Pregnancy Classifications are Sex-Based Classifications: A Proposal to Overrule Geduldig – N.Y.U. Journal of Legislation & Public Policy

By Isabel Gutenplan

March 21, 2023

*This is the second post in our series Life after Roe: Grappling with the New Abortion Rights Reality. You can find the rest of the posts in the series here.*

In Dobbs v. Jackson Women’s Health Org.,¹ the Supreme Court eviscerated constitutional law doctrine protecting the right to abortion. Yet, Dobbs is far from the first precedent limiting reproductive rights and justice. For any attempt to move past Dobbs to reestablish constitutional reproductive legal protections, there is one other case that seems to provide a roadblock at every turn: Geduldig v. Aiello.²

In brief, Geduldig concerned a state disability insurance program for private employees who were temporarily disabled from working. The insurance program disqualified disabilities resulting from “normal” pregnancy. In Geduldig, the Supreme Court held that California’s exclusion of disabilities resulting from “normal” pregnancy did not violate the Equal Protection Clause of the Fourteenth Amendment. The reach of Geduldig’s holding and its implication for Equal Protection doctrine is in debate, as evidenced by the majority and dissent in Bray v. Alexandria Women’s Health Clinic.³ Justice Stevens in dissent argued that Geduldig’s holding only extends so far as to say that “not every legislative classification based on pregnancy was equivalent, for equal protection purposes, to . . . explici[t] gender-based distinctions.”⁴ Justice Scalia, on the other hand, seemed to interpret Geduldig to stand for the proposition that pregnancy-based classifications are never facially sex-based classifications.⁵

This distinction—if pregnancy can ever be considered a facial sex-based classification—is critical for ensuring that laws that discriminate on the basis of pregnancy are analyzed under heightened scrutiny.⁶ If pregnancy is not perceived as a sex-based classification, the law or policy under scrutiny will only receive rational basis review, unless it can be shown that the pregnancy classification was intended to discriminate on the basis of sex.⁷ Regardless of the expansiveness of the interpretation, Geduldig is a harmful precedent, as even in its narrowest form, it establishes that not all pregnancy classifications are sex-based.

Geduldig is a major hurdle, not only for abortion rights, but for reproductive rights and justice writ large. We should not concede that Geduldig stands for the proposition that pregnancy-related classifications are never sex-based classifications, as there is a strong argument for reading Geduldig’s holding narrowly.⁸ However, after Dobbs, it is necessary to face this interpretation head on. Though Equal Protection was not directly in front of the Court in Dobbs, Geduldig enabled Justice Alito to dismiss the Equal Protection Clause as source of constitutional protection for the abortion right in a mere paragraph, describing it as “squarely foreclosed by our precedents,” and quoting Geduldig for the proposition that “a State’s regulation of abortion is not a sex-based classification and is thus not subject to the “heightened scrutiny” that applies to such classifications . . . unless the regulation is a “mere pretex[t] designed to effect an invidious discrimination against members of one sex or the other.””⁹ Thus, it appears as though the current majority of the court reads Geduldig expansively, as Justice Scalia did in Bray.

Geduldig must be overruled. Not only is it completely illogical, but it will also prevent the Equal Protection Clause from protecting against any sort of pregnancy-related classification, including those concerning abortion and contraceptives predominately used by women, such as IUDs. It permits discrimination against pregnant people as medical patients, workers, and human beings. Although a head-on attack such as this proposal has the potential to entrench this harmful precedent, the Dobbs decision already assumes a broad interpretation of Geduldig, foreclosing other viable options.

Overruling Precedent Post-Dobbs

This begs the question: how does one go about overruling precedent these days?

Planned Parenthood v. Casey¹⁰ has been the canonical “precedent on precedent” for the past 30 years. It established four factors the Court must consider when tasked with potentially overruling its own precedent: workability, reliance, the law’s growth in intervening years, and changes to factual premises.¹¹ It’s not clear if the stare decisis holding in Casey was overruled by Dobbs. Within the opinion, the Dobbs majority chastises the Casey authors for not considering the nature of the error, but later cites to Casey regarding the importance of stare decisis.¹² Dobbs itself purported to consider five factors in overruling Roe v. Wade and Casey: nature of the error, quality of reasoning, workability, effects on other areas of law, and reliance.¹³ Two of these factors, workability and reliance, are the same as the Casey factors. Two Dobbs factors are completely new: nature of error and quality of reasoning. Effects on other areas of law is similar to the Casey factors focusing on the development of law and fact. However, while Casey’s factors look to see if changes in law and fact in intervening years functionally discount the precedent in question, Dobbs considers if the precedent forced change or constrained other areas of law.

This essay will provide arguments for overruling Geduldig based on a combination of the Casey and Dobbs factors: nature of error, quality of reasoning, workability, reliance, and the development of law and facts, including if the law’s growth has left the central rule discounted by society, if the premises of fact have changed such that the holding is irrelevant or unjustifiable, and/or if the precedent has constrained, warped, or in any way negatively impacted other areas of law.

**Analyzing Geduldig under the Casey and Dobbs Factors**

I. Nature of error

In Dobbs, the Court held “[a]n erroneous interpretation of the Constitution is always important, but some are more damaging than others.”¹⁴ As Dobbs was evaluating a constitutional holding that extended protection, and I am evaluating a holding that removes protection, this analysis will undoubtably take a different form that that in Dobbs. However, the Dobbs analysis is illustrative.

Dobbs stated that Roe “short-circuited” the democratic process by superseding the opportunity for legislative agreement.¹⁵ Geduldig, in many ways, presents the opposite conundrum. The Court’s error was corrected by the democratic process as directly as it possibly could be, outside of passing a constitutional amendment. The Pregnancy Discrimination Act was passed as a direct response to Geduldig and Gilbert v. General Electric Company, which explicitly relied on Geduldig’s logic and applied it to the Title VII context.¹⁶ As Justice Stevens noted in his dissent in Bray,

“[I]n the Pregnancy Discrimination Act, Congress rejected Geduldig’s focus on benefits and overall impact, instead insisting that discrimination on the basis of pregnancy necessarily constitutes prohibited sex discrimination. The statements of the bill’s proponents demonstrate their disapproval of the Court’s reluctance in Gilbert and Geduldig to recognize that discrimination on the basis of pregnancy is always gender-based discrimination.”¹⁷

Yet, Geduldig’s constitutional holding still stands, which, as explored below, has direct effects on other areas of constitutional law despite the swift and absolute rebuke by the people of the Court’s interpretation of “sex discrimination.” This is a particularly damaging precedent that has been continuously rejected by the people. It must be overruled.

II. Quality of reasoning

To use Dobbs’ language, Geduldig is “more than just wrong. It [stands] on exceptionally weak grounds.”¹⁸ In Geduldig, the law in question treated disabilities faced predominantly by men, such as prostatectomies, circumcision, hemophilia, and gout, which received favorable treatment, differently than those faced predominantly by women, such as “normal” pregnancy-related disabilities, which received unfavorable treatment.¹⁹ The majority found this difference to not constitute sex discrimination because “[t]here is no risk from which men are protected and women are not” and “there is no risk from which women are protected and men are not.”²⁰ Though I disagree with the underlying factual assumption that no men have the capacity for pregnancy, even after accepting the premise as true, the reasoning is circular and exposes the weakness of the logic on its own terms.

To illustrate, it appears as though the majority is analogizing, implicitly, to an insurance policy that compensates men and women differently based on a condition that affects men and women relatively equally—for example, an insurance policy that covers car-accident related injuries, but only for men. By this logic, the classification at issue in Geduldig would have received heightened scrutiny had it insured “normal” pregnancy related disabilities for men but left “normal” pregnancy related disabilities for women uninsured. Of course, the reason that the policy was not written this way is precisely because the classification of “normal” pregnancy will only (according to the Court) affect women. When taken to its natural conclusion, the Court’s logic is difficult to justify.

To be clear, not all people affected by pregnancy classifications are women, because not all people who have the capacity for pregnancy are women. Some trans men and non-binary people can become pregnant, just as some cis-gendered women cannot get pregnant. This fact does not weaken the claim that pregnancy classifications are sex-based classifications, because the claim does not rest on the supposition that the classification divides all women versus all men. While not all pregnancy classifications are themselves discriminatory, the classification heightens the possibility of sex discrimination because pregnancy classifications are historically tied to stereotypes about a woman’s place in society as a mother.²¹ Pregnancy classifications may affect trans and non-binary people with capacity for pregnancy in different ways than cis-gendered women, and may be reflective of different stereotypes, but the same invidious discrimination on the basis of sex-stereotyping is still present.²² Pregnancy is not tied to one gender—however, classifications are often tied to stereotypical ideas about how people of certain gender identities and assignments of sex at birth should be operating. As is discussed at length later in this article, under Equal Protection jurisprudence, sex-based classifications may not be justified by “archaic and stereotypic notions.”²³

Further, pregnancy classifications affect all people with the potential of becoming pregnant, regardless of if they are actively carrying a child or not. It ignores reality to claim that the potential to receive disability payments based on pregnancy only affects those that are currently pregnant, an assumption on which the majority presumably relied to claim that the pregnancy classification at issue was not a sex-based classification. The ability to access disability payments influences anyone with the capacity for pregnancy in decisions to get pregnant, continue a pregnancy, work in certain jobs, and conduct personal relationships.

This refusal to recognize pregnancy as a facial sex-based classification enables pregnancy-based classifications to escape heightened scrutiny so long as there are no apparent discriminatory motives for the classification.²⁴ In Bray, Justice Scalia’s majority opinion attempted to differentiate Geduldig’s logic from classifications involving “irrational” objects of disfavor that are exclusively or predominantly held by particular people, from which one can properly infer an intent to disfavor the protected class.²⁵ “A tax on wearing yarmulkes is a tax on Jews.”²⁶ According to the majority, this is different than abortion, because there are rational bases on which to disfavor abortion.²⁷

As the dissent points out, the yarmulkes analogy fails on its own terms. “The yarmulke tax would not become less of a tax on Jews if the taxing authorities really did wish to burden the wearing of yarmulkes. And the fact that many Jews do not wear yarmulkes—like the fact that many women do not seek abortions—would not prevent a finding that the tax—like petitioners’ blockade—targeted a particular class.”²⁸ “A tax on yarmulkes is a tax on Jews” implies that a tax on yarmulkes is a facial classification on the basis of religion, and thus the motives of those imposing the tax should be legally irrelevant to the level of scrutiny required. To say that it is not a facial classification, but only infers invidious discrimination, leaves a gaping hole in Equal Protection jurisprudence by allowing for framing and world play to dictate the promises of our Constitution. It is a legal fiction that stretches reality such that it boggles the mind.

Geduldig “is not just wrong, but…egregiously wrong.”²⁹

III. Workability

Due in large part to its poor logic, Geduldig presents a rule that is exceedingly difficult to “underst[and] and appl[y] in a consistent and predictable manner.”³⁰ To illustrate: in International Union v. Johnson Controls, the Supreme Court addressed a workplace policy in which the company excluded women capable of pregnancy—defined as “all women except those whose inability to bear children is medically documented”—from certain positions in which they would experience exposure to lead.³¹ As the defendant in this case was a private employer, the Court considered a challenged to the policy under Title VII as opposed to the Equal Protection Clause.³² Yet, the Court’s logic in reaching the conclusion that this policy constitutes a facial classification on the basis of sex illustrates the weakness of Geduldig’s constitutional holding. The Court found that the company’s “fetal-protection policy explicitly discriminates against women on the basis of their sex. The policy excludes women with childbearing capacity from lead-exposed jobs and so creates a facial classification based on gender.”³³

The Court in International Union reached this conclusion despite the fact that, as with all pregnancy classifications, the law does not affect all women. Geduldig’s classification was essentially identical, save for a change in time: instead of classifying by those with the capacity for pregnancy, Geduldig’s classification only concerned those currently exercising this capacity. This recognition of a facial classification is difficult to square with Geduldig’s refusal to find one. Further, by the Court’s own logic it does not appear as though the difference in holdings hinged on the use of “women” in the policy considered in International Union. “In its use of the words ‘capable of bearing children’ . . . Johnson Controls explicitly classifies on the basis of potential for pregnancy.”³⁴ Therefore, the Court found that “[u]nder the PDA, such a classification must be regarded, for Title VII purposes, in the same light as explicit sex discrimination.”³⁵

Thus, it appears as though the difference in outcomes can only be explained by a difference in Title VII and Equal Protection jurisprudence. This is hard to square with the Dobbs Court’s insistence that removing constitutional protections returns the issue to the people. If this is truly how our democratic and constitutional systems are supposed to work in concert, when the people respond with explicit and sustained rebuke, instead of developing two completely different doctrinal standards, the Court must accept its mistake. As constitutional law scholars argued in an amicus brief submitted in support of Jackson Women’s Health Organization in the Dobbs litigation, in light of “the Court[’s] gained experience interpreting the Pregnancy Discrimination Act… explain[ing] how certain laws regulating pregnancy could be based on impermissible sex-role stereotypes,”³⁶ Geduldig’s rule proves unworkable.

IV. Reliance

As discussed, the Pregnancy Discrimination Act has nullified any potential reliance interests in the private sphere. As for federal and state actors, it is difficult to imagine how Geduldig could produce reliance interests such that the overruling this precedent would significant damage or destabilize society. Perhaps states have current regulations discriminating on the basis of pregnancy that they would be forced to change, but it’s difficult to think of an example that would present more of a reliance interest than, say, “[t]he ability of women to participate equally in the economic and social life of the Nation,”³⁷ an interest that was discarded in Dobbs. Geduldig, as a law stripping constitutional protection rather than extending it, presents no persuasive reliance interests.

V. Development of Law and Facts

The Pregnancy Discrimination Act was not the only relevant change in law after Geduldig. Geduldig was decided two decades before the Court developed doctrine prohibiting classifications based on sex stereotypes. After United States v. Virginia, “generalizations about ‘the way women are,’” and “estimates of what is appropriate for most women, no longer justify” sex-based classifications.³⁸ Further, later cases “made clear that pregnancy-based regulations anchored in archaic stereotypes about gender roles” to “justify discrimination against women . . . when they were mothers or mothers-to-be . . . can violate the Equal Protection Clause.”³⁹ As many pregnancy-related classifications are steeped in stereotypes about women’s role as mothers and the fallacy of a “normal” pregnancy, the development of sex-stereotyping Equal Protection doctrine demands Geduldig’s logic be revisited.

Geduldig was also based on the faulty factual assumption that there is such a thing as a “normal” pregnancy such that classifications on those grounds are rational. This assumption is likely based in stereotypes itself—since people, from the beginning of time, have carried pregnancies, there must be a baseline pregnancy experience that does not require or deserve accommodations. Moreover, this is likely based on assumptions that pregnancy is a standard life experience for all or most women, and thus accommodations are really “privileges.” This ignores the lived experience of pregnant people, whether that pregnancy is wanted or not.⁴⁰

As discussed above, Geduldig’s holding—or at least, what the current majority interprets Geduldig’s holding to be—has prevented the Court from considering how limitations on the right to abortion discriminate on the basis of sex.⁴¹ Abortion restrictions impermissibly stereotype on the basis of sex, including by the use of assumptions that all women regret their abortions because motherhood is where women naturally thrive. In addition, if future cases challenge Griswold v. Connecticut⁴² and Eisenstadt v. Baird,⁴³ Geduldig will likely stand in the way of the Court’s consideration of the right to certain types of contraceptives under the Equal Protection Clause. While Eisenstadt grounded the right to access contraceptives in the Equal Protection Clause,⁴⁴ the discrimination at issue was between married and unmarried persons, rather than a sex-based classification. If future laws or policies classify on the basis of contraceptives used primarily by men, such as condoms, and those used primarily by women, such as IUDs, Geduldig may stand in the way of the Court’s ability or appetite to analyze the classification as sex-based.

Finally, Geduldig has had other negative impacts, “distort[ing]…important but unrelated legal doctrines.”⁴⁵ In re A.C. demonstrates how the constitutional doctrines of bodily integrity and informed consent are warped by the law’s willingness to treat pregnant people differently than non-pregnant people. ⁴⁶ In re A.C. concerned a tragic case of a pregnant woman in the District of Columbia whose condition was rapidly deteriorating from terminal cancer. Though while competent she consented to a cesarean section at 28 weeks of pregnancy,⁴⁷ it later became clear at 26 and ½ weeks that she would likely die within 24 to 48 hours.⁴⁸ After finding that A.C. was not competent to give her consent to a cesarean section, the trial court ordered the procedure done in an emergency declaration, despite evidence from family and treating physicians that A.C. would have refused consent prior to 28 weeks.⁴⁹ The court denied a stay of the cesarean section even after evidence was put forth that A.C. indicated she did not want the procedure done after being told of the court’s initial decision.⁵⁰ The procedure was done, the baby lived only a few hours, and A.C. died from cancer two days later.⁵¹

On appeal, the D.C. Court of Appeals made clear that the trial court violated A.C.’s “right to accept or forego medical treatment.”⁵² The Court of Appeals found that when deciding the motions, the trial court should have followed the procedure of substituted judgment,⁵³ as is required by the Constitution and common law.⁵⁴ Substituted judgment demands the court attempt to determine what a person deemed incompetent to consent would have consented to for themselves had they been competent.⁵⁵ Instead, the trial court conducted a balancing test, weighing the perceived interests of A.C. of not having the procedure done against the state’s interest, which included explicit consideration of A.C.’s pregnancy and the potential viability of the fetus.⁵⁶ Though the Court of Appeals corrected the legal error, the trial court’s damage could not be reversed.

To be clear, In re A.C. contemplates neither Equal Protection nor Geduldig directly. Yet, the trial court’s refusal to grant A.C. the same substituted judgement it would have offered a non-pregnant person reflects Geduldig’s understanding that the law can be applied differently to pregnant people by virtue of their pregnancy, and in fact in some instances must be applied differently. This is the world order that Geduldig approves and sustains: pregnant people are, in fact, treated unequally.

Conclusion

Every factor considered in Casey and Dobbs counsels in favor of overruling Geduldig. Moreover, Dobbs, by stripping the Constitutional right to abortion under the Due Process Clause and flippantly dismissing the Equal Protection argument, illustrates the need for overruling this destructive precedent. Though this proposal is far from a comprehensive map of future legal doctrine relating to reproductive rights and justice, overruling Geduldig would be a critical step in the right direction.

Isabel Gutenplan, J.D. Class of 2024, N.Y.U. School of Law.

Suggested Citation: Isabel Gutenplan, Pregnancy Classifications are Sex-Based Classifications: A Proposal to Overrule Geduldig, N.Y.U. J. Legis. & Pub. Pol’y Quorum (2023).

1
142 S. Ct. 2228 (2022).
2
417 U.S. 484 (1974).
3
506 U.S. 263 (1993). In Bray, abortion clinics and abortion rights groups sued under the Civil Rights statute §1985 to prevent anti-abortion protesters from trespassing on or obstructing access to abortion clinics. Id. at 266. Although Bray is not an Equal Protection case, as §1985 requires a showing of class-based discriminatory animus, Id. at 267-68, Geduldig was discussed in the context of the clinics’ argument that the clinic protesters were discriminating against women as a class. Id. at 269-71.
4
Id. at 327 (Stevens, J., dissenting).
5
Id. at 271-73, 273 n.3 (1993).
6
See U.S. v. Virginia, 518 U.S. 515, 531 (1996); Personnel Adm’r of Mass. v. Feeney, 442 U.S. 256, 274 (1979).
7
Dobbs, 142 S. Ct. at 2246.
8
See, e.g., Bray at 327 (Stevens, J., dissenting).
9
Dobbs, 142 S. Ct. at 2246 (quoting Geduldig, 417 U.S. at 496 n. 20).
10
505 U.S. 833 (1992).
11
Id. at 855.
12
142 S. Ct. at 2244, 2261-62.
13
Dobbs, 142 S. Ct. at 2265.
14
Id. at 2265. As I hope is evident, my continuous citation of Dobbs is not meant to legitimize its reasoning in any way. Rather, it is an acceptance of the path as it lays before us, and, more than anything, an effort to beat them at their own game.
15
Id. at 2265.
16
Deborah L. Brake, On Not “Having it Both Ways” and Still Losing: Reflections on Fifty Years of Pregnancy Litigation under Title VII, 98 B.U. L. Rev. 995, 998 (2015); see also Gilbert v. Gen. Elec. Co., 429 U.S. 125, 136 (1976).
17
Bray, 506 U.S. at 330-331.
18
Dobbs, 142 S. Ct. at 2266.
19
Geduldig, 417 U.S. at 501 (Brennan, J., dissenting).
20
Id. at 496-97.
21
Brief of Equal Protection Constitutional Law Scholars Serena Mayeri, et al. as Amici Curiae Supporting Respondents at 10-11, Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022) (No. 19-1392) (“Hibbs made clear that pregnancy-based regulations anchored in archaic stereotypes about gender roles can violate the Equal Protection Clause.”)
22
Though not the topic of this article, a similar argument can be made to extend Bostock v. Clayton Cnty, Ga., 140 S. Ct. 1731 (2020), which interpreted “sex-discrimination” in Title VII to prohibit discrimination on the basis of gender-identity and sexual orientation, to the Equal Protection Clause.
23
Miss. Univ. for Women v. Hogan, 458 U.S. 718, 725 (1982).
24
See e.g., Dobbs, 142 S. Ct. at 2246.
25
506 U.S. at 270-71.
26
Id.
27
Id.
28
Bray, 506 U.S. at 325 n. 23 (Stevens, J., dissenting).
29
Dobbs, 142 S. Ct. at 2307.
30
Id. at 2272.
31
499 U.S. 187, 192 (1991).
32
Id.
33
Id. at 197.
34
Id. at 199.
35
Id.
36
Brief of Equal Protection Constitutional Law Scholars, supra note 21, at 8.
37
Casey, 505 U.S. at 856.
38
518 U.S. at 550.
39
Brief of Equal Protection Constitutional Law Scholars, supra note 21, at 10 (quoting Nevada Dep’t. of Hum. Res. v. Hibbs, 538 U.S. 721, 736 (2003)).
40
See Geduldig, 417 U.S. at 500-01 (Brennan, J., dissenting) (“Despite the Code’s broad goals and scope of coverage, compensation is denied for disabilities suffered in connection with a ‘normal’ pregnancy—disabilities suffered only by women . . . Disabilities caused by pregnancy, however, like other physically disabling conditions covered by the Code, require medical care, often include hospitalization, anesthesia and surgical procedures, and may involve genuine risk to life. Moreover, the economic effects caused by pregnancy-related disabilities are functionally indistinguishable from the effects caused by any other disability: wages are lost due to a physical inability to work, and medical expenses are incurred for the delivery of the child and for postpartum care.”).
41
Dobbs, 142 S. Ct. at 2246.
42
381 U.S. 479 (1965) (holding the Due Process Clause’s right to privacy encompasses a married couple’s right to choose to use contraceptives).
43
405 U.S. 438 (1972) (holding the Equal Protection Clause demands that unmarried people cannot be barred from access to contraceptives).
44
405 U.S. at 454-55.
45
Dobbs, 142 S. Ct. at 2275.
46
573 A.2d 1235 (D.C. 1990).
47
Id. at 1238-39.
48
Id. at 1240.
49
Id. at 1239-40.
50
Id. at 1241.
51
Id.
52
Id. at 1243.
53
Id. at 1251-52.
54
Id. at 1247.
55
Id. at 1249.
56
Id. at 1252; In re A.C., 533 A.2d 611, 617 (D.C. 1987).