By: Michael Selmi*


The Civil Rights Act of 1991 (“CRA”) sought to change the employment discrimination landscape.  The CRA overturned or repudiated eight Supreme Court decisions that had narrowed the scope of Title VII in a way that Congress determined was inconsistent with the broad purpose of eradicating employment discrimination.[1]  Relatedly, the CRA transformed Title VII from an equitable relief statute—under which attorneys were most commonly compensated through attorney fee petitions—to a tort-like statute that allows for jury trials and damages for claims relating to intentional discrimination.  The CRA also marked the most comprehensive amendment to the original Civil Rights Act; although Title VII had been amended several times previously, the CRA was, by far, the most substantial amendment then or now.[2]

In this Article, I will explore the effect the CRA has had on employment discrimination litigation, primarily in the Supreme Court, but I will also glance toward litigation in the lower courts.  Based on a review of the Title VII cases and other related cases the Court has decided over the last twenty years,[3] it appears that the CRA had a meaningful restraining effect on the Supreme Court’s jurisprudence.  Since the CRA was passed, the Court has generally, though by no means always, been more supportive of plaintiffs’ Title VII claims than it had been in the years immediately preceding the CRA.  There is, however, an important caveat: in the most ideological cases, or those cases that might have the most dramatic effect on litigation, the Court has remained decidedly pro-defendant.  In other words, in the most meaningful cases, plaintiffs continue to encounter a hostile Supreme Court.  It also appears that the changes wrought by the CRA did not substantially improve outcomes for plaintiffs, though there was, especially in the early years, a dramatic increase in filings.  On the whole, plaintiffs have fared only marginally better on the merits, and employment discrimination cases continue to be more difficult to win than most other comparable civil cases.

This Article will proceed in two primary parts.  The first Part will explore the cases that preceded the CRA and introduce the positive political theory framework through which I want to analyze the Court’s response.  Positive political theory sees the relationship between the judicial, legislative, and executive branches as a political game designed to assert preferences; several scholars have assessed the origins of the CRA against the positive political theory framework.[4]  The second Part of the Article will go beyond those analyses to assess the Supreme Court’s response to see how the Court has adopted a strategically sophisticated approach that has diverged significantly from its interpretative path prior to the CRA.

I. The Making of the Civil Rights Act of 1991

The history behind the CRA is well known and I will provide only a cursory outline, some of which is informed by my experiences as a staff attorney with the Lawyers’ Committee for Civil Rights at the time the Act was under consideration.  Although I played a rather minor monitoring role, the Lawyers’ Committee and my direct supervisor, Rick Seymour, were heavily involved in the drafting and negotiating of the legislation, and I later became enmeshed in some of the early litigation interpreting the CRA.

During the 1980s, the Supreme Court took a deeply conservative turn on issues of civil rights, particularly with respect to employment discrimination.  The Court repeatedly reached adverse results for plaintiffs, and even in cases in which the plaintiffs prevailed, the Court would often impose significant limitations on the employment discrimination doctrine.[5]  There were a substantial number of cases that limited the rights of plaintiffs, but three cases decided during the 1989 term were particularly important in prompting congressional action.

Probably the most significant departure from prior precedent came in the case of Patterson v. McLean Credit Union, in which the Court severely restricted the scope of § 1981.[6]  The statute had never been a major source of employment rights, but it was one of the original civil rights statutes enacted during Reconstruction, and it had a long history that the Supreme Court effectively ignored in holding that the statute only applied to contract formation, and not to acts of discrimination that occurred thereafter.[7]  Not only did the Court limit the statute’s reach, it did so aggressively and on its own initiative.  After the case was initially briefed and argued on the question of the statute’s scope, the Court, on its own motion, called for reargument on whether the statute should apply to private parties, an issue the Supreme Court had addressed in Runyon v. McCrary[8] just a decade earlier and that the parties had not raised.[9]  Ultimately, the Court backed away from that radical reinterpretation, in part because of an outpouring of briefs and critical public reaction,[10] but the Court’s gesture, and its limitation of the statutory scope, sent a signal that settled civil rights principles were up for reconsideration.

During the same term, the Court also rewrote the law with respect to class action settlements.  In Martin v. Wilks, a group of white firefighters sought to intervene in a case that had been resolved through a consent decree many years earlier.[11]  The white firefighters sought to challenge the remedial provisions in the decree, which they argued impermissibly limited their opportunities within the fire department; by a five-to-four vote, the Court permitted the intervention, even though the firefighters had the opportunity to contest the decree when it was originally entered.[12]  This meant that, despite the Court’s frequent admonitions regarding the importance of finality in litigation, it would often be difficult to determine when a settlement embodied in a consent decree could be assumed to be final and free from challenge.  It also meant that a new group of firefighters could challenge settlements that their predecessors had accepted.[13]

The Wilks case also offers important context for understanding the Court’s direction during this time period.  The consent decree at issue in Wilks provided for preferential treatment of African American firefighters as part of the remedies that had been incorporated into the decree.[14]  The case therefore became part of the affirmative action debate that was raging throughout much of the 1980s, and was seen as integral to the Reagan administration’s broader assault on civil rights and affirmative action.[15]  Even though the administration had been unsuccessful in many of the cases,[16] the Wilks case was seen as giving a green light to efforts to dismantle remedial orders.[17]  From this perspective, the Court’s decision was fully consistent with the position espoused by the Reagan administration—and to a lesser extent by the Bush administration—and the Democratic Congress became concerned with what the future might hold.[18]

While Wilks and Patterson reflected the Court’s hostility toward employment discrimination claims, the coup de grâce came with the Court’s decision in Wards Cove Packing Co. v. Atonio, a complicated case in which the Court rewrote part of the disparate impact law to place the burden of proof of the business-necessity defense squarely on the plaintiffs.[19]  Although the Court’s decision was seen as a significant change in the law, it had also become clear that a majority was developing on the Court in favor of eliminating the disparate impact theory altogether—a theory the Court had created in common law fashion in its landmark Griggs decision.[20]  Like Wilks, Wards Cove was also seen as indirectly connected with the affirmative action debate, given that the disparate impact theory had often, and mistakenly, been seen as prompting employers to adopt quotas as a way of avoiding litigation.[21]  The Wards Cove case was decided on the last day of the same term that produced the Wilks and Patterson decisions; shortly after the term ended, Congress began work on what was then dubbed the Civil Rights Act of 1990.[22]

While the three cases discussed above played the strongest role in motivating Congress to act, the Court issued a number of other controversial decisions on smaller issues, all of which made it more difficult for discrimination plaintiffs to obtain relief on their claims.  Some of the issues involved interest on judgments and awards of expert fees,[23] while others involved the timing of claims[24] or the extraterritorial application of Title VII.[25]  Together these cases represented a clear hostility to the interests of plaintiffs—a hostility that became particularly apparent when seen in connection with a series of nonemployment discrimination cases decided at the same time that reflected a broader hostility to civil rights claims, particularly in the area of affirmative action.[26]  When we turn to an assessment of the effects of the CRA, it is important to keep these smaller cases in mind, as in the 1980s no case seemed too small for the Court to side with employers.

Once the legislation began to move through Congress, several obstacles became apparent.  With President George H.W. Bush in office, the Democratically controlled Senate had to secure the votes necessary to overcome a presidential veto.  Many of the proposed provisions were uncontroversial, and there was also widespread agreement that Title VII plaintiffs should be afforded jury trials with damages available.[27]  At the same time, a group of Republican Senators were intent on making the bill part of a larger tort-reform effort and therefore sought to place caps on the damages provisions.[28]  Among some in Congress, there was a sense that capping the damages in Title VII cases would lead to damage caps in other federal statutes, although efforts to impose broader tort reform stalled shortly after the passage of the CRA.[29]

The most controversial part of the legislation was the provision designed to overturn the Wards Cove decision.  As I have argued elsewhere, the disparate impact theory has always rested uneasily within antidiscrimination law and it has likewise always been equated with affirmative action, an issue that was particularly divisive at the time.[30]  The Supreme Court, and politicians, had cautioned against aggressive interpretations of the disparate impact law for fear that employers would be forced to adopt quotas as a way to avoid lawsuits.[31]  This always seemed mostly a specious argument given that the disparate impact theory had been in existence since 1971, with reasonably aggressive interpretations in the 1970s, without any hint of broadscale quota-motivated hiring.[32]  In any event, the rhetoric proved powerful and the disparate impact provisions became hotly contested and produced a series of innovative legislative provisions.

Within the Senate, a debate broke out regarding whether Wards Cove was truly a departure from past precedent, eventually leading to dueling legislative memoranda that were written primarily by interest groups.[33]  At least in this particular instance, Justice Scalia’s theory of statutory interpretation was given full credence, as the memoranda were naked attempts to influence how the statute should be interpreted almost entirely independent of the legislators themselves, though not independent of their staffs, which had been deeply involved in the process.[34]  As a result, the Senate inserted a most peculiar provision into the statute forbidding courts from looking to the legislative history.[35]  The Wards Cove company also got its hands in the legislative cookie jar, as it was worried that the legislative fix might undo its ten-year victory, and the company eventually purchased its own statutory provision that exempted the Wards Cove case from the legislation.[36]

Despite all of the legislative maneuvering, the controversy refused to die, and President Bush vetoed the Civil Rights Act of 1990 due to the disparate impact provision.[37]  Congress failed in its override attempt but immediately set out to craft a new bill, although the prospects for passage remained dim until the Clarence Thomas hearings intervened.[38]  While Congress was debating what was then known as the Civil Rights Act of 1991, Anita Hill’s allegations of sexual harassment surfaced, which led to a public debate over the emerging sexual harassment doctrine.  Those hearings ultimately contributed to the passage of the CRA; indeed, I think it is fair to say that without the hearings, there may not have been a CRA.[39]  There were two reasons for this connection.  First, Missouri Senator John Danforth was both a Republican sponsor of the CRA and the shepherd for Clarence Thomas, who had worked for the Senator many years earlier in the Missouri Attorney General’s office.[40]  After the Anita Hill allegations arose, Senator Danforth pledged to ensure Thomas’s nomination and the passage of the CRA—a pledge he ultimately lived up to.[41]

Second, and sometimes lost in the story, was the realization during the hearings that victims of sexual harassment were often left without any meaningful remedy.  Not only did civil rights advocates have Justice Thomas to thank for the passage of the CRA, Judge Daniel Manion from the Seventh Circuit also chipped in with his own contribution.[42]  In a case involving clear and uncontested sexual harassment, Diane Swanson had been denied any relief since she did not lose her job and therefore did not suffer monetary loss.[43]  Not content to simply deny her relief, the Seventh Circuit went on to conclude that because she was not eligible for any relief, she had no claim and therefore was responsible for the defendant’s court costs, which were taken directly out of her paycheck.[44]

As noted at the outset, the CRA overturned parts or all of eight Supreme Court decisions, and it added important new remedial provisions to the statute.[45]  Equally important, the CRA sent a strong signal that Congress believed the Court was interpreting Title VII too narrowly, and there was language to this effect included in the statutory preface.[46]  The debate over the Act occurred in a very public forum over the course of two years and tainted the arrival of the Court’s newest member.  All of this is to suggest that it would have been difficult for the Supreme Court to ignore the message behind the CRA, but it was less clear whether the Act would have its intended effect.  Some of the provisions were effectively self-executing: the interest and attorney’s fees provisions allowed for little judicial interpretation, and the provision designed to overturn Martin v. Wilks also turned out to be a clear directive that produced no meaningful subsequent litigation.  But the real question was whether the Supreme Court would take Congress’s broader message seriously and interpret Title VII with an eye toward fulfilling the underlying purpose of the CRA rather than with an eye toward protecting employers.

The message sent by—as opposed to the substantive provisions of—the 1991 Act raises important questions about the relationship between Congress and the Court.  Here Congress was not only reversing specific decisions but was also seeking to change the Court’s interpretive direction.  Congress’s oversight powers, however, are limited; Congress could always pass new legislation to change or modify Supreme Court decisions, but short of new legislation the Court is largely free to ignore congressional directives.  Suggesting that the Court is free to ignore congressional directives assumes that the Court may have its own interests or preferences in mind in interpreting statutes.  Most scholars who concentrate on statutory interpretation assume that a court’s judicial duty is to interpret the statute consistent with congressional intent, with the primary area of contention being what Congress intended.[47]  Positive political theorists, on the other hand, treat courts as political actors who desire to implement their own preferences; these theorists typically see modes of statutory interpretation as rhetorical, rather than restraining.[48]  Numerous empirical scholars have also documented that courts frequently decide cases based on judges’ presumed ideological preferences.[49]  For a conservative court, when it comes to issues of employment discrimination, those preferences are most likely to include insulating employers from liability, and as we have just seen, that was how the pre-CRA Court proceeded.

Yet the Court was ultimately unsuccessful, and within the positive political theory framework, a court that wants to implement its preferences must avoid having its decisions overturned.  As a strategic matter, this can lead to some complex analysis, as the Supreme Court would be primarily concerned not with the Congress that enacted a statute but with the current Congress that would be responsible for passing any new legislation, and, similarly, with the President who might veto the legislation.[50]  Under this guise, the Court clearly played the game poorly in its discrimination decisions of the late 1980s, since those decisions all had a very short shelf life.  In light of the CRA’s repudiation of those decisions, we might expect the Supreme Court to change its game plan.  As we will see, that is precisely what it did—and it did so in a strategic way that has protected most of the decisions the Court seems to care most about.

II. The Post-1991 Act Cases

This Part will assess the Supreme Court’s behavior following the passage of the CRA in 1991 and will demonstrate that the Court has acted as positive political theory would predict.  Apparently chastened by the CRA rebuke, the Court has proceeded more wisely, ruling for plaintiffs in the majority of cases, often unanimously, while siding with the interests of employers in the cases that matter most.  During this time period, from 1993 to 2009, the Court’s composition has changed but it has remained a fundamentally conservative Court, one that arguably is more conservative than the Court that issued the decisions that led to the CRA.[51]  It is important, however, to highlight the Court’s process—many of the cases the Court has adjudicated over the last two decades have raised rather minor issues, and a surprising number of cases were simply necessary to reverse plainly incorrect lower court cases.  In these minor cases, the plaintiffs have uniformly prevailed.  But there were also a handful of controversial and important cases, and in those cases the defendants have prevailed, suggesting that the Court was still willing to implement its preferences, at the risk of reversal, in the cases of greatest significance.  Finally, there were cases addressing issues of intermediate importance in which the plaintiffs fared well, and it is in this handful of cases that the Court likely exercised the most judicial restraint.

Before discussing the cases, I must address several preliminary matters.  As noted previously, I have excluded disability cases from the analysis, primarily because the CRA was not aimed at the disability statute.  The Court unquestionably interpreted the ADA narrowly, and Congress recently passed legislation intended to modify the Court’s approach in several of the cases;[52] it will be interesting to see how the Court responds, and if the CRA offers any guidance to the Court’s likely reaction to the statutory repudiation.  I have also excluded most of the cases that involve arbitration agreements since the cases have primarily involved interpretation of a different statute—the Federal Arbitration Act[53]—or issues not directly related to discrimination claims.[54]  I have, however, included the several cases that directly involve discrimination issues.  Finally, I should note that classifying several of the cases has required subjective determinations as to who prevailed in a case, and also as to whom the doctrine is most likely to benefit.  I will highlight where I have made such judgments.

Immediately following the passage of the CRA, the Supreme Court appeared to be up to its old tricks.  The first two cases the Supreme Court addressed involved the retroactivity of the statute—specifically, whether the CRA applied to cases that were pending at the time of enactment.[55]  In both cases, the Court held that the CRA did not apply retroactively but instead only applied to controversies that arose after it was passed.[56]  This effectively delayed implementation for several years.  Notably, however, both decisions were written by Justice Stevens with Justice Blackmun as the lone dissenter, and there was substantial support for the Court’s decision both in the legislative history, which was purposefully left unresolved, and in the body of law that had developed regarding the retroactive application of legislation.[57]  The Court also decided the controversial St. Mary’s Honor Center v. Hicks case after the CRA was passed, but the case itself had little to do with the statute even though the Court’s decision made it more difficult for plaintiffs to prevail in certain cases.[58]  Yet if we view Hicks as part of the post-CRA history, the Court’s treatment is consistent with the pattern we observe in other cases—namely, that the most significant cases remain solidly in the defendants’ camp.

After Hicks, the Court heard a series of cases in which plaintiffs prevailed, often through unanimous decisions.  In Appendix A, I provide a list of the cases decided since the CRA was passed, noting the party that prevailed, the year the case was decided, and the Supreme Court vote breakdown.  The results are revealing: the Supreme Court decided forty-three cases in connection with Title VII, the ADEA, and § 1981, and in twenty-nine, or 67.4%, of those decisions found in favor of the plaintiffs.  The defendants prevailed in thirteen, or 30.2%, of the cases.[59]  Of the forty-three decisions, twenty were unanimous, 46.5% of the total, and, remarkably, eighteen of the unanimous decisions were in favor of plaintiffs.  Indeed, nearly two thirds (62.1%) of the decisions favoring plaintiffs were unanimous.[60]

Table 1: Employment Discrimination Cases Decided 1993–2010

Decision Plaintiff Defendant
Prevail 29 13
Unanimous 18 2
5–4 1 5


The high number of unanimous decisions in favor of plaintiffs offers a sharp contrast to the Court’s decisions rendered prior to the CRA.  During the period 1986–1989, there were five unanimous decisions and all but one had a significant concurring opinion supporting a more limited approach.[61]  Interestingly, none of the unanimous decisions were issued during 1989, when the Court was most active in limiting the scope of Title VII.[62]  In terms of substance, only one of the unanimous cases involved race discrimination, while three involved important issues relating to sex discrimination.  In contrast, the cases that most clearly prompted the CRA—Patterson, Wilks, and, to an extent, Wards Cove—all involved issues of race discrimination, as was true for most of the controversial affirmative action cases that arose during this time period.[63]


Table 2: Employment Discrimination Cases Decided 1987–1991

Decision Plaintiff Defendant
Prevail 10 12
Unanimous 5 0
5–4 1 3
6–3 2 4


The substantial rise in unanimous decisions not only represents a change in course for the Supreme Court but also highlights an important issue embedded in these cases—namely, just how conservative some of the lower courts have become.  Perhaps more accurately, the cases indicate how much more conservative some of the lower courts are compared to what is generally viewed as a very conservative Supreme Court.  While it is difficult to draw any conclusions based on this small sample, it is worth noting that of the eighteen unanimous decisions, eleven originated from the Fourth, Fifth, and Sixth Circuits, with the Sixth leading the way with five unanimous reversals.  Only one of the unanimous decisions was an outright affirmance, and that was the mixed-motives case Desert Palace, Inc. v. Costa,[64] which arose out of the Ninth Circuit, often considered the most liberal appellate court.  Some of the cases involved appellate decisions that were clearly outliers and were essentially summarily reversed.  For example, the Fourth Circuit held that to establish a prima facie age discrimination claim, an individual had to demonstrate that she had been replaced by someone outside of the protected class, a holding that had no support in the statutory language and that the Supreme Court reversed in a seven-paragraph opinion.[65]  The same court also held that former employees could not bring Title VII claims, excising from the statute anyone who was no longer employed.[66]  This latter case had some resemblance to the Supreme Court’s decision in Patterson—as both cases involved the scope of the statute—and suggests that the Supreme Court may have taken seriously Congress’s directive to interpret the statutes consistently with their underlying purposes.

Perhaps the most interesting of the unanimous decisions reversing a hostile lower court was a case that involved racial epithets.  In Ash v. Tyson Foods, Inc., the Eleventh Circuit Court of Appeals had held that the use of “boy,” when directed at an African American man, was not evidence of discriminatory intent unless it was qualified by a racial term such as “black.”[67]  In a per curiam rebuke, the Supreme Court rejected the need for the racial qualifier, noting that whether the term was evidence of discrimination should be considered within its context and in conjunction with additional evidence the plaintiff produced.[68]  In the same case, the Court also rejected the appellate court’s standard—that pretext could only be established by comparison to the employer’s treatment of others to the extent the “disparity in qualifications is so apparent as virtually to jump off the page and slap you in the face.”[69]  Despite the Supreme Court’s sound rejection, the Eleventh Circuit recently reaffirmed a dismissal of the case, albeit under slightly different legal standards.[70]

Another noteworthy aspect of the unanimous cases is that most of the cases were of minor significance.  Several had to do with procedural issues that had not been resolved in the thirty-year history of Title VII, such as the requirements for verifying a complaint and the method for counting employees to meet the statutory coverage requirement.[71]  The Supreme Court decided only two cases during the Wards Cove era that presented similar interpretive questions,[72] and in these and other cases, the Court has recently taken a pragmatic rather than literal linguistic approach.  This was true in the Court’s definition of “employee” and its determination that the number of employees is not a jurisdictional issue, even though there were substantial arguments in support of the other side on both issues.[73]  In the earlier era, it seems quite likely that the Court would have ruled differently, or, more likely, allowed the lower court decisions to stand without review.[74]

Not all of the unanimous cases have turned on minor issues.  The Court held that Title VII prohibits same-sex harassment,[75] an issue that had caused considerable havoc in the lower courts; tossed aside any remnant of the pretext-plus issue;[76] and in two cases, crafted quite liberal principles of law relating to retaliation claims.[77]  Indeed, if there has been any major and surprising turn of events, it has been the Supreme Court’s protective approach to retaliation claims.  Plaintiffs have prevailed in all five retaliation claims the Court has considered, and the Court has adopted an expansive interpretation of the statute in each case.  In one of the cases, the Court had to identify a retaliation claim when the statute was arguably silent or at best ambiguous on the issue.[78]  These cases have helped spark a sharp rise in retaliation claims, but despite that increase, the Court has not sought to cut back on its broad interpretations, and it is difficult to see the Court as anything other than genuinely protective of retaliation claims.[79]

The plaintiff-friendly cases demonstrate that the Supreme Court has moved in a direction that has been more protective of victims of discrimination, but there is an important countertrend that offers a clear balance and also suggests that the Court may be playing a sophisticated political game.  In the most significant cases—including the sole employment case to touch on questions relating to affirmative action[80]—the defendants continue to prevail, and often by five-to-four majorities.  In these cases, the Court continues to impose its preferences, but now does so while also issuing a series of pro-plaintiff decisions, most of which likely do not implicate clear preferences of the Court.  There have been five decisions in favor of defendants by a five-to-four margin, and at least four of the cases are among the most significant decided since the CRA.[81]  As noted previously, the Supreme Court began the post-CRA era with a five-to-four decision in St. Mary’s Honor Center v. Hicks, in which it held that proof of pretext leads to a permissive inference of discrimination rather than the mandatory presumption advocated by the plaintiffs and adopted by the lower court.[82]  That decision also kept alive the damaging pretext-plus theory, though only by disingenuous lower court interpretations, which the Supreme Court abrogated nearly a decade later.[83]

More recently, the Supreme Court has issued several controversial decisions favoring defendants.  In Ricci v. DeStefano, the conservative majority of the Court invalidated an employer’s voluntary efforts to remedy the adverse impact of several promotion tests it had administered.[84]  Reverting to its Wards Cove days, the Court deemed the tests valid even though the tests had not been subject to any legal scrutiny and despite strong arguments that the tests could not be validated under existing law.[85]  The Ricci case has drawn considerable attention and harkens back to the assault on affirmative action from the 1980s, as the Court appeared to view the city’s remedial action as akin to instituting racial preferences for the minority firefighters.[86]  The following term, the Court also held that the mixed-motives theory, often seen as a boon to plaintiffs, was not available under the ADEA, even though the language at issue in the ADEA was quite similar to the language in Title VII that permits such claims.[87]  The difference in results between cases decided under Title VII and the ADEA may be a sign that statutory language can, in fact, restrain the Court.  Although the Court announced a liberal standard for Title VII claims based on language from the CRA, that language did not apply to the ADEA, leaving the Court free to implement its preference on age claims.


Table 3: Five-to-Four Decisions: 1993–2010

  • St. Mary’s Honor Center v. Hicks
  • Ledbetter v. Goodyear Tire & Rubber Co.
  • Kentucky Retirement Systems v. EEOC
  • Ricci v. DeStefano
  • Gross v. FBL Financial Services, Inc.









The other noteworthy five-to-four decision reveals that the Court can still overplay its hand.  In the only employment discrimination case to receive more attention than Ricci, the Supreme Court held that Lilly Ledbetter had waited too long to file her wage discrimination claim.[88]  There were, to be sure, some pragmatic aspects to the case that led the Court to side with the employer.  However, in doing so, the Court imposed a restrictive standard that would have likely foreclosed most wage discrimination claims since it can often take employees years to learn that pay raises were issued in a discriminatory fashion.  In her dissenting opinion, Justice Ginsburg called on Congress to act,[89] and it quickly did so.  Just over a year after the decision was issued, the Lilly Ledbetter Fair Pay Act (“Fair Pay Act”) became the first bill President Barack Obama signed into law, thus reversing the Supreme Court’s decision and likely expanding the statute of limitations beyond what had existed in most of the lower courts.[90]  The case, and the subsequent Fair Pay Act, also drew attention to the issue of pay equity in a way the Court likely did not intend, and there is little question that to the extent the Court was seeking to insulate employers from wage discrimination claims, its Ledbetter decision ultimately had the opposite effect.[91]  Nevertheless, the unintended consequences may prove more theoretical than real, as to date there has not been any significant increase in wage claims, and a bill to address pay equity issues has failed to gain traction.[92]

I should also note that rather than overplaying its hand, the Supreme Court may have misjudged future election results.  The Ledbetter decision was issued toward the end of the Bush presidency but before the Democrats took over the presidency and both houses of Congress.  It is certainly possible that had the decision been issued the following year, the Court may have sought a more moderate path, although its decision in the Ricci case may suggest otherwise.  There is, however, an important distinction between those two cases: Ledbetter was purely a matter of statutory interpretation and relatively easy to overturn, while Ricci represented an amalgam of interpretations of past Supreme Court precedent with overlays of constitutional considerations.  The Ricci case also involved race, whereas Ledbetter presented a more appealing sex discrimination claim that was ripe for congressional review.

In addition to the unanimous decisions for plaintiffs and the five-to-four decisions for defendants, there was a series of cases decided by various margins and also a set of cases in which it was difficult to determine what party would ultimately come out ahead.  These latter cases included three in which the Court provided a legal standard that was generally protective of plaintiff interests, and then carved out an affirmative defense to encourage employers to take precautionary measures.[93]  The Court first took this step in a pair of sexual harassment cases in which it crafted an affirmative defense out of thin air—but a defense that also seemed consistent with the purpose behind the CRA, which is to prevent rather than remedy harassment.[94]  What is perhaps most revealing is that although the language of the affirmative defense should make it difficult for employers to proceed, lower courts have frequently construed the defense more broadly so as to deny plaintiffs relief.[95]  In another case—this one unanimous—the Supreme Court resolved a long-standing split in the circuits by holding that the age discrimination statute permitted disparate impact claims, while creating a very loose standard for employers to justify their practices.[96]

Viewed in their entirety, the cases decided after 1991 reveal a decidedly different Supreme Court from the one that prompted passage of the CRA.  The current Court seems more moderate and less hostile to employment discrimination plaintiffs and remarkably protective of the right to be free from retaliation, but at the same time continues to implement its own preferences when it matters the most.  As a matter of positive political theory, the Court has responded not with timidity but in a strategically sophisticated fashion, and most of its decisions have remained in force.  In other words, the CRA provided a meaningful but not total restraint on the Court’s impulses.

III. The 1991 Act and the Lower Courts

Although the CRA appears to have restrained the Supreme Court, it has had significantly less force in the lower courts.  While I will not go into great detail to demonstrate the hostility to employment discrimination claims at the appellate level, I will highlight three different indicators.  One has already been discussed, and that is the number of cases in which the Supreme Court unanimously reversed lower courts.  In addition, appellate courts have created a number of legal doctrines that make it more difficult for plaintiffs to prove their cases.  Many of the doctrines are evidentiary in nature, but all of them make it more, rather than less, difficult for plaintiffs to prevail.  These doctrines include the creation of a fourth element of the prima facie case that requires plaintiffs (in some circuits) to prove that there is a similarly situated individual who was treated differently, with strict requirements governing who will satisfy the requirement; the stray remarks doctrine; and the same actor inference.[97]  Equally important, no evidentiary rule or legal doctrine has arisen that favors plaintiffs, with the possible exception of some of the emerging case law regarding mixed-motives claims.[98]

Perhaps the strongest indicator of the difficulty plaintiffs face in lower courts is revealed by the many studies that have documented low success rates both at trial and on appeal.  In her Symposium contribution, Professor Wendy Parker surveys the studies,[99] and I will only add a brief summary of my own.  The various studies are all consistent in their findings—more employment discrimination cases go to trial than do other kinds of cases, but plaintiffs typically have a lower success rate.  Plaintiffs succeed in somewhere between 35–40% of their cases tried before a jury, with a significantly lower success rate before a judge.[100]  Even though most cases are now tried before juries, this latter finding is important because judges handle the pre- and post-trial motions, and plaintiffs tend to have a low success rate in defending summary judgment motions.[101]  The data also demonstrate that employment discrimination plaintiffs fare worse than other civil plaintiffs both at trial and on appeal.[102]

It might be that the lower success rates reflect weaker cases, but it is not at all clear why this might be so.  There is no clear reason why employees would file weaker cases, particularly given the filtering process that requires individuals to first proceed through the federal EEOC or the state analogue.[103]  While there may be a higher number of pro se plaintiffs, the absolute number remains very small, and very few ever get to trial.[104]  The settlement values are also typically modest,[105] so these cases should not be particularly attractive to profit-motivated attorneys, though the availability of attorney’s fees might be an additional incentive.  Nevertheless, if the monetary value is not the inducement, the prospect of success surely should be, and again, unless the cases were worth significantly more, attorneys should have the same incentives to bring strong employment discrimination claims as they would to bring other civil cases.[106]  It also strikes me as problematic to assume it is the cases rather than the judges that drive the disparate results—it seems to me the burden should be on explaining what those differences might be rather than simply suggesting employment discrimination cases are less meritorious.  Indeed, the “blame the cases” mentality—which arises in most presentations of the data—mirrors the judicial hostility to employment discrimination cases.

Although it may come as a surprise to some that in the context of employment discrimination cases the lower courts now appear to be more conservative than the Supreme Court, this is less of a surprise within the positive political theory framework.  Congressional action is almost always aimed at the Supreme Court rather than lower courts, and as a result, Congress poses less of a threat to the lower courts.  Instead, the Supreme Court plays the primary restraining role on the appellate courts, and it may be that the chance of review and reversal is so low as to pose only a limited constraint.  At the same time, the prospect of congressional reversal also seems quite low, and it is not clear why one would pose a greater restraint than the other.  It may be that the difference lies in the assumptions behind the process: Supreme Court review is a normal part of the appellate process, whereas congressional action is an extraordinary and public process that typically is directed at cases of greater magnitude.

Whatever the reason, the problem for plaintiffs pursuing employment discrimination claims lies primarily in the lower courts rather than in the Supreme Court; this also makes the prospect for meaningful change more complicated since congressional action is less likely to reshape judicial approaches in the appellate courts.  The Supreme Court might be able to prompt change, but, outside of a handful of aberrational cases, that does not seem to be the Court’s interest.  I think there is little question that the current Supreme Court remains fundamentally conservative and is not likely to have a preference for greater plaintiff success in the lower courts.


The CRA not only reversed a series of decisions but also prompted the Supreme Court to change its interpretive position.  Plaintiffs have fared considerably better in the last two decades than they did in the period immediately preceding the passage of the CRA.  But the Supreme Court has clearly not entirely relented, as it continues to reach conservative results in the cases in which it appears to have the strongest preferences.  Close decisions continue to trend for defendants without much variation, whereas the decisions that side with plaintiffs are now most commonly unanimous, and often short, decisions.  Yet, as noted, the real obstacles for plaintiffs have simply moved to the appellate courts, in which plaintiffs now continually face hostile forums, ones that the Supreme Court is generally willing to accept and that avoid the glare of Congress.  So while the Supreme Court has become a more favorable forum for employment discrimination plaintiffs, conditions on the whole have not significantly improved.

Appendix A

Case Outcome Margin
Harris v. Forklift Sys., 510 U.S. 17 (1993) Pl. 9-0
McKennon v. Nashville Banner Publ’g Co., 513 U.S. 352 (1995) Pl. 9-0
O’Connor v. Consol. Coin Caterers Corp., 517 U.S. 308 (1996) Pl. 9-0
Walters v. Metro. Educ. Enters., Inc., 519 U.S. 202 (1997) Pl. 9-0
Robinson v. Shell Oil, 519 U.S. 337 (1997) Pl. 9-0
Oubre v. Entergy Operations, Inc., 522 U.S. 422 (1998) Pl. 6-3
Oncale v. Sundowner Offshore Servs., 523 U.S. 75 (1998) Pl. 9-0
Burlington Indus., Inc. v. Ellerth, 524 U.S. 742 (1998) Pl. 7-2
Faragher v. City of Boca Raton, 524 U.S. 775 (1998) Pl. 7-2
West v. Gibson, 527 U.S. 212 (1999) Pl. 5-4
Kolstad v. Am. Dental Ass’n, 527 U.S. 526 (1999) Pl. 9-0 & 5-4
Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133 (2000) Pl. 9-0
Pollard v. E.I. du Pont de Nemours & Co., 532 U.S. 843 (2001) Pl. 9-0
EEOC v. Waffle House, Inc., 534 U.S. 279 (2002) Pl. 6-3
Edelman v. Lynchburg Coll., 535 U.S. 106 (2002) Pl. 9-0
Nat’l R.R. Passenger Corp. v. Morgan, 536 U.S. 101 (2002) Pl. 9-0 & 5-4
Clackamas Gastroenterology Assocs., P.C. v. Wells, 538 U.S. 440 (2002) Pl. 7-2
Desert Palace, Inc. v. Costa, 539 U.S. 90 (2003) Pl. 9-0
Jones v. R.R. Donnelley & Sons Co., 541 U.S. 369 (2004) Pl. 9-0
Smith v. City of Jackson, 544 U.S. 228 (2005) Pl. 9-0
Ash v. Tyson Foods, Inc., 546 U.S. 454 (2006) Pl. 9-0
Arbaugh v. Y&H Corp., 546 U.S. 500 (2006) Pl. 8-0
Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53 (2006) Pl. 9-0
Fed. Express Corp. v. Holowecki, 552 U.S. 389 (2008) Pl. 7-2
CBOCS W., Inc. v. Humphries, 553 U.S. 442 (2008) Pl. 7-2
Gomez-Perez v. Potter, 553 U.S. 474 (2008) Pl. 6-3
Meacham v. Knolls Atomic Power Lab., 554 U.S. 84 (2008) Pl. 9-0
Crawford v. Metro. Gov’t of Nashville & Davidson Cnty., 129 S. Ct. 846 (2009) Pl. 9-0
Lewis v. City of Chi., 130 S. Ct. 2191 (2010) Pl. 9-0
Sprint/United Mgmt. Co. v. Mendelsohn, 552 U.S. 379 (2008) None 9-0
Hazen Paper Co. v. Biggins, 507 U.S. 604 (1993) Def. 9-0
St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502 (1993) Def. 5-4
Landgraf v. USI Film Prods., 511 U.S. 244 (1994) Def. 8-1
Rivers v. Roadway Express, Inc., 511 U.S. 298 (1994) Def. 8-1
Comm’r v. Schleier, 515 U.S. 323 (1995) Def. 6-3
Clark Cnty. Sch. Dist. v. Breeden, 532 U.S. 268 (2001) Def. 9-0
Gen. Dynamics Land Sys. v. Cline, 540 U.S. 581 (2004) Def. 6-3
Pa. State Police v. Suders, 542 U.S. 129 (2004) Def. 8-1
Ledbetter v. Goodyear Tire & Rubber Co., 550 U.S. 618 (2007) Def. 5-4
Ky. Ret. Sys. v. EEOC, 554 U.S. 135 (2008) Def. 5-4
AT&T Corp. v. Hulteen, 129 S. Ct. 1962 (2009) Def. 7-2
Gross v. FBL Fin. Servs., Inc., 129 S. Ct. 2343 (2009) Def. 5-4
Ricci v. DeStefano, 129 S. Ct. 2658 (2009) Def. 5-4

Appendix B

Case Outcome Margin
Cal. Fed. Sav. & Loan Ass’n v. Guerra, 479 U.S. 272 (1987) Pl. 6-3
United States v. Paradise, 480 U.S. 149 (1987) Pl. 5-4
Johnson v. Transp. Agency, 480 U.S. 616 (1987) Pl. 6-3
St. Francis Coll. v. Al-Khazraj, 481 U.S. 604 (1987) Pl. 9-0
Goodman v. Lukens Steel Co., 482 U.S. 656 (1987) Def. 6-3
EEOC v. Commercial Office Prods. Co., 486 U.S. 107 (1988) Pl. 6-3
Watson v. Fort Worth Bank & Trust, 487 U.S. 977 (1988) Pl. 9-0
Price Waterhouse v. Hopkins, 490 U.S. 228 (1989) Def. 6-3
Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989) Def. 5-4
Martin v. Wilks, 490 U.S. 755 (1989) Def. 5-4
Lorance v. AT&T Techs., Inc., 490 U.S. 900 (1989) Def. 5-3
Patterson v. McClean Credit Union, 491 U.S. 164 (1989) Def. 5-4
Indep. Fed’n of Flight Attendants v. Zipes, 491 U.S. 754 (1989) Def. 6-2
Pub. Emps. Ret. Sys. v. Betts, 492 U.S. 158 (1989) Def. 7-2
Univ. of Pa. v. EEOC, 493 U.S. 182 (1990) Pl. 9-0
W. Va. Hosps., Inc. v. Casey, 499 U.S. 83 (1991) Def. 6-3
Int’l Union v. Johnson Controls, Inc., 499 U.S. 187 (1991) Pl. 9-0
EEOC v. Arabian Am. Oil Co., 499 U.S. 244 (1991) Def. 6-3
Stevens v. Dep’t of Treasury, 500 U.S. 1 (1991) Pl. 8-1
Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20 (1991) Def. 7-2
Astoria Fed. Sav. & Loan Ass’n v. Solimino, 501 U.S. 104 (1991) Pl. 9-0
Gregory v. Ashcroft, 501 U.S. 452 (1991) Def. 6-3



* Samuel Tyler Research Professor of Law, George Washington University Law School.  An earlier version of this Article was presented at a Symposium held at Wake Forest University School of Law, where I benefitted from the comments and conversations I had at the time.  Particular thanks to Professor Wendy Parker for the invitation to participate and for very helpful suggestions.

[1]. The statute overturned portions of or the entire decisions in eight cases.  See EEOC v. Arabian Am. Oil Co., 499 U.S. 244 (1991) (extraterritorial application of Title VII), superseded by statute, Civil Rights Act of 1991, Pub. L. No. 102-166, § 109, 105 Stat. 1071, 1077; W. Va. Univ. Hosps., Inc. v. Casey, 499 U.S. 83 (1991) (expert witness fees), superseded by statute, Civil Rights Act of 1991 § 113; Patterson v. McLean Credit Union, 491 U.S. 164 (1989) (§ 1981), superseded by statute, Civil Rights Act of 1991 § 101(b); Lorance v. AT&T Techs., Inc., 490 U.S. 900 (1989) (seniority system), superseded by statute, Civil Rights Act of 1991 § 112; Martin v. Wilks, 490 U.S. 755 (1989) (white employees allowed to challenge affirmative action consent decrees), superseded by statute, Civil Rights Act of 1991 § 108; Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989) (disparate impact), superseded by statute, Civil Rights Act of 1991 § 105; Price Waterhouse v. Hopkins, 490 U.S. 228 (1989) (mixed motives), superseded by statute, Civil Rights Act of 1991 § 107(a); Library of Cong. v. Shaw, 478 U.S. 310 (1986) (government’s immunity from interest), superseded by statute, Civil Rights Act of 1991 § 114.

[2]. Prior to the passage of the CRA, Title VII had been amended twice.  In 1972, Title VII was amended to apply the statute to public employers (and to address other smaller issues), and in 1978, the Pregnancy Discrimination Act was passed to amend Title VII so that discrimination based on pregnancy would be considered part of sex discrimination.  See Roy L. Brooks et al., Civil Rights Litigation 402–03 (3d ed. 2005).

[3]. In this Article, I am concentrating on Title VII, the Age Discrimination in Employment Act (“ADEA”), and § 1981, while consciously excluding cases involving the Americans with Disabilities Act (“ADA”).  The ADA was passed in 1990 and the Supreme Court has generally interpreted it very narrowly—so narrowly, in fact, that Congress recently passed a statute to modify several of the decisions.  See ADA Amendments Act of 2008, Pub. L. No. 110-325, 122 Stat. 3553 (codified as amended in scattered sections of 42 U.S.C.).  The 1991 Act, however, was not aimed at the ADA, and the ADA also posed unusual interpretive issues that may have led the Court to effectively rewrite the statute.  Most of what proved to be controversial decisions were not ideologically charged, as several of the cases were decided by seven-to-two margins.  For a discussion of the ADA and the Court’s interpretive approach, see Michael Selmi, Interpreting the Americans with Disabilities Act: Why the Supreme Court Rewrote the Statute, and Why Congress Did Not Care, 76 Geo. Wash. L. Rev. 522 (2008).  Obviously, the article’s title proved a bit too cute.

[4]. See, e.g., William N. Eskridge, Jr., Overriding Supreme Court Statutory Interpretation Decisions, 101 Yale L.J. 331, 385–90 (1991); William N. Eskridge, Jr., Reneging on History?  Playing the Court/Congress/President Civil Rights Game, 79 Calif. L. Rev. 613, 615–16 (1991).

[5]. A case that perhaps fits this mold best was Watson v. Fort Worth Bank & Trust, in which the Court held that subjective employment practices could be challenged under a disparate impact theory but also began to carve out a more rigorous proof structure for those claims.  Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 999–1000 (1988).  This project gained a critical fifth vote three years later in Wards Cove, 490 U.S. at 660.

[6]. Patterson, 491 U.S. at 164.

[7]. See id.

[8]. 427 U.S. 160, 168 (1976).

[9]. Patterson, 491 U.S. at 190–91 (Brennan, J., concurring in the judgment in part and dissenting in part).  For a discussion of the controversy surrounding Patterson, see Donald R. Livingston & Samuel A. Marcosson, The Court at the Crossroads: Runyon, Section 1981, and the Meaning of Precedent, 37 Emory L.J. 949 (1988).  To offer a flavor of what was at stake, the authors explain:

The NAACP Legal Defense Fund sought and obtained participation as amicus curiae by 47 of the 50 states, the American Bar Association, and a bipartisan group of 66 Senators and 119 Representatives.  It sought also the participation of the executive branch as amicus, but the administration decided not to participate.

Id. at 952 n.18; see also Al Kamen, Administration Won’t Argue Rights Case: Solicitor General Upsets Conservatives, Wash. Post, June 24, 1988, at A1.  “The Government had filed a brief on Patterson’s behalf on initial argument which had assumed the validity of Runyon’s interpretation of § 1981, and, given that premise, had supported Patterson’s position that racial harassment could give rise to a valid § 1981 claim against a private employer.”  Livingston & Marcosson, supra, at 952 n.18.

[10]. For a discussion of the case and the public reaction, see John Hope Franklin, The Civil Rights Act of 1866 Revisited, 41 Hastings L.J. 1135 (1990).

[11]. Martin v. Wilks, 490 U.S. 755, 758–59 (1989), superseded by statute, Civil Rights Act of 1991, Pub. L. No. 102-166, § 108, 105 Stat. 1071, 1076–77.

[12]. Id. at 758, 761–63.

[13]. For a critical analysis of the case, see George M. Strickler, Jr., Martin v. Wilks, 64 Tul. L. Rev. 1557 (1990).

[14]. Wilks, 490 U.S. at 758.

[15]. For an excellent discussion of the efforts of the Reagan and Bush administrations to dismantle affirmative action and to challenge other civil rights orthodoxies, see Neal Devins, Reagan Redux: Civil Rights Under Bush, 68 Notre Dame L. Rev. 955 (1992).

[16]. On two previous occasions, the Reagan administration’s efforts to challenge existing decrees had been rebuffed.  See Local No. 93, Int’l Ass’n of Firefighters v. City of Cleveland, 478 U.S. 501, 530 (1986); Firefighters Local Union No. 1784 v. Stotts, 467 U.S. 561, 583 (1984).

[17]. A similar phenomenon was occurring with school desegregation decrees as the Supreme Court was loosening the standards for dismantling those decrees.  The most significant case was decided while the CRA was under consideration.  See Bd. of Educ. v. Dowell, 498 U.S. 237, 247 (1991) (establishing the standard for dissolving a desegregation decree); cf. Missouri v. Jenkins, 495 U.S. 33, 50 (1990) (rejecting the district court’s assertion of power to raise taxes to fund plan implementation).

[18]. Prior to the 1988 term, the Court had vacillated some.  For example, it approved the use of preferential hiring in a gender discrimination case, Johnson v. Transportation Agency, 480 U.S. 616 (1987), and upheld a remedial affirmative action order in United States v. Paradise, 480 U.S. 149 (1987), while invalidating a preferential layoff policy in Wygant v. Jackson Board of Education, 476 U.S. 267 (1986).

[19]. Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 660 (1989), superseded by statute, Civil Rights Act of 1991, Pub. L. No. 102-166, § 105, 105 Stat. 1071, 1074.

[20]. The Griggs decision, recognizing what has come to be known as the disparate impact theory, was not tethered to any particular statutory language but was entirely a product of statutory interpretation.  See Griggs v. Duke Power Co., 401 U.S. 424, 433–34 (1971).  The notion that the Supreme Court was likely to eviscerate the disparate impact theory is based, in large part, on the author’s recollections of the negotiations surrounding the CRA.

[21]. The Court has explicitly made this connection.  See Wards Cove, 490 U.S. at 653; Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 992 (1988).  For a discussion regarding how affirmative action is frequently linked with disparate impact claims, see Mary C. Daly, Affirmative Action, Equal Access and the Supreme Court’s 1988 Term: The Rehnquist Court Takes a Sharp Turn to the Right, 18 Hofstra L. Rev. 1057, 1080–81 (1990).

[22]. See Susan F. Rasky, Rights Groups Work on Measure to Reverse Court’s Bias Rulings, N.Y. Times, Dec. 30, 1989, at 11.

[23]. See W. Va. Univ. Hosps., Inc. v. Casey, 499 U.S 83, 84 (1991) (restricting expert witness fees), superseded by statute, Civil Rights Act of 1991 § 113; Library of Cong. v. Shaw, 478 U.S. 310, 311 (1986), superseded by statute, Civil Rights Act of 1991 § 114.

[24]. Lorance v. AT&T Techs., Inc., 490 U.S. 900, 911–12 (1989) (holding that a seniority system could be challenged as discriminatory only when it was first adopted rather than when it became applicable), superseded by statute, Civil Rights Act of 1991 § 112.

[25]. EEOC v. Arabian Am. Oil Co., 499 U.S. 244, 246 (1991), superseded by statute, Civil Rights Act of 1991 § 109.

[26]. In the same term as Wilks, Patterson, and Wards Cove, the Supreme Court applied strict scrutiny to review and strike down a set-aside program in the city of Richmond, the former capital of the Confederacy.  See City of Richmond v. J.A. Croson Co., 488 U.S. 469, 493, 511 (1989).  That same term, in a rather extraordinary showing, the Court decided five other notable cases, all of which went against the plaintiffs.  See Pub. Emps. Ret. Sys. of Ohio v. Betts, 492 U.S. 158, 168–69 (1989) (holding that a disability plan adopted prior to passage of the ADEA could not be considered a subterfuge for discrimination); Indep. Fed’n of Flight Attendants v. Zipes, 491 U.S. 754, 766 (1989) (holding that no attorneys’ fees would be granted against losing intervenors); Jett v. Dallas Indep. Sch. Dist., 491 U.S. 701, 731 (1989) (holding that § 1983 provided the exclusive remedy for rights violations by state governmental actors); Will v. Mich. Dep’t of State Police, 491 U.S. 58, 71 (1989) (holding that the State is not a person under § 1983); DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 202 (1989) (finding the State not liable under § 1983 for failing to protect a child from abuse).

[27]. M. Isabel Medina, A Matter of Fact: Hostile Environments and Summary Judgments, 8 S. Cal. Rev. L. & Women’s Stud. 311, 325 (1999).

[28]. In an op-ed defending the President’s signing of the bill, White House Counsel C. Boyden Gray made the connection explicit.  See C. Boyden Gray, Civil Rights: We Won, They Capitulated, Wash. Post, Nov. 14, 1991, at A23 (noting that the damages provisions in the CRA [set] “an important precedent for tort reform”).

[29]. 137 Cong. Rec. 30,693 (Nov. 7, 1991) (statement of Rep. Chester Atkins).

[30]. See Michael Selmi, Was the Disparate Impact Theory a Mistake?, 53 UCLA L. Rev. 701, 706, 763–67 (2006).

[31]. Ian Ayres & Peter Siegelman, The Q-Word as Red Herring: Why Disparate Impact Liability Does Not Induce Hiring Quotas, 74 Tex. L. Rev. 1487, 1489–90 (1996); see also Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 993 (1988); Albemarle Paper Co. v. Moody, 422 U.S. 405, 449 (1975).

[32]. For a strong refutation of the argument, see generally Ayres & Siegelman, supra note 31.

[33]. See Nathan Oman, Statutory Interpretation in Econotopia, 25 Pace L. Rev. 49, 68 (2004).

[34]. See Daniel A. Farber, Statutory Interpretation and Legislative Supremacy, 78 Geo. L.J. 281, 289 & n.39 (1989); Oman, supra note 33, at 68.

[35]. Section 105(b) of the CRA reads:

No statements other than the interpretive memorandum appearing at Vol. 137 Congressional Record S 15276 (daily ed. Oct. 25, 1991) shall be considered legislative history of, or relied upon in any way as legislative history in construing or applying, any provision of this Act that relates to Wards Cove—Business necessity/cumulation
/alternative business practice.

Civil Rights Act of 1991, Pub. L. No. 102-166, § 105(b), 105 Stat. 1071, 1075 (codified at 42 U.S.C. § 1981 (2006)).

[36]. The very last provision of the Act, section 402(b), states: “Certain Disparate Impact Cases.—Notwithstanding any other provision of this Act, nothing in this Act shall apply to any disparate impact case for which a complaint was filed before March 1, 1975, and for which an initial decision was rendered after October 30, 1983.”  Id. § 402(b).  It was generally understood that the Wards Cove case was the only case that satisfied this definition.  See Stewart Kwoh, Congress Votes a Cure for All but the Victims, L.A. Times, Nov. 19, 1991, at B37 (noting that the company paid lobbyists more than $175,000 to secure the unusual provision).

[37]. See Steven A. Holmes, President Vetoes Bill on Job Rights; Showdown Is Set, N.Y. Times, Oct. 23, 1990, at A1.

[38]. Initial efforts to override the veto failed by one vote in the Senate.  See Neil A. Lewis, President’s Veto of Rights Measure Survives by 1 Vote, N.Y. Times, Oct. 25, 1990, at A1.  A new bill was later introduced with some modifications.

[39]. Other scholars have reached the same conclusion.  See, e.g., Jerome McCristal Culp, Jr., Neutrality, the Race Question, and the 1991 Civil Rights Act: The “Impossibility” of Permanent Reform, 45 Rutgers L. Rev. 965, 965 (1993) (giving primacy to the role of the Thomas hearings in the passage of the CRA); Devins, supra note 15, at 996 (emphasizing the importance of the Thomas hearings and of David Duke’s run for governor of Louisiana to the passage of the CRA).

[40]. See Linda P. Campbell, GOP’s Danforth Faces a Political Maelstrom: Senator Must Play 2 Roles with Bush, Chi. Trib., Sept. 1, 1991, § 1, at 19.

[41]. See Helen Dewar, White House, Senate Agree on Civil Rights Bill Revisions, Wash. Post, Oct. 25, 1991, at A1, A8.

[42]. See Swanson v. Elmhurst Chrysler Plymouth, Inc., 882 F.2d 1235, 1240 (7th Cir. 1989) (quoting Bohen v. City of E. Chi., 799 F.2d 1180, 1184 (7th Cir. 1986) (“If Congress wishes to amend the provisions of Title VII to provide a remedy of damages, it can do so.”)).

[43]. Id.

[44]. See Judith Lichtman, The Labor Force Needs the 1990 Civil Rights Act, Newsday (N.Y.), May 31, 1990, at 69.

[45]. See supra note 1.

[46]. Civil Rights Act of 1991, Pub. L. No. 102-166, § 3(4), 105 Stat. 1071, 1071.

[47]. See, e.g., Jonathan T. Molot, The Rise and Fall of Textualism, 106 Colum. L. Rev. 1, 1 (2006); Jonathan R. Siegel, The Inexorable Radicalism of Textualism, 158 U. Pa. L. Rev. 117 (2009); Lawrence M. Solan, Private Language, Public Laws: The Central Role of Legislative Intent in Statutory Interpretation, 93 Geo. L.J. 427 (2005).

[48]. The literature on positive political theory is now extensive, but a good summary can be found in McNollgast, Politics and the Courts: A Positive Theory of Judicial Doctrine and the Rule of Law, 68 S. Cal. L. Rev. 1631 (1995).  For a recent literature review, see Michael Abramowicz & Emerson H. Tiller, Citation to Legislative History: Empirical Evidence on Positive Political and Contextual Theories of Judicial Decisionmaking, 38 J. Legal Stud. 419 (2009).

[49]. There is now an extensive empirical literature on judicial decisions, most of which reveals that judges’ political ideology has a statistically significant effect on their decisions.  It is generally the case that political ideology is not determinative in most cases but is clearly significant.  For several recent discussions of the literature, and some critiques, see Cass R. Sunstein et al., Are Judges Political? (2006); William M. Landes & Richard A. Posner, Rational Behavior: A Statistical Study, 1 J. Legal Analysis 775 (2009); and Theodore W. Ruger et al., The Supreme Court Forecasting Project: Legal and Political Science Approaches to Predicting Supreme Court Decisionmaking, 104 Colum. L. Rev. 1150 (2004).  Although the empirical approach has now migrated to law schools, much of the work is concentrated in political science.  For a work coauthored by a leading contributor in the field, see Jeffrey A. Segal & Harold J. Spaeth, The Supreme Court and the Attitudinal Model (1993).

[50]. See John A. Ferejohn & Barry R. Weingast, A Positive Theory of Statutory Interpretation, 12 Int’l Rev. L. & Econ. 263, 270 (1992) (noting that “the preference configuration of the current legislature is far more important for the results of statutory interpretation than is that of the enacting legislature”).

[51]. Justice Alito is more conservative than the Justice he replaced—Justice O’Connor—and this is particularly true on employment discrimination issues.  Chief Justice Roberts is also more conservative than was Chief Justice Rehnquist.  Perhaps most significantly, Justice Clarence Thomas, who joined the Court in 1991, is undeniably far more conservative than Justice Thurgood Marshall.  Justice Kennedy, who joined the Court in 1988, has also solidified his foothold in the conservative wing of the Court on many issues.  On the other side, Justice Ginsburg is more liberal than the Justice she replaced—Justice White—and when Justice Souter replaced Justice Brennan in 1990, he was more conservative but drifted consistently to the left thereafter.

[52]. ADA Amendments Act of 2008, Pub. L. No. 110-325, 122 Stat. 3553 (codified as amended in scattered sections of 42 U.S.C.).

[53]. 9 U.S.C. §§ 1–307 (2006).

[54]. For example, I have excluded Circuit City Stores, Inc. v. Adams, 532 U.S. 105, 109 (2001) (interpreting the Federal Arbitration Act’s exemption for employment of transportation workers), but included EEOC v. Waffle House, Inc., 534 U.S. 279, 282 (2002) (holding that an agreement between employer and employee to arbitrate employment-related disputes does not bar the EEOC from seeking victim-related damages) and Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 23 (1991) (upholding compulsory arbitration agreement for claims arising under the ADEA).

[55]. See Rivers v. Roadway Express, Inc., 511 U.S. 298, 300 (1994) (holding that § 1981 does not apply retroactively); Landgraf v. USI Film Prods., 511 U.S. 244, 247 (1994) (holding that amendments to Title VII do not apply retroactively).

[56]. Rivers, 511 U.S. at 300; Landgraf, 511 U.S. at 247.

[57]. These cases resonated personally with me since I had spent a significant amount of time arguing that the CRA did apply retroactively and was the lead appellate counsel on a case in which the argument was successful (a case in which I prevailed over one of my now colleagues).  See Estate of Reynolds v. Martin, 985 F.2d 470, 471 (9th Cir. 1993).

[58]. St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502, 510–11 (1993) (holding that proof of pretext may, but need not, lead to a finding of discrimination).  The effects of the Hicks case were muted by the introduction of jury trials as part of the CRA.  Civil Rights Act of 1991, Pub. L. No. 102-166, § 102(c), 105 Stat. 1071, 1072.  The Court’s holding in Hicks has less relevance to a jury than to a judge, and has been particularly important at the summary judgment stage.

[59]. One of the cases was functionally a tie, as neither party’s position was adopted.  See Sprint/United Mgmt. Co. v. Mendelsohn, 552 U.S. 379, 387–88 (2008) (holding that so-called me-too evidence was subject to general evidence standards of relevancy rather than to any automatic rule).

[60]. See Table 1, infra.

[61]. For a breakdown of the decisions, see Table 2, infra.  The only decision without some qualification was St. Francis College v. Al-Khazraj, 481 US. 604, 613 (1987), which defined § 1981 to include national origin claims.  The other cases all had some limitations.  See Cal. Fed. Sav. & Loan Ass’n v. Guerra, 479 U.S. 272, 272–73 (1987) (holding, in an opinion by a fractured Court, that a leave policy that offered preferential treatment for women and exceeded the federal standards for pregnancy nondiscrimination was permissible); Bazemore v. Friday, 478 U.S. 385, 400, 407–10 (1986) (handing down a unanimous decision on the relevance of regression analysis but a split decision on the public-accommodations provision); Meritor Sav. Bank v. Vinson, 477 U.S. 57, 66, 72 (1986) (handing down a unanimous decision on the permissibility of the hostile-work-environment theory but withholding judgment on the scope of liability).

[62]. See Int’l Union v. Johnson Controls, Inc., 499 U.S. 187, 197, 212–13 (1991) (analyzing a fetal-protection policy as facially discriminatory with a concurring opinion suggesting that the possibility of third-party lawsuits might constitute adequate justification for the policy); Univ. of Pa. v. EEOC, 493 U.S. 182, 189 (1990) (permitting the EEOC to obtain information relating to a tenure decision).

[63]. Only one of the cases overturned or modified by the CRA substantively involved sex discrimination.  Lorance v. AT&T Techs., Inc., 490 U.S. 900 (1989), superseded by statute, Civil Rights Act of 1991, Pub. L. No. 102-166, § 112, 105 Stat. 1071, 1078–79.

[64]. 539 U.S. 90 (2003) .

[65]. See O’Connor v. Consol. Coin Caterers Corp., 517 U.S. 308 (1996), rev’g 56 F.3d 542 (4th Cir. 1995).

[66]. See Robinson v. Shell Oil Co., 70 F.3d 325 (4th Cir. 1995) (en banc), rev’d, 519 U.S. 337 (1997).

[67]. See Ash v. Tyson Foods, Inc., 129 F. App’x 529, 533 (11th Cir. 2005), rev’d, 546 U.S. 454 (2006).

[68]. See Ash v. Tyson Foods, Inc., 546 U.S. 454, 456 (2006).

[69]. Id. at 456–57 (quoting Ash, 129 F. App’x at 533 (internal quotation marks omitted)).

[70]. See Ash v. Tyson Foods, Inc., 392 F. App’x 817 (11th Cir. 2010).

[71]. See Arbaugh v. Y&H Corp., 546 U.S. 500, 516 (2006) (determining that employee numerosity is not a jurisdictional issue); Edelman v. Lynchburg Coll., 535 U.S. 106 (2002) (upholding an EEOC regulation allowing for post-charge verification); Walters v. Metro. Educ. Enters., Inc., 519 U.S. 202, 212 (1997) (adopting the payroll method for counting employees).

[72]. See Stevens v. Dep’t of Treasury, 500 U.S. 1, 6 (1991) (requiring an employee to file a notice of intent to sue with the EEOC within 180 days of the discriminatory act and at least 30 days before filing suit); Martin v. Wilks, 490 U.S. 755, 764 (1989) (permitting challenges to consent decrees by employees not present at the time the decree was entered), superseded by statute, Civil Rights Act of 1991, Pub. L. No. 102-166, § 108, 105 Stat. 1071, 1076–77.

[73]. See Arbaugh, 546 U.S. at 510–15.

[74]. If the Court were so inclined, it could have reached a different conclusion in Arbaugh, which makes its unanimity all the more surprising.  Given that Title VII only applies to employers with fifteen or more employees, permitting the statute to apply to a smaller employer, as appeared to be the case in Arbaugh, seems to be a stretch.  But the question the Court was likely asking itself is why this issue had not been raised earlier, and it probably relented in the face of a completed trial.  A similar pragmatic result was reached in a case involving whether an EEOC intake form can constitute a charge of discrimination.  See Fed. Express Corp. v. Holowecki, 552 U.S. 389, 402–03 (2008).  The Court affirmed an EEOC regulation permitting this scenario, given that a different conclusion likely would have simply meant that the person would have filed a charge much later to avoid penalizing the plaintiff for the way the EEOC had handled the case.  Id. at 406–07.

[75]. See Oncale v. Sundowner Offshore Servs., Inc., 523 U.S. 75, 82 (1998) (permitting same-sex harassment claims).

[76]. See Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 148–49 (2000) (abrogating pretext-plus decisions by the lower courts).

[77]. See Crawford v. Metro. Gov’t of Nashville & Davidson Cnty., 129 S. Ct. 846, 849 (2009) (holding that an individual who participates in an internal proceeding can proceed on a retaliation claim); Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53, 67–68 (2006) (establishing a standard for a retaliation claim when an employee would be discouraged from filing a claim).

[78]. In Gomez-Perez v. Potter, the Court read into the ADEA a retaliation provision in an opinion by Justice Alito that drew dissents from Chief Justice Roberts and Justices Scalia and Thomas.  Gomez-Perez v. Potter, 553 U.S. 474, 477–79 (2008).  This case can likewise be seen as a purely pragmatic decision, as there was little question that Congress would have inserted the standard retaliation provision into the statute since this appeared to be little more than a drafting error.  See also Crawford, 129 S. Ct. at 849; CBOCS W., Inc. v. Humphries, 553 U.S. 442, 445 (2008) (implying a retaliation cause of action in § 1981); Burlington N., 548 U.S. at 68.  The Burlington Northern case provides an example of just how protective the Court has been, as the plaintiff prevailed in the lower court on a stricter standard than the Supreme Court adopted.  For a discussion of the cases, see Michael Zimmer, A Pro-Employee Supreme Court?  The Retaliation Decisions, 60 S.C. L. Rev. 917, 917 n.2, 919–23 (2009).

[79]. Retaliation claims filed with the EEOC have increased from 21,613 in Fiscal Year 2000 to 33,613 in Fiscal Year 2009.  Charge Statistics, U.S. EEOC, (last visited Apr. 8, 2011).  This term the Court held that certain third-party retaliation claims could be pursued under Title VII; in that particular case, the employer appeared to retaliate against the employee’s fiancée.  See Thompson v. N. Am. Stainless, L.P., 131 S. Ct. 863 (2011).  I should note that I did not include Thompson in the statistical count, given that, as I write this, the term is not yet completed and including only some of the cases might appear misleading.  One might define Clark County School District v. Breeden, 532 U.S. 268 (2001), as a loss for plaintiffs on a retaliation claim, but that case seems to be more about pleading than about retaliation and was a unanimous per curiam decision.  Id. at 271, 274.

[80]. See Ricci v. DeStefano, 129 S. Ct. 2658, 2663–64 (2009).

[81]. See Table 3, infra.

[82]. St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502, 523–24 (1993).

[83]. See Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 148–49 (2000).

[84]. Ricci, 129 S. Ct. at 2664.

[85]. See id. at 2678 (“There is no genuine dispute that the examinations were job-related and consistent with business necessity”).  This statement ignored the numerous objections that had been raised about the test, including its limited utility for assessing skills relevant to higher-level positions and its use as a rank-order device.  Id. at 2707 n.16.  It is my own sense that the tests at issue in Ricci would have been very difficult to justify under existing validation guidelines.  The case has been the subject of extensive critical commentary.  See, e.g., Cheryl I. Harris & Kimberly West-Faulcon, Reading Ricci: Whitening Discrimination, Racing Test Fairness, 58 UCLA L. Rev. 73 (2010).

[86]. Ricci, 129 S. Ct. at 2677.  See generally Helen Norton, The Supreme Court’s Post-Racial Turn Towards a Zero-Sum Understanding of Equality, 52 Wm. & Mary L. Rev. 197 (2010); Richard Primus, The Future of Disparate Impact, 108 Mich. L. Rev. 1341 (2010).

[87]. Gross v. FBL Fin. Servs., Inc., 129 S. Ct. 2343 (2009).

[88]. See Ledbetter v. Goodyear Tire & Rubber Co., 550 U.S. 618, 621 (2007), superseded by statute, Lilly Ledbetter Fair Pay Act of 2009, Pub. L. No. 111-2, 123 Stat. 5.

[89]. Id. at 661 (Ginsburg, J., dissenting).

[90]. See Lilly Ledbetter Fair Pay Act of 2009, Pub. L. No. 111-2, 123 Stat. 5 (codified as amended in scattered sections of 29 U.S.C. and 42 U.S.C.).  The Act amends Title VII’s filing requirement for compensation cases so that an employment practice occurs “when a discriminatory compensation decision or other practice is adopted, when an individual becomes subject to a discriminatory compensation decision or other practice, or when an individual is affected by application of a discriminatory compensation decision or other practice, including each time wages, benefits or other compensation is paid.”  42 U.S.C. § 2000e-5(e)(3)(A).  That standard is potentially broader than any standard that had existed in the lower courts.

[91]. Not only was the Act amended to overturn Ledbetter, but also another bill has been introduced to address pay equity issues.  See Paycheck Fairness Act, S. 182, 111th Cong. (2009); Paycheck Fairness Act, H.R. 12, 111th Cong. (2009).

[92]. The other five-to-four decision was a complicated age discrimination case in which the defendant prevailed, but it would be difficult to characterize the case as significantly disadvantaging older employees because the plan at issue was unusual and the particular case was enmeshed in peculiar facts.  See Ky. Ret. Sys. v. EEOC, 554 U.S. 135, 143–47 (2008).

[93]. See Kolstad v. Am. Dental Ass’n, 527 U.S. 526, 545–46 (1999) (adopting an affirmative defense in connection with a punitive damages claim); Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 765 (1998) (adopting an affirmative defense in sexual harassment cases); Faragher v. City of Boca Raton, 524 U.S. 775, 807–08 (1998) (same).

[94]. As best I have been able to determine, the affirmative defense had not previously been adopted by any court in a sexual harassment case, and it was not presented in any of the briefs filed in the case.  See generally Elizabeth M. Brama, Note, The Changing Burden of Employer Liability for Workplace Discrimination, 83 Minn. L. Rev. 1481 (1999) (discussing prior case law).  This gives credence to Justice Thomas’s claim in dissent that the defense was made up out of “whole cloth.”  See Burlington Indus., 524 U.S. at 771 (Thomas, J., dissenting).

[95]. An early assessment demonstrated that lower courts were interpreting the affirmative defense so that employers who acted appropriately were generally immunized from liability, even if the defense did not technically apply.  See David Sherwyn et al., Don’t Train Your Employees and Cancel Your “1-800” Harassment Hotline: An Empirical Examination and Correction of the Flaws in the Affirmative Defense to Sexual Harassment Charges, 69 Fordham L. Rev. 1265, 1266, 1294 (2001).

[96]. See Smith v. City of Jackson, 544 U.S. 228, 232, 239 (2005).  In Smith, the Supreme Court adopted a “reasonableness” standard based loosely on the portion of the Wards Cove decision that had been overruled by the CRA; much like in the Ricci case, the Court went on to uphold the City’s practice even though the reasonableness of the practice had not been briefed or argued.  Id. at 240–41.

[97]. These doctrines are all discussed in Dianne Avery et al., Employment Discrimination Law 124–32 (8th ed. 2010).

[98]. See, e.g., Rachid v. Jack in the Box, Inc., 376 F.3d 305, 312 (5th Cir. 2004) (adopting a modified standard for summary judgment purposes).  Although the mixed-motives theory has drawn considerable interest among academics, and can be a useful theory at the summary judgment stage, the limited remedies available under the theory render it less useful at trial.

[99]. See generally Wendy Parker, Juries, Race, and Gender: A Story of Today’s Inequality, 46 Wake Forest L. Rev. 185, 192–99 (2011).

[100]. Id. at 195–96.  In charting the win rates for plaintiffs from 1990–2001, Professors Laura Beth Nielson and Robert L. Nelson demonstrated a success rate that ranged from a low of 35.8% (1996) to a high of 43.6% (1992), with an average of 40%.  See Laura Beth Nielson & Robert L. Nelson, Rights Realized?: An Empirical Analysis of Employment Discrimination Litigation as a Claiming System, 2005 Wis. L. Rev. 663, 699 tbl.4.A.  The success rates in bench trials were typically one half that of those in cases tried to juries, with the notable exception of 2001, when plaintiffs prevailed in one third of their bench trials.  Id.

[101]. The difficulty plaintiffs encounter on summary judgment has been well and repeatedly documented.  For two recent analyses, see Elizabeth M. Schneider, The Changing Shape of Federal Civil Pretrial Practice: The Disparate Impact on Civil Rights and Employment Discrimination Cases, 158 U. Pa. L. Rev. 517 (2010), and Elizabeth Schneider, The Dangers of Summary Judgment: Gender and Federal Civil Litigation, 59 Rutgers L. Rev. 705 (2007).

[102]. See Kevin Clermont & Stewart J. Schwab, How Employment Discrimination Plaintiffs Fare in Federal Court, 1 J. Empirical L. Stud. 429, 455 fig.12 (2004).  The difference varies, but for both 1995 and 2001, plaintiffs had a 5% differential at trial, with similar differences on appeal.  Id. at 441 fig.7.

[103]. See Civil Rights Act of 1964, 42 U.S.C. § 2000e-5 (2006); Laura Beth Nielsen et al., Individual Justice or Collective Legal Mobilization?  Employment Discrimination Litigation in the Post Civil Rights United States, 7 J. Empirical L. Stud. 175, 177 (2010).

[104]. Clermont & Schwab, supra note 102, at 434 tbl.1, 440.

[105]. Nielsen et al., supra note 103, at 188.

[106]. I have commented on these issues previously.  See Michael Selmi, Why Are Employment Discrimination Cases So Difficult to Win?, 61 La. L. Rev. 555, 570 (2001).  For an interesting alternative explanation for appellate court hostility that focuses on judicial workload, see Lee Reeves, Pragmatism over Politics: Recent Trends in Lower Court Employment Discrimination Jurisprudence, 73 Mo. L. Rev. 481, 512–22 (2008).

Article in PDF Form