aramco was a coverage decision reversed by amendment. Bostock is a coverage decision in the sense that it settled which employees the sex provision protects, though its reasoning is about causation rather than about who counts as an employer.
What unifies this family is that its answers are close to binary and tend to stay settled. An establishment is covered or it is not. When Congress has disagreed with a coverage holding, it has amended the definitions, which is the cheapest kind of legislative correction to draft.
Proof frameworks
This is the largest family and the one that generates the most litigation. Griggs created the effects theory, Wards Cove narrowed it, and the 1991 act restored and codified it with the specific-practice requirement retained in modified form. McDonnell Douglas, Burdine and Hicks built the circumstantial-evidence framework for individual claims. Price Waterhouse and the 1991 amendment built the mixed-motive rule. Nassar carved retaliation out of the motivating-factor standard and returned it to but-for causation.
These frameworks are where most cases are actually decided, usually at summary judgment, and they are also where the greatest distance opens between what the statute says and what a party must do to win. Nothing in the 1964 text describes a prima facie case or a burden-shifting sequence. Those are judicial constructions built to administer a prohibition that Congress wrote in a single sentence, and their existence is the clearest available demonstration that operative law is text plus construction rather than text alone.
Employer responsibility
Meritor established that harassment can violate the title. Faragher and Ellerth allocated responsibility between the employer and the employee and gave the employer a defense conditioned on its own preventive conduct. Vance narrowed the class of people whose conduct triggers strict vicarious liability. Harris set the severity threshold, and Oncale removed the requirement that harasser and victim be of different sexes.
This family is unusual because its practical output is not litigation doctrine but organizational design. The affirmative defense created an incentive structure, and employers responded by building the compliance apparatus the defense rewards. A reader who wants to see judicial construction converted into daily working reality should look here first, and then at the enforcement process traced in the Title VII charge process from filing to court.
Remedies and timing
Albemarle set the back pay expectation. Piggie Park set the fee expectation under Title II. The 1991 act supplied damages, juries and caps. Lorance and Ledbetter set claim clocks that Congress then reset. Green v. Brennan set the constructive discharge clock. Fort Bend held that the charge requirement is waivable if not timely raised.
Timing decisions look technical and decide an enormous number of real cases, because a claim that is filed late is over regardless of its merits. Two of the four congressional overrides in this article were responses to timing holdings, which is a fair indication of how much practical weight the family carries.
What the courts left standing and what they neutralized
An inventory is more useful here than a summary, and it divides into three groups.
Standing without qualification: the constitutional foundation of Title II under the commerce power; the availability of injunctive relief and attorney’s fees under Title II; the disparate impact theory, now codified in the text rather than resting on Griggs alone; the McDonnell Douglas framework for circumstantial individual claims; hostile environment liability; the vicarious liability rule for supervisor harassment with the two-element defense; the retaliation standard from Burlington Northern; and the coverage of sexual orientation and gender identity discrimination under the sex provision after 2020.
Standing in modified form: the mixed-motive rule, which survives as a statutory motivating-factor test with a same-decision defense that limits remedies rather than eliminating liability; the specific-practice requirement from Wards Cove, which the 1991 act retained while providing that elements incapable of separation may be analyzed as one practice; and the section 1981 contracts guarantee, which reaches post-formation conduct after the 1991 amendment but retains the intent requirement that distinguishes it from Title VII’s effects theory.
Neutralized or superseded: the Wards Cove burden allocation; the Gilbert holding on pregnancy; the Grove City reading of program or activity; the Lorance seniority clock; the Aramco extraterritoriality holding; and the Ledbetter accrual rule. Each of these describes law that operated for a defined period and then stopped, and citing any of them as a current statement of the law is the most common serious error in secondary writing about this act.
One genuinely narrowed area has not been restored. Alexander v. Sandoval eliminated the private enforcement of Title VI effects regulations, and Congress has not overturned it by statute. The substantive regulations remain valid, agencies retain administrative enforcement authority, and private plaintiffs must plead intentional discrimination. A reader who wants to understand why Title VI enforcement looks administrative while Title VII enforcement looks adversarial has the answer in that one decision plus the funding-termination design Congress wrote in 1964.
The cases people cite that are not Civil Rights Act of 1964 cases
A large share of the errors that appear in writing about this statute involve decisions that belong to other laws or to the Constitution. Sorting them out is not pedantry, because the applicable standard, the available remedy and the identity of the defendant all change depending on which body of law supplies the claim.
Brown v. Board of Education, 347 U.S. 483 (1954), is a Fourteenth Amendment decision about state-imposed school segregation, decided a decade before the act existed. It could not have been a Title IV or Title VI case, because neither existed, and it did not enforce a statute. The relationship runs the other way: the slow pace of compliance with Brown across the decade after 1954 is part of the record Congress legislated against, and Titles IV and VI supplied the enforcement tools that the judiciary alone had lacked. That relationship is developed in what the act measurably changed, which owns the outcomes analysis and separates what the statute plausibly caused from what was already moving.
Runyon v. McCrary, 427 U.S. 160 (1976), held that section 1981 prohibits private, commercially operated non-sectarian schools from excluding Black students. It rests on the Reconstruction-era statute, not on Title VI, which is why it reached schools that took no federal money. Jones v. Alfred H. Mayer Co., 392 U.S. 409 (1968), similarly rests on section 1982 and the Thirteenth Amendment rather than on the Fair Housing Act enacted that same year. Both are constantly attributed to the 1964 act.
Washington v. Davis, 426 U.S. 229 (1976), is the constitutional counterpart to Griggs, holding that disproportionate impact alone does not establish an equal protection violation. Writers who read the two decisions as inconsistent have conflated a statutory standard with a constitutional one. Congress may forbid conduct the Constitution permits, and in Title VII it did.
Roberts v. United States Jaycees, 468 U.S. 609 (1984), upheld the application of a state public accommodations law to a private membership organization against a freedom of association challenge. Title II’s covered categories are narrow and do not extend to most membership organizations, which is why discrimination claims against clubs, gyms, retail establishments and professional associations generally proceed under state public accommodations statutes rather than federal law. Readers who assume the federal act covers every business open to the public are describing the state law of many states rather than Public Law 88-352.
Shelby County v. Holder, 570 U.S. 529 (2013), concerns the Voting Rights Act of 1965 and not the Civil Rights Act of 1964. The confusion is understandable, since the statutes are adjacent in time and purpose, and it is worth resolving precisely because the two litigation histories point in opposite directions. One statute’s core survived constitutional review and was later refined by construction and amendment; the other’s coverage formula was held unconstitutional, which suspended the operation of a mechanism Congress had reauthorized four times. The comparison is drawn out fully in how the two statutes differ in structure and effect.
Finally, the Americans with Disabilities Act of 1990 and the Age Discrimination in Employment Act of 1967 borrow Title VII’s procedures and much of its vocabulary, so decisions under those statutes are often cited as Title VII authority. Sometimes the borrowing is legitimate, since the charge-filing machinery is shared. Sometimes it is not, since the causation standards differ. Naming the statute a decision construes, every time, costs one clause and prevents the most common category of error.
Reading a holding without overreading it
The single most valuable skill for a reader working through this material is separating what a decision held from what commentary says it stands for. Four habits do most of the work, and they generalize to any statute in this series.
Start with the question presented, not the outcome. Bostock held that firing an individual for being gay or transgender is discrimination because of sex under Title VII. It did not hold anything about facilities, dress codes, or religious exemptions, because those questions were not presented, and the opinion said so. A decision resolves the dispute in front of it, and the sentence beginning “this case does not address” is often the most load-bearing sentence in the opinion.
Identify the operative provision by number. Griggs construes section 703(a)(2), the provision about limiting, segregating or classifying employees in ways that deprive them of opportunities, and that textual anchor is why an effects theory was available at all. Sandoval turns on the difference between section 601 and section 602 of Title VI. Nassar turns on the fact that the 1991 motivating-factor amendment lists protected characteristics and does not list retaliation. In each case the holding is unintelligible without the section number, and summaries that omit it produce readers who cannot check the reasoning.
Ask what happened to the holding afterward. This is where the ledger earns its place. A decision may have been codified, adjusted, superseded, distinguished into irrelevance, or left untouched. Wards Cove and Ledbetter are the two most frequently miscited decisions in this area precisely because they were prominent when decided and displaced shortly afterward, so a great deal of writing describing them as law was accurate when published and is not accurate now.
Distinguish the holding from the framework built on it. Faragher and Ellerth allocate liability; they do not require any particular training program. The training program is a rational response to an affirmative defense, and the difference between a legal requirement and a rational response to a legal incentive is one that compliance literature blurs constantly. The same distinction explains why a written policy is evidence rather than an element, and why an employer with a policy nobody can find is worse off than one with no policy at all.
For a reader building a working file on these cases, the mechanics of moving from a decision to the statutory text it construes to the amended text that followed are covered in how to read a federal statute, and the sourcing questions that arise when a committee report or a floor statement is offered as evidence of meaning belong to how to find legislative history. Keeping a case chronology, the statutory sections each decision touches, and the amendment that answered it in one place is exactly the kind of research file that gets rebuilt from scratch three times before someone finally organizes it, and you can keep your statute notes, citations and case chronologies together free on VaultBook instead of reassembling it each term.
Where litigation sits in the life of this statute
This series treats a federal law as passing through four stages: enactment, rulemaking, litigation, and amendment. The Civil Rights Act of 1964 is the cleanest available illustration, because all four stages are documented and none of them can be skipped without producing a false picture of the law.
Enactment settled the text and the constitutional theory. Rulemaking and enforcement practice, especially the EEOC’s guidelines on selection procedures and the funding agencies’ Title VI regulations, determined what the text meant in the first instance for regulated parties who never litigate. Litigation then tested both, and produced the proof frameworks, the liability rules and the coverage answers this article has traced. Amendment resolved the disagreements between the second and third stages, six times.
The order matters. A great deal of writing treats litigation as the end of the story, as though a Supreme Court decision fixes a statute’s meaning permanently. In this statute’s history, a Supreme Court decision has repeatedly been the beginning of a legislative process rather than the end of an interpretive one. The three-round pattern named at the start of this article is the specific form that took here: expansion by construction from 1971 through the late 1980s, contraction by construction in 1989 and again in 2007, and restoration by statute in 1991 and 2009, with the Bostock holding sitting outside the cycle as an expansion produced by textual reasoning rather than by purposive reasoning.
The pattern also explains a persistent feature of public argument about this act. Two people can hold opposite and sincerely held views about what Title VII requires while both relying on real Supreme Court decisions, because they are describing different rounds. Establishing which round a claim comes from resolves a surprising share of these disputes before anyone has to reach the merits, and that is a more useful analytical move than deciding who has the better policy argument.
The disagreements inside the Court, stated as the Justices stated them
A neutral treatment of this case law has to report the disputes without adjudicating them, and the honest way to do that is to attribute each position to the opinion that carried it rather than to an abstract camp.
The disagreement over the effects theory is the oldest. Griggs was unanimous, but the unanimity did not survive contact with the theory’s application. In Wards Cove the majority wrote that requiring an employer to demonstrate business necessity for each challenged practice would push employers toward adopting quotas to avoid liability, and it treated the specific-practice requirement as necessary to keep the theory from becoming a challenge to workforce composition as such. Justice Stevens, in dissent, wrote that the Court had turned settled law around without acknowledging that it was doing so, and Justice Blackmun’s separate dissent questioned whether the majority still regarded racial discrimination as a serious problem. Congress in 1991 sided with the dissents on the burden question and with the majority on the specific-practice question, which is the shape of a compromise rather than a repudiation, and the enacted text reflects both.
The disagreement over mixed motives divided the Price Waterhouse Court four ways. The plurality accepted that the employer could escape liability by proving it would have made the same decision, Justice White and Justice O’Connor concurred on narrower grounds, and Justice Kennedy’s dissent argued that the plaintiff should bear the burden of proving but-for causation throughout. The 1991 act adopted neither position cleanly: it made a motivating factor sufficient for liability while allowing the same-decision showing to cut off damages and reinstatement.
The disagreement over the private enforcement of regulations produced the sharpest exchange in the Title VI line. The Sandoval majority reasoned that the right of action recognized under section 601 reaches only what section 601 forbids, and that agencies cannot create private rights by regulation. Justice Stevens, in dissent, argued that the Court had disregarded decades of practice in which private suits under the regulations had been assumed by Congress and the courts alike. Congress has not resolved the disagreement by statute, which distinguishes this dispute from the Title VII ones and leaves the administrative enforcement route as the operative answer.
The disagreement over Ledbetter was procedural on the surface and substantive underneath. The majority reasoned that the statute requires a charge within a fixed period of the discriminatory act, and that treating each paycheck as a new violation would nullify the filing period Congress set. The dissent argued that pay discrimination is typically hidden and accrues in increments too small to detect, so a rule keyed to the original decision immunizes exactly the conduct the statute targets, and it invited a legislative response in terms unusual for their directness. The response came less than two years later.
The disagreement in Bostock was about interpretive method as much as about outcome. The majority reasoned from the text’s focus on the individual and from but-for causation. Justice Alito’s dissent argued that the Court had substituted a logical exercise for the ordinary public meaning of the words in 1964 and had legislated under the label of textualism. Justice Kavanaugh’s dissent accepted that the outcome might be desirable as policy while arguing that the ordinary meaning of the phrase controls and that the amendment power belongs to Congress. Reporting the case without reporting that disagreement produces a distorted picture of what happened, since the dissents are the reason the decision is discussed as a methodological event rather than only as a coverage ruling.
Title II after 1969 and why its litigation record went quiet
The public accommodations title generated intense litigation for five years and then produced comparatively little. That pattern is itself a finding, and it has three explanations that can be stated from the record.
The first is that the constitutional question was closed early and completely, so the category of case that would have dominated the docket disappeared. The second is that the title’s coverage categories are specific and short, which left fewer boundary disputes than a general prohibition would have produced, and pushed novel claims toward state public accommodations laws with broader coverage. The third is remedial. Title II authorizes injunctive relief and, after Piggie Park, attorney’s fees, but it does not authorize damages, so a plaintiff who has already been refused service has no financial remedy to pursue and often no continuing interest in an injunction. A statute that offers no damages generates fewer suits, independent of how much conduct it prohibits.
Reading that quiet as evidence that discrimination in public accommodations ended would be a mistake, and reading it as evidence that the title failed would be a different mistake. Compliance in the covered categories was rapid and broad, which is the strongest single piece of evidence for the title’s effectiveness, and is treated with the underlying evidence in the outcomes article. The remedial design is why the enforcement record is thin even where the behavioral record is strong. The two facts are compatible and are frequently confused for each other.
What the decisions left unresolved
Several questions in this body of law remain genuinely open, and identifying them precisely is more useful than pretending the doctrine is complete.
The relationship between the disparate impact provision and constitutional equal protection has never been squarely resolved. Ricci raised the question of what an employer must do when avoiding one form of liability creates exposure to the other, and answered it with an evidentiary standard rather than a constitutional holding. Justice Scalia’s concurrence stated that the underlying tension would have to be addressed eventually. It has not been.
The scope of religious exemptions in Title VII cases is unsettled at the intersection with other federal statutes. The act’s own exemption for religious organizations addresses employment of individuals of a particular religion to carry on the organization’s activities, and the Religious Freedom Restoration Act of 1993 supplies a separate framework whose application to Title VII claims the Bostock majority expressly declined to address. The accommodation standard for individual religious practice, set out in Trans World Airlines v. Hardison, 432 U.S. 63 (1977), as requiring no more than a minimal cost, has been criticized in later separate opinions as inconsistent with the statutory phrase undue hardship.
The treatment of the specific-practice requirement where a decision process is genuinely undocumented remains contested in the lower courts, since the 1991 text permits challenging a process as a single practice when its elements cannot be separated for analysis, and what counts as incapable of separation varies by circuit.
A fifth open question concerns the interaction between Title VII and arbitration agreements. The statute says nothing about whether a claim may be routed out of federal court by a pre-dispute agreement, and the answer has been supplied by decisions construing the Federal Arbitration Act rather than by decisions construing the Civil Rights Act. The practical effect on where these claims are heard is substantial, and the doctrinal source sits entirely outside the statute this article traces, which is a useful reminder that a statute’s operative reach can be reshaped by law it never mentions.
And the boundary between the negligence standard for coworker harassment and vicarious liability for supervisor harassment, after Vance narrowed the supervisor category, produces recurring disputes in workplaces where operational authority and formal authority sit in different people. None of these questions has an answer that can be stated from the record, and an article that supplied one would be inventing law rather than describing it.
The chronology in brief
Stated as a sequence rather than as a table, the litigation history has a clear shape. The first period, from 1964 to 1969, is constitutional and is over quickly: two decisions in December 1964 secure Title II, a 1964 decision abates pending sit-in prosecutions, a 1968 decision makes fee awards routine, and a 1969 decision closes the private-club evasion. Nothing about the act’s validity is seriously litigated afterward.
The second period, from 1971 to 1987, builds Title VII. Griggs supplies the effects theory in 1971, the same year that Phillips v. Martin Marietta rejects the sex-plus argument. McDonnell Douglas supplies the circumstantial framework in 1973, Albemarle supplies the back pay expectation in 1975, Franks supplies retroactive seniority in 1976, Dothard applies the effects theory to physical standards and gives the BFOQ its narrow illustration in 1977, Weber opens room for voluntary plans in 1979, Teal closes the bottom-line defense in 1982, and Meritor establishes harassment liability in 1986. Congress intervenes once during this period, in 1978, to reverse the Gilbert holding on pregnancy, and once more in 1988 to reverse the Grove City reading of program or activity across four statutes.
The third period is short and concentrated. Between January and June of 1989 the Court decides Wards Cove, Price Waterhouse, Patterson, Lorance and Martin v. Wilks, and in March 1991 it adds Aramco. A bill responding to the 1989 decisions passes in 1990 and is vetoed, with the Senate override failing by a single vote. The Civil Rights Act of 1991 follows in November of the next year.
The fourth period, from 1993 onward, is one of consolidation punctuated by a second override. Harris sets the harassment threshold in 1993, Hicks refines the pretext rule the same year, Oncale removes the different-sex requirement in 1998, Faragher and Ellerth allocate employer liability in 1998, Sandoval closes private effects enforcement under Title VI in 2001, Burlington Northern sets the retaliation standard in 2006, Ledbetter sets a pay clock in 2007 that Congress resets in 2009, Ricci addresses the tension between the two liability theories in 2009, Vance and Nassar narrow supervisor status and retaliation causation in 2013, and Bostock settles the coverage of sexual orientation and gender identity in 2020.
Read as a whole, the sequence has a property worth naming: the constitutional questions were resolved in the first five years and stayed resolved, while the statutory questions have never stopped moving. That is the opposite of the pattern most readers expect, and it is the reason a person who wants to know what this act requires needs the amendment history as much as the case list.
How the government’s litigating positions shaped the record
The United States is not a bystander in this body of law, and several of the doctrines described above exist in their current form because of positions the executive branch took in litigation.
The Justice Department chose Ollie’s Barbecue as a test case deliberately, because winning on a neighborhood restaurant with no interstate clientele would establish the outer boundary of Title II coverage in the first term rather than leaving it to be litigated establishment by establishment for a decade. Choosing the hardest available facts to litigate first is a strategy with obvious risk, and it worked.
The EEOC’s guidelines on employee selection procedures, issued in the years before Griggs, supplied the validation vocabulary the Court used, and the opinion gave those guidelines substantial deference. In Meritor the Court relied on the agency’s guidelines on sexual harassment in accepting a theory that the statutory text does not name. The agency’s interpretive positions have not always prevailed, and the Court’s willingness to defer to them has varied considerably across the period, but a reader tracing where doctrines came from will find agency guidance upstream of several of them. The rulemaking and guidance record, and the enforcement posture that produced it, belong to the implementation article rather than to this one.
The Solicitor General’s office has appeared on both sides of these questions across administrations, and the EEOC has occasionally taken a position at odds with the Solicitor General’s, most visibly in the litigation that produced the extraterritoriality holding in 1991. Reporting those disagreements institutionally rather than politically is the correct treatment, since the pattern reflects the structural independence of a commission with its own litigating authority more than it reflects any particular administration’s preferences.
The consequence for a reader is a practical one. When a decision seems to have arrived at a framework the statute does not describe, the source is frequently an agency document that predated the case, and locating that document explains the reasoning far better than reading the opinion alone. That is a research move, and the sourcing techniques for it belong to the research cluster of this series.
The remedies map the cases and amendments produced
Holdings about liability are only half of what determines whether a statute changes behavior. The other half is what a successful claimant obtains, and on that question the three titles diverge sharply.
Under Title II, the available relief is an injunction and, following Piggie Park, attorney’s fees for a prevailing plaintiff absent special circumstances. No damages. A person turned away from a covered establishment who wins recovers an order that the establishment stop, plus fees for counsel.
Under Title VI, the statutory sanction is termination of federal financial assistance following an administrative process, and the private right of action recognized by the courts reaches intentional discrimination. After Sandoval, effects claims must proceed administratively. The remedy structure explains why Title VI functions as leverage exercised by funding agencies over recipients rather than as a litigation-driven regime, and why its largest measurable effects came through the threat of fund termination rather than through court judgments.
Under Title VII, the remedial picture changed twice. From 1965 the relief available was equitable: injunctions, reinstatement, back pay under Albemarle, retroactive seniority after Franks v. Bowman Transportation Co., 424 U.S. 747 (1976), and attorney’s fees. From 1991 the statute added compensatory and punitive damages for intentional discrimination, with a jury trial right attached and combined caps scaled to employer size. Back pay remained outside the cap as an equitable make-whole remedy. Disparate impact claims did not gain damages, since the 1991 damages provision reaches intentional discrimination, which is a distinction that surprises many readers and shapes how claims are pleaded.
The practical consequence of the whole map is that the three titles produce three different enforcement cultures from one statute. Public accommodations enforcement is sparse and injunction-shaped. Federal funding enforcement is administrative and negotiated. Employment enforcement is adversarial, insured, and driven by the availability of damages and fees. Anyone comparing the titles’ effectiveness without holding the remedy structure constant is comparing enforcement designs while believing they are comparing commitments.
Three worked examples of the three-round test
The pattern named at the start of this article is useful only if it can be applied, so here are three claims a reader might encounter, resolved by asking which round they come from.
The first claim: Title VII does not require an employer to justify a hiring test unless the plaintiff proves the employer intended to discriminate. This states the law as it stood between June 1989 and November 1991 and not before or after. Griggs placed the justification burden on the employer, Wards Cove reduced it to a burden of production, and the 1991 amendment restored the employer’s burden of proving job relatedness and business necessity once a plaintiff identifies a particular practice causing a disparate impact. A person making this claim has stopped reading at round two.
The second claim: a woman who discovers after a decade that she has been paid less than her male peers has no federal remedy because the discrimination happened years earlier. This states the law as it stood between May 2007 and January 2009. The Ledbetter accrual rule was displaced by the 2009 statute, which provides that each paycheck issued pursuant to a discriminatory compensation decision is itself an unlawful practice. A person making this claim has stopped at the last narrowing decision and missed the override.
The third claim: the Supreme Court rewrote the Civil Rights Act in 2020 to add sexual orientation and gender identity. This misdescribes the mechanism rather than the round. The text of the sex provision was not changed in 2020; the Court construed the words Congress enacted in 1964 and held that they reach these cases as a matter of but-for causation. A reader can disagree with that construction, as three Justices did in writing, and the disagreement is about whether the reading is correct, not about whether the statute was amended. The distinction is not a technicality: an amendment would have required Congress, and a construction can be revisited by the same Court or overridden by the same Congress.
Applied consistently, the test does three things. It dates every claim, which prevents the most common error. It identifies the mechanism, which prevents the second most common one. And it locates the burden of proof for the reader, since the difference between rounds in this statute is usually a difference about who has to prove what. For orientation across the whole statute, including the titles this article does not trace in detail, the complete guide to the act is the hub the rest of this cluster reports to.
Why this statute’s litigation history looks different from other landmark laws
Set the record described above beside the litigation histories of other major postwar statutes and three differences stand out. Naming them makes the pattern in this article usable elsewhere rather than leaving it as a fact about one law.
The first difference is the ratio of statutory to constitutional litigation. Most civil rights statutes of the period faced sustained constitutional attack for years. This one faced a concentrated attack that failed in under six months, after which essentially all of its litigation has been about what words mean. That is a consequence of the drafting choice Congress made in committee, and it is the strongest available argument that constitutional theory selected during drafting is a practical decision with measurable downstream effects rather than a matter of lawyerly preference.
The second difference is the frequency of legislative correction. Congress has amended this act in direct response to judicial construction at least six times. Most statutes accumulate judicial glosses that stand indefinitely, either because no legislative coalition exists to reverse them or because the affected parties lack the visibility to force the question. The overrides here happened because the affected constituencies were organized, the decisions were legible to non-lawyers, and the remedy required only ordinary legislation. Where a statute’s constituency is diffuse or its doctrines technical, the same judicial narrowing produces no legislative answer at all, which is roughly what happened to the Title VI effects regulations after 2001.
The third difference is the division of a single statute into three separate enforcement cultures, described in the remedies map above. Most statutes of comparable importance operate through one enforcement channel. This one operates through three, and the differences among them are the best natural experiment available in postwar federal law on the question of whether a right without a remedy changes behavior. The answer the record supports is qualified: the title with no damages produced fast and broad compliance in its covered categories, and the title with the fullest remedial arsenal produced the slowest measurable change in the conduct it targeted. Anyone drawing a general lesson from that should be careful, because the two titles addressed conduct that differs in visibility, in cost to the regulated party, and in how easily a violation can be proved, and the outcomes evidence and its contested readings are handled in the impact article rather than here.
Held together, these three differences make the act unusually informative for a reader trying to understand how American statutes actually work. It has a clean constitutional test, a long construction record, a repeated override cycle, and internal variation in enforcement design, all in one law. That is why this cluster spends ten articles on it, and why the procedural mechanisms that made each of those legislative corrections possible, including the cloture threshold that governed the 1964 debate and the override attempt that failed by one vote in 1990, are treated in their own right in the filibuster and cloture explained and the veto and the override explained.
The judicially settled shape of the statute
After five decades of construction, the act’s operative shape can be stated in a paragraph, and the fact that it can is the achievement of the litigation record rather than of the enacted text alone.
Title II is constitutionally secure, broadly self-executing in the categories it names, and enforced rarely because it offers no damages. Title VI prohibits intentional discrimination enforceable by private suit and prohibits practices with discriminatory effects enforceable only by the funding agencies, a division created by one decision in 2001 that Congress has not disturbed. Title VII carries two theories of liability, one requiring intent and one requiring only a demonstrated causal relationship between a particular practice and a disparate outcome; a burden-shifting framework for circumstantial claims; a motivating-factor rule with a damages-limiting employer defense; vicarious liability for supervisor harassment subject to a two-element affirmative defense; a retaliation standard reaching any action that would dissuade a reasonable worker, proved to a but-for standard; damages and jury trials capped by employer size; and coverage of sexual orientation and gender identity discrimination as discrimination because of sex.
Not one sentence of that description can be read off the 1964 text. Every clause in it is the product of a decision, an amendment, or a decision followed by an amendment, which is the point this article exists to establish. The statute Congress passed is the beginning of the law, and the law is the text plus the rules the agencies wrote, plus the constructions the courts supplied, plus the corrections Congress made when it disagreed with them.
A reader who takes only one habit from this article should take the dating habit. Every sentence about what this act requires is a sentence about a moment, because the operative content of the law changed in 1972, 1978, 1988, 1991, 2009 and 2020, and changed again with each construction between those dates. Writing “as the statute stood after the 1991 amendments” or “under the rule announced in 1989 and displaced two years later” costs a clause and prevents the failure mode that runs through most secondary writing on the subject, which is confident description of a rule that stopped operating decades ago.
For a reader deciding where to go next: the operative text and its exemptions are walked through in the title-by-title explainer, the amendments that produced six of the entries in the construction ledger are traced in the article on the 1991 act and its predecessors, and the measured effects of all of it are assessed against the statute’s own aims in the outcomes article. The roll calls that produced the text in the first place, and the regional pattern in those votes that most accounts report only half of, are the subject of the vote breakdown for the 1964 act.
Frequently Asked Questions
Q: How did Heart of Atlanta Motel uphold the Civil Rights Act of 1964?
Heart of Atlanta Motel v. United States, 379 U.S. 241 (1964), decided on December 14, 1964, upheld Title II as a valid exercise of the commerce power. The motel had 216 rooms, sat near two interstate highways in downtown Atlanta, advertised nationally, and drew most of its guests from outside Georgia. It refused to rent to Black travelers and sued to have the new statute declared unconstitutional. The Court held that racial discrimination by establishments serving interstate travelers burdens interstate travel, and that Congress may remove that burden under the Commerce Clause. Justice Clark wrote the opinion. The decision arrived less than six months after enactment and permanently closed the constitutional challenge to the act’s most visible title.
Q: Why was the Civil Rights Act of 1964 based on the Commerce Clause?
Because the Civil Rights Cases, 109 U.S. 3 (1883), had held that the Fourteenth Amendment’s enforcement power reaches state action rather than the conduct of private individuals, and had struck down the public accommodations provisions of the Civil Rights Act of 1875 on that ground. That precedent was still standing in 1964. Grounding Title II in the Fourteenth Amendment would have required the Supreme Court to overrule an eighty-one-year-old decision before the statute could operate, with the entire law at risk. Congress instead defined coverage establishment by establishment through commerce connections: transient lodging, food that moved in commerce, entertainment that moved in commerce. The Court upheld that theory in December 1964, and the constitutional question has not seriously recurred.
Q: What did Griggs v. Duke Power decide about the Civil Rights Act of 1964?
Griggs v. Duke Power Co., 401 U.S. 424 (1971), held that Title VII prohibits employment practices that are neutral on their face and in intent but that operate to exclude protected groups, unless the employer shows the practice is related to job performance. Duke Power had required a high school diploma and passing scores on two general aptitude tests for transfer into better-paid departments, neither of which had been shown to predict performance, and both of which disqualified Black applicants at higher rates. The decision established the disparate impact theory, shifted the justification burden to the employer, and prompted employers nationwide to validate or abandon unvalidated tests and credential requirements. Congress codified the theory in the Civil Rights Act of 1991.
Q: Did the Supreme Court ever strike down part of the Civil Rights Act of 1964?
No. The Court upheld Title II against constitutional attack in December 1964 and has never invalidated an operative provision of Public Law 88-352. What it has done is construe specific provisions narrowly, and Congress reversed six of those constructions by statute in 1978, 1988, 1991 and 2009. The confusion usually comes from the Voting Rights Act of 1965, whose coverage formula was held unconstitutional in 2013, suspending the operation of the preclearance mechanism. That is a genuinely different outcome. A statutory construction can be reversed by ordinary legislation, and this statute’s history is largely a record of Congress doing exactly that, while a constitutional holding cannot be reversed the same way.
Q: What is disparate impact under Title VII?
Disparate impact is the theory that an employment practice can violate Title VII because of how it operates, regardless of whether the employer intended to discriminate. A plaintiff demonstrates that a particular practice, such as a written test, a credential requirement, or a physical standard, causes a disparate impact on a protected group. The employer must then demonstrate that the practice is job related for the position and consistent with business necessity. If it does, the plaintiff may still prevail by showing an alternative practice with less impact that the employer refused to adopt. The theory came from Griggs in 1971, was narrowed in 1989, and was written into the statutory text by the Civil Rights Act of 1991.
Q: How did Bostock v. Clayton County change Title VII?
Bostock, decided on June 15, 2020, held that an employer who fires an individual for being gay or transgender discriminates because of sex under Title VII, since it is impossible to act on those grounds without taking the individual’s sex into account. The vote was six to three, with Justice Gorsuch writing for the majority and dissents from Justice Alito, joined by Justice Thomas, and from Justice Kavanaugh. The decision resolved a division among the federal courts of appeals. It did not amend the statute; it construed the text Congress enacted in 1964. The opinion expressly declined to address facilities, dress codes, or the operation of the Religious Freedom Restoration Act in Title VII cases.
Q: What is the McDonnell Douglas framework under Title VII?
It is the burden-shifting method the Court adopted in McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973), for individual discrimination claims proved by circumstantial rather than direct evidence. The plaintiff first establishes a prima facie case by showing protected class membership, qualification, rejection, and that the position remained available. The employer then articulates a legitimate non-discriminatory reason, which Texas Department of Community Affairs v. Burdine, 450 U.S. 248 (1981), clarified is a burden of production only. The plaintiff may then show that the stated reason is pretext. St. Mary’s Honor Center v. Hicks, 509 U.S. 502 (1993), held that disbelieving the employer’s explanation permits but does not compel a finding of discrimination.
Q: Which Civil Rights Act of 1964 cases established sexual harassment law?
Four decisions built the doctrine. Meritor Savings Bank v. Vinson, 477 U.S. 57 (1986), held that a hostile or abusive working environment based on sex violates Title VII even without economic loss, and that the question is whether the conduct was unwelcome. Harris v. Forklift Systems, 510 U.S. 17 (1993), set the threshold at conduct severe or pervasive enough that a reasonable person would find the environment hostile, with no requirement of psychological injury. Oncale v. Sundowner Offshore Services, 523 U.S. 75 (1998), held that harassment between members of the same sex is actionable. Faragher v. City of Boca Raton and Burlington Industries v. Ellerth, both decided in 1998, then allocated employer liability.
Q: What is the Faragher-Ellerth defense?
It is the affirmative defense the Supreme Court created in two decisions issued on the same day in June 1998. An employer is vicariously liable for actionable harassment by a supervisor with authority over the employee. If the harassment culminated in a tangible employment action such as discharge, demotion, or undesirable reassignment, no defense is available. If it did not, the employer may avoid liability by proving both that it exercised reasonable care to prevent and promptly correct harassing behavior, and that the employee unreasonably failed to use the preventive or corrective opportunities provided. Nearly every corporate anti-harassment policy, reporting hotline, training requirement and investigation file exists because of the first element of that defense.
Q: Is Wards Cove Packing still good law?
Only in part, and citing it for the burden rules describes law that stopped operating in November 1991. Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989), held that a disparate impact plaintiff must identify a specific practice, that the comparison must be to the qualified labor market rather than to another internal workforce, and that the employer bears only a burden of production on justification. The Civil Rights Act of 1991 displaced the burden allocation, placing on the employer the burden of demonstrating job relatedness and business necessity. It retained a modified specific-practice requirement, allowing a decision process to be challenged as one practice where its elements cannot be separated for analysis.
Q: What did the Supreme Court hold in Ledbetter v. Goodyear?
Ledbetter v. Goodyear Tire and Rubber Co., 550 U.S. 618 (2007), held by a vote of five to four that the charge-filing period for a pay discrimination claim runs from the discriminatory compensation decision itself, and that later paychecks reflecting that decision do not restart the clock. The plaintiff, who had learned years afterward that her pay had been set below that of comparable male supervisors, had therefore filed too late. Justice Alito wrote for the majority; Justice Ginsburg dissented, arguing that pay disparities are typically concealed and accumulate incrementally, and calling for a legislative response. Congress responded in the Lilly Ledbetter Fair Pay Act of 2009, which makes each paycheck issued under a discriminatory decision a fresh unlawful practice.
Q: Why can private plaintiffs no longer sue over Title VI disparate impact?
Because of Alexander v. Sandoval, 532 U.S. 275 (2001), which held by a vote of five to four that no private right of action exists to enforce disparate impact regulations issued under section 602 of Title VI. The private right recognized under section 601 reaches intentional discrimination only, and the Court held that an agency regulation cannot create a right of action Congress did not create. The regulations themselves remain valid, and funding agencies retain authority to investigate and to pursue administrative remedies including termination of assistance. The practical result was that environmental justice, language access and school discipline claims built on effects regulations lost their private litigation vehicle in a single decision.
Q: Is Brown v. Board of Education a Civil Rights Act case?
No. Brown v. Board of Education, 347 U.S. 483 (1954), is a Fourteenth Amendment decision about state-imposed school segregation, decided ten years before the statute existed. It enforced the Constitution, not a federal statute, and it reached only public institutions. The relationship runs the other direction: the limited compliance with Brown across the following decade is part of the record Congress legislated against, and Titles IV and VI of the 1964 act supplied enforcement tools the judiciary alone had lacked, particularly the threat of terminating federal funds. Attributing Brown to the act is among the most common errors in secondary writing about this period, and it obscures what the statute actually contributed.
Q: What did Price Waterhouse v. Hopkins hold about sex stereotyping?
Price Waterhouse v. Hopkins, 490 U.S. 228 (1989), arose from an accounting firm’s refusal to admit a candidate to partnership after evaluations criticizing her interpersonal style in terms tied to her being a woman, including advice about dressing and behaving more femininely. The Court held that evidence of sex stereotyping is probative of discrimination because of sex, and that where a protected characteristic played a motivating part in a decision, the employer could avoid liability by proving it would have made the same decision anyway. The Court fractured, producing a plurality opinion by Justice Brennan and separate concurrences. The Civil Rights Act of 1991 kept motivating-factor liability while converting the same-decision showing into a limit on remedies.
Q: What did Ricci v. DeStefano decide about discarded test results?
Ricci v. DeStefano, 557 U.S. 557 (2009), held that before an employer may take a race-conscious action to avoid disparate impact liability, it must have a strong basis in evidence to believe it would face such liability if it did not act. New Haven had administered promotional examinations for firefighters, obtained results with a pronounced racial disparity, and discarded them rather than certify the list. Firefighters who scored well sued, arguing that discarding the results because of their racial consequences was intentional discrimination. The Court agreed on that record. The decision is the clearest illustration of how Title VII’s two liability theories can point in opposite directions on identical facts.
Q: Did Title VII ever fail to protect Americans working overseas?
Yes, between March 1991 and November 1991. In EEOC v. Arabian American Oil Co., 499 U.S. 244 (1991), the Court held that Title VII did not apply to United States citizens employed by American companies outside the United States, applying the presumption that federal statutes do not operate extraterritorially unless Congress says so clearly. The decision issued while the legislative response to the 1989 employment cases was already in progress, and the Civil Rights Act of 1991 extended the title’s coverage to American citizens employed abroad by American-controlled employers, subject to an exception where compliance would violate the law of the country where the workplace is located.