By a vote of 6-3, the Supreme Court just ruled in favor of Peggy Young, a UPS driver who claimed she was illegally discriminated against when she was denied a workplace accommodation that was made available to other employees with similar physical restrictions. The Court concluded that her case should not have been dismissed on summary judgment; rather, she should have had the opportunity to prove that UPS, by refusing her an accommodation it made available to many other restricted workers, committed pregnancy discrimination. Justice Alito agreed that UPS’s victory on summary judgment should be vacated and the case remanded, but based on a different interpretation of applicable law.
This ruling caps off more than a decade of controversial litigation over the application of the Pregnancy Discrimination Act to accommodation claims. And while the Court’s reasoning raises many questions, the holding restores protection that pregnant workers should have had by statute, but which they have been denied by many federal courts.
In the first part of a two-part series, we will examine the ruling and explain its scope and reasoning. In the second part, which will appear on this site in two weeks, we will consider the implications of the ruling on other aspects of employment discrimination law.
Hard Labor: Peggy Young and UPS’s Light-Duty Policy
When Peggy Young finally became pregnant after a series of miscarriages, her doctor imposed a restriction on how much she could lift. At the time, she worked as an “air driver” for UPS, a delivery driver who carried lighter letters and packs for United Parcel Service that had arrived by air. UPS decided that it could not continue to allow her to work unless she could lift the amount listed in her job description, even though she rarely if ever was asked to lift things that heavy. She requested a light-duty assignment, but was denied, despite the fact that UPS made such accommodations available to three large groups of employees; those who were injured on the job, those who were eligible for an accommodation under the Americans with Disabilities Act, and those who had lost their commercial driver’s licenses due to a medical condition such as a diabetic complication or a legal condition such as the loss of a license after a drunk driving conviction. Young, however, was denied a similar accommodation. She was forced out of her job, lost her health insurance, and was allowed to return to work only after giving birth.
Birth of a Statute: The Pregnancy Discrimination Act of 1978
Peggy Young sued, claiming that UPS had violated the Pregnancy Discrimination Act (PDA), the heart of federal protection for pregnant workers. The law was passed in 1978, in response to a bizarre opinion by the Supreme Court, General Electric v. Gilbert, which held that pregnancy discrimination was not a form of sex discrimination under Title VII of the Civil Rights Act of 1964, the main federal anti-discrimination law. The Gilbert decision tracked the Court’s equally obtuse ruling under the U.S. Constitution’s Equal Protection Clause, in which the Court infamously distinguished sex discrimination from pregnancy discrimination because the former discriminates between women and men while the latter discriminates between pregnant persons and non-pregnant persons. The Court’s benighted approach allowed employers to continue openly discriminating against pregnant women in a wide variety of ways.
The PDA was designed to disrupt this status quo—to open workplace doors to pregnant women, and to force employers to abandon those policies and practices that treated all pregnant women as incompetent or otherwise made it difficult for them to form or retain labor force attachments. The first clause of the PDA amended Title VII to make clear that discrimination on the basis of “pregnancy, childbirth, or related medical conditions” is unlawful as a form of sex discrimination under Title VII—expressly repudiating both the Court’s reasoning and its holding in Gilbert.
The second clause—the one at issue in Young—provides that pregnant women have the right to be treated the same as others who are “similar in their ability or inability to work,” but “not so affected” by pregnancy. This approach has been criticized for giving pregnant workers only a comparative, rather than an absolute, right of accommodation. But, even setting aside that limitation, courts have struggled for years over what this clause means and how to define the proper comparison group. This disagreement was at the heart of the Young case, as well as several other similar cases over the last several years.
Both the trial and appellate courts held that UPS had not violated the PDA because the UPS light-duty policies did not exclude only pregnancy; there were at least some temporarily disabled employees who, in theory, might need accommodation but would not receive it. As long as the policy was “pregnancy-blind,” it did not run afoul of the PDA. This type of reasoning essentially read the second clause out of the PDA, since the first clause already expressly forbids employment actions taken “because of” pregnancy. In so doing, the Fourth Circuit joined several other circuits, although most were dealing with less egregiously discriminatory policies than UPS’s.
Splitting the Baby: The Supreme Court’s Ruling in Young v. UPS
Peggy Young argued that UPS’s policy was discriminatory because it permitted light-duty accommodations to some workers—potentially many—who had similar types of work restrictions, but did not allow the same accommodation for her. Under the second clause of the PDA, she argued, UPS must grant her the same accommodations available to other workers with similar restrictions.
UPS argued that no policy could violate the PDA if it was pregnancy-neutral—that is, if it did not single out pregnancy as the only condition that did not merit some particular accommodation.
The majority opinion, written by Justice Breyer, split the baby. It rejected the interpretations offered by both parties. With respect to Young’s interpretation, Breyer wrote that pregnant women were not entitled to “most favored nation” status, under which they could demand an accommodation that was offered to any other worker. This, the majority wrote, was too broad a reading of the second clause. (At least Justice Alito, who otherwise took a more narrow approach to clause two, avoided the oddly abstract and impersonal “most favored nation” terminology and instead referred to “most favored employees.”) With respect to UPS’s interpretation, the majority reasoned that such an interpretation would collapse the second clause into the first, in violation of an important principle of statutory construction. And even more damningly, this reading would have allowed the employer’s policy in Gilbert—which covered all sicknesses and accidents—to be upheld despite the incontrovertible fact that the PDA was enacted expressly to overrule that opinion.
The majority, instead, crafted a new approach to applying the Second Clause of the PDA, which, it claims, “minimizes the problems [of the parties’ interpretations], responds directly to Gilbert, and is consistent with longstanding interpretations of Title VII.” The Court’s approach makes use of the so-called McDonnell-Douglas test, which is used to smoke out discriminatory intent by employers accused of unlawful disparate treatment. Under that test, a plaintiff must first make out a prima facie case, demonstrating that she was treated differently from someone similarly situated but outside the protected class. The district court in Young’s case had held that she failed to make out a prima facie case because none of the proposed comparators were “similarly situated”—she hadn’t been injured on the job, she did not qualify for ADA protection, and she hadn’t lost her commercial driver’s license. The policy, in other words, was used in its own defense—she was not similarly situated to anyone covered by the policy because she was not covered by the policy.
Justice Breyer’s opinion rejects this application of McDonnell-Douglas. A plaintiff can establish a prima facie case of pregnancy discrimination simply by showing that “she belongs to the protected class, that she sought accommodation, that the employer did not accommodate her, and that the employer did accommodate others ‘similar in their ability or inability to work.’” This alone will fix a number of the problems in lower court rulings, where courts had found a variety of different ways to prevent light-duty claims from reaching the next stage.
Upon establishment of the prima facie case, the burden of production then shifts to the employer, who must articulate a legitimate, non-discriminatory reason for its differential treatment. Here, the Court interposes another rule to protect pregnancy discrimination plaintiffs. “[C]onsistent with the Act’s basic objective, that reason normally cannot consist simply of a claim that it is more expensive or less convenient to add pregnant women to the category of those (‘similar in their ability or inability to work’) whom the employer accommodates.” Those reasons animate many policies that restrict light-duty accommodations and, indeed, the “employer in Gilbert could in all likelihood have made such a claim.”
After the employer articulates a legitimate, non-discriminatory reason for its treatment of the plaintiff, the plaintiff has the opportunity nonetheless to reach a jury by “providing sufficient evidence that the employer’s policies impose a significant burden on pregnant workers, and that the employer’s ‘legitimate, nondiscriminatory’ reasons are not sufficiently strong to justify the burden, but rather—when considered along with the burden imposed—give rise to an inference of intentional discrimination.” This, too, is new, and gives pregnant workers the opportunity to force the real question underlying all these cases: why would the employer categorically exclude pregnant women from an accommodation that it provides to potentially large numbers of other workers?
For Young, the Court wrote, she could prevail on remand by showing that UPS does not have a sufficiently strong reason for refusing to accommodate pregnant employees with lifting restrictions while accommodating non-pregnant employees with lifting restrictions—“to the point that a jury could find that its reasons for failing to accommodate pregnant employees give rise to an inference of intentional discrimination.”
This was a clear victory for Peggy Young—the ruling below was vacated—but it invites more litigation in pregnancy discrimination cases, as the standard is far from clear or categorical. But it is certainly better than the alternative approach, under which several federal appellate courts upheld blatantly discriminatory policies that excluded pregnant women from routinely available, and sometimes costless, accommodations for no apparent reason.
The Baby Grows Up: The Future of Pregnancy Discrimination Act Claims
The Court’s opinion leaves several issues that will have to be sorted out as pregnancy discrimination claims move forward in the lower courts. The remainder of this column lays out our understanding of what the Court’s decision means for these issues.
How “many” comparators does it take to support an inference of discrimination?
As the Court explains the PDA, an inference of pregnancy discrimination may arise from the non-accommodation of pregnancy despite the accommodation of “many”—but not all—workers with non-pregnancy related disabilities. Lower courts will now have to parse exactly what this means. In our view, this should not require a pregnant worker to identify numerous specific individuals whose conditions the employer treated more favorably. Since the ultimate question is, as the Court says, “the extent to which an employer’s policy treats pregnant workers less favorably than it treats non-pregnant workers similar in their ability to work” (emphasis added), it should be enough for a plaintiff to show that the more favorable treatment extends to numerous non-pregnancy conditions. The fortuity that few individuals may have actually presented with the conditions favored by the employer’s policies should not defeat a pregnancy discrimination claim if such conditions would be accommodated under the employer’s policies.
For example, if an employer’s policy gives paid time off for a wide range of illnesses but not pregnancy, a pregnant worker should still be able to make out a discrimination claim even if, by good fortune, few of the employer’s employees had become ill and taken advantage of the policy. Similarly, if the employer has never before had a disabled employee with a lifting restriction, but its policies make clear that it would accommodate such a (non-pregnant) worker, that too may support the plaintiff’s prima facie case. The bottom line is, where an employer’s policies by their own terms treat a broad range of conditions more favorably than pregnancy—as distinguished from a singular exception or fluke—that should suffice to support an inference of pregnancy discrimination.
What counts as a “legitimate nondiscriminatory reason”?
Clearly, the employer must do more to meet this burden than merely coming up with a way to describe an exclusionary policy in pregnancy-neutral terms. As the Court points out, that kind of formalism is exactly what Congress repudiated in enacting the PDA. Even the General Electric policy in Gilbert, covering illness and accident, can be stated in pregnancy-neutral terms: the policy covers illness and accidents, and pregnancy does not qualify as either. Similarly, as the Court explains, it is not enough to defend the unfavorable treatment of pregnancy by reference to the fact that it would cost more to include it or create added administrative burdens. These explanations too could have supported the General Electric policy. Rather, the employer must provide a reasoned justification of the unfavorable treatment of pregnancy that does not discriminate against pregnancy as the source of the condition causing the work-related effects.
This understanding of the legitimate nondiscriminatory reason (LNDR) makes sense of the Court’s examples of permissible reasons for treating non-pregnant workers more favorably: where such favorable treatment is meted out according to job class, the employer’s needs, age or seniority. In any of those cases, the source of the condition affecting the employee’s ability to work—pregnancy or not-pregnancy—is not the reason for the conferral or denial of the accommodation.
While this means that preferential treatment of employees in some jobs, such as the Court’s example of more dangerous assignments, may be a LNDR, it does not follow that preferential treatment of all on-the-job injuries will necessarily qualify as a LNDR. The latter discriminates based on the source of condition (pregnancy is not an injury, nor is it incurred on the job), while the former does not. Nor do the justifications for singling out employees who hold dangerous jobs for preferential treatment extend to blanket policies limiting accommodations to on-the-job injuries. Many on-the-job injuries result from mundane, everyday misfortunes that could occur anywhere and are not the product of particular dangers on a job. One of us has a friend who holds an office job in marketing at a bank. This otherwise graceful woman stood up from sitting at her computer without noticing that her foot had fallen asleep; she inadvertently landed on her foot wrong, breaking several bones. Hers was classified as an on-the-job injury, triggering her employer’s policies for such conditions. Blanket rules accommodating all on-the-job injuries but not pregnancy necessarily discriminate by source of condition, and are not necessarily (at least when framed as a blanket rule) nondiscriminatory responses to dangerous conditions or special risks on the job.
What proof is required to show pretext?
Much of the time, the McDonnell-Douglas proof method comes down to the pretext stage. Here, the plaintiff has the opportunity to rebut the employer’s non-discriminatory justification by exposing it as a pretext for discrimination. Under the ruling in Young, that means showing that the employer lacks “reasons sufficiently strong” to justify the “significant burden” on pregnant workers. Although Justice Scalia’s dissent accuses the majority of mistakenly conflating the disparate treatment claim with proof of disparate impact and an absence of business necessity, the Court’s approach better fits the theory of the PDA. A core lesson of the PDA, and the history behind it, is that the disfavored treatment of pregnancy often rests on the devaluation of pregnant employees as future mothers and unreliable workers, and the view that pregnant employees are not worth the same investments as other workers needing accommodations for other reasons. Too often, employer refusals to accommodate pregnant employees, despite accommodating other workers with similar impairments, and without good business reasons for such distinctions, rest precisely on the kind of stereotyping that amounts to intentional discrimination. The Court’s test should provide meaningful protection to pregnant workers to challenge the very practices that the PDA intended to eliminate.
Thanks, Justice Kennedy! (Um, well, kind of)
Finally, we can’t end the column without commenting on a citation to the amicus brief we coauthored in the case in support of Peggy Young. Of course, one always hopes to be cited by a majority opinion, but it can be a consolation to be cited in a dissent, knowing at least that your argument helped persuade someone to come out on the “right” side of a case. But it is downright perplexing to be cited by a dissenting Justice when your brief was filed in support of the winning party. Justice Kennedy filed a separate dissenting opinion in which he waxes eloquently about the hardships of discrimination against pregnant women, citing (among other sources), our amicus brief. If only he had found a way to interpret the PDA so as to protect the class for whose rights he expressed such concern.