From Medical Exceptions to Reproductive Freedom

David S. Cohen & Greer Donley* December, 2025

Since the Supreme Court overruled Roe v. Wade in 2022, there has been a significant focus on pregnancy complications. This focus has created some risks. Highlighting medically indicated abortion stories could stigmatize the vast majority of abortion seekers who lack medical reasons for abortion and result in narrow remedies that do nothing to undermine abortion bans. This Article presents a way to convert these risks into an opportunity: to use pregnancy complications as a wedge to challenge abortion bans more generally and make abortion more accessible for everyone. We present a long-term strategy to dismantle abortion bans, using several legal theories surrounding pregnancy complications to show that abortion bans are inherently vague, religiously discriminatory, and arbitrary. We then zoom out and show that pregnancy complication cases reveal that Dobbs itself is unworkable and must be overturned. Though the current Supreme Court is unlikely to adopt the theories we describe, there is already evidence that state and lower federal courts are open to them. Moreover, legal losses have the potential to sway public attitudes by revealing how all abortions are health-saving.

Introduction

In the aftermath of Dobbs v. Jackson Women’s Health Organization, which overruled Roe v. Wade, abortion stories have dominated media coverage of the issue.¹ Though most people who have abortions do so because they lack the financial resources, capacity, or desire to have a(nother) child,² these are not the stories filling the media landscape. Rather, the public has been bombarded with stories about pregnancy complications and the challenges of finding proper reproductive healthcare in states where abortion is banned.³ This narrow post-Dobbs focus, which has the seemingly limited goal of easing medical care in emergency situations, risks stigmatizing the vast majority of abortion seekers, who lack medical reasons for abortion. Our Article, however, reveals how medical exceptions could be an essential key to the long-term strategy of overturning Dobbs and restoring reproductive freedom for all.

The broad dissemination of stories about women dying, nearly dying, or birthing dying children because of pregnancy complications has been a key development post-Dobbs.⁴ These stories have been one of the many reasons that support for abortion rights—which was stubbornly stable during Roe’s reign—has increased since Roe was overturned.⁵ They have played a big role in the fourteen successful pro-choice ballot initiatives since 2022.⁶ And they have been partly responsible for changes to abortion laws in states with bans.⁷

One of the most prominent stories has been Amanda Zurawski’s.⁸ Four months into a wanted pregnancy, Amanda was diagnosed with preterm premature rupture of membranes (PPROM), meaning her pregnancy would almost certainly not survive.⁹ Ordinarily, Amanda’s doctor would have recommended an abortion; however, Amanda lived in Texas, where abortion is banned.¹⁰ Consequently, she was unable to get an abortion there and became too sick to travel outside the state.¹¹ Because Texas’s ban has only a narrow exception for pregnancies that threaten the life of the pregnant woman, Amanda and her doctors had to wait until the infection she developed became so severe that she was rushed to the hospital while unresponsive. Only then did her doctors believe they could legally perform an abortion under Texas law, which they did. Amanda barely survived the experience, and the health complications she endured threatened her future fertility.¹²

Amanda survived her horror story, but Porsha Ngumezi did not.¹³ Also in Texas, Porsha was a mother of two boys and pregnant for a third time, hoping for a girl. At ten weeks pregnant, Porsha noticed bleeding. Her doctor told her to go to the emergency room if it got worse, which she did a few days later. In the emergency room, she was passing large clots and in significant pain. Doctors who later reviewed Porsha’s case when it became public agreed that the proper treatment for Porsha’s apparent miscarriage was a procedure known as a dilation and curettage (D&C). However, because that procedure is also used to perform abortions, doctors in states with bans have been reluctant to offer it as miscarriage care, fearful that it would raise legal suspicion.¹⁴

Porsha’s doctor followed this watch-and-wait approach. When the bleeding became even worse, instead of performing the D&C, Porsha’s doctor gave her misoprostol, a drug that causes uterine contractions. Misoprostol is often used to complete a miscarriage, but experts agreed that it was not ideal for Porsha’s case because the drug causes more bleeding and would take too long to work given the severity of her condition. Three hours after taking the medication, Porsha was dead.¹⁵

Despite the poignancy of these stories and how undeniably important they have been in changing public discourse, there are risks in emphasizing them. First, intense focus on the people suffering emergency pregnancy complications might paint other abortion seekers as less deserving of abortion care (if deserving at all).¹⁶ Second, any remedy or reform that might come from pregnancy-complication cases and media attention is likely to be narrow, only applying in rare circumstances. Though fixing the problems with medical exceptions may spare some needless suffering, the risk is that this approach will not eliminate or significantly soften the impact of an abortion ban on those who seek an abortion for non-emergent reasons.

The goal of this Article is to show that these exceptional and moving stories, and the legal issues they produce, have a much broader potential that has been largely ignored. In fact, given the emotional pull of these stories and their resonance with the public, continuing to focus on them could be an essential part of the long-term strategy to dismantle abortion bans generally and to undermine and ultimately overturn Dobbs. Almost from the moment the Supreme Court decided Roe in 1973, the antiabortion movement cultivated wedge issues where they had greater public support—such as abortion for minors, later abortion care, or time and information requirements for patients’ decision-making—to undermine Roe and Casey’s standards and pave the way for the precedents’ ultimate demise.¹⁷ By focusing on these isolated matters, the movement used the narrow to eventually accomplish the broad. Medical exceptions now present the same opportunity for the abortion rights and justice movement to first attack bans and then challenge Dobbs.

At least three distinct constitutional doctrines reveal the potential for these cases to thwart abortion bans more generally. First, pregnancy complications reveal an unconstitutional vagueness within abortion exceptions that violates due process. This vagueness cannot be cured without broad health exceptions that antiabortion legislatures have been unwilling to accept out of a fear that the exception would eat the rule.¹⁸ Second, the existence of secular health exceptions, including exceptions for fetal anomaly and the patient’s life and health, without exceptions for religious belief reveals a serious constitutional infirmity in most abortion bans. Pursuant to the Court’s recent religious liberty jurisprudence under the First Amendment’s Free Exercise Clause, laws are religiously discriminatory when they do not give religious exceptions the same status as secular ones. This is a problem in every abortion ban. Third, pregnancy complications expose how abortion bans lack a rational connection to a legitimate government interest. Though rational basis review is very deferential to state legislatures, there are many irrationalities associated with abortion bans that could invalidate them even under that standard.

Finally, moving from a ground-level to a bird’s-eye view, we argue that medical exception litigation will ultimately contribute to one of the key factors for overturning precedent: unworkability. The inconsistencies coming down the pike on all these issues and more will prove that Dobbs itself is unworkable and should be overturned by a future Supreme Court. Put simply, pregnancy is too complicated for the Dobbs framework to work. Antiabortion legislatures are in a double bind: create a strict ban with no exceptions, which would be extremely unpopular and likely unconstitutional,¹⁹ or try to create narrow exceptions that will inevitably be too vague, discriminatory, or irrational to properly function.

Though we recognize that these arguments are somewhat novel, each has support in the case law and should be a fundamental part of the long-term strategy that advocates must press to reinstate abortion rights and overturn Dobbs.²⁰ The current Supreme Court will almost certainly not view these arguments sympathetically, but a future Court may.²¹ In the meantime, the theories advanced here could deliver big payoffs in lower federal court and state court opinions, as well as in the court of public opinion. One lesson from Dobbs—the culmination of an almost-fifty-year strategy to overturn Roe—is the importance of investing in bold, long-term strategies and remaining steadfast in the face of losses.²²

This Article proceeds as follows. First, in Part I, we explain the landscape of abortion ban exclusions and exceptions for pregnancy complications, review the most publicized pregnancy complication cases to date, and describe the criticisms some have levied against an overemphasis on these cases. Then, in Part II, we apply the doctrines of void-for-vagueness, religious liberty, and rational basis review to medical exceptions and show how they can move us closer to eliminating abortion bans. In Part III, we broaden our lens and argue that the focus on medical exceptions can be an integral part of arguing that Dobbs is unworkable—one of the factors courts use to evaluate whether to overturn precedent. As a result, we ultimately conclude that seemingly narrow pregnancy complication cases should be a key ingredient to restoring reproductive freedom in the future.

I. Pregnancy Complications and Abortion Bans

There is a pervasive narrative that pregnancy is safe, natural, and bound to end with a healthy baby in the arms of a happy mother. But the reality of pregnancy is much more complicated. Nearly 42% of pregnancies in the United States are unintended,²³ and roughly 20% of pregnancies are terminated through abortion.²⁴ Moreover, even when a person is happily pregnant, things go wrong. Around a quarter of pregnancies will end in pregnancy loss—miscarriage before twenty weeks or stillbirth thereafter.²⁵ And roughly 3 to 5% of fetuses are diagnosed with a serious health condition, known as a fetal anomaly, many of which are life-threatening or fatal.²⁶ Pregnant patients can also experience serious risks to their own life and health. The United States has by far the highest maternal mortality of any high-income country: 32.9 maternal deaths per 100,000 live births in 2021.²⁷ This translated into 1,205 deaths that year.²⁸ Moreover, 50,000 to 60,000 Americans each year nearly die or suffer significant harms to their health because of birth or pregnancy, even though they survive.²⁹ There is no denying this simple fact: Pregnancy is much more complicated than people often believe.

Abortion has always been a crucial element to maximizing safety in pregnancy. It ensures that women do not endure the risks and side effects of pregnancy when they do not want to be pregnant.³⁰ It can help complete an inevitable, incomplete, or threatened miscarriage.³¹ It can be compassionate end-of-life care for a baby suffering from a severe fetal anomaly.³² And it is the standard of care for some serious pregnancy complications.³³ Though these unfortunate realities of pregnancy have long been shrouded in stigma and silence,³⁴ the post-Dobbs reality has brought them screaming to the surface.³⁵ Abortion bans have revealed the interconnectedness of abortion and other types of reproductive healthcare. Nearly every day brings a new story of a person denied miscarriage care, ectopic pregnancy treatment, abortion for fetal anomaly, or medically urgent abortions because of abortion bans.³⁶

This Part explains first how states that ban abortion use definitional exclusions and exceptions to attempt to carve out some medically necessary reproductive healthcare—and why those attempts have been inadequate and unsuccessful. It then describes the various litigation strategies that have emerged since Dobbs to clarify and potentially expand those exceptions. It concludes by describing the risk in focusing so intensely on medical emergencies.

A. Carveouts for Pregnancy Complications

Abortion bans have two different mechanisms to exclude medically necessary care. First, some reproductive healthcare is excluded in the statutory definition of abortion—in other words, it is not covered by an abortion ban because the state declares via statute it is not an abortion at all.³⁷ Second, some reproductive healthcare that meets the definition of abortion is permitted through an exception to the abortion ban.³⁸ We refer to the former as exclusions and the latter as exceptions. This Section describes both statutory mechanisms and why they have failed: The ambiguity of exclusions and exceptions combined with the exorbitant penalties for violating abortion bans chills lifesaving healthcare.

1. Definitional Exclusions

Many of the medical carve-outs in abortion bans are not actually created as an exception to the ban. Rather, they are baked into the statute’s definition of abortion: States explicitly define certain types of care as not being an abortion at all.³⁹ For instance, almost all states that ban abortion exclude the “removal of a dead fetus” from their definition of abortion, and a majority exclude treatments for ectopic pregnancy.⁴⁰ Less common definitional exclusions include birth control, molar pregnancy, fertility treatment, accidental fetal death, and abortions for fatal fetal anomaly or for a threat to the pregnant person’s life or health.⁴¹

One might wonder why such exclusions are necessary. First, medically, the term “abortion” refers to what people colloquially call abortion and miscarriage.⁴² Second, most abortion definitions are written broadly enough to include any act done with the intention of ending a pregnancy. And because “the act of terminating a pregnancy is not related to whether the pregnancy is alive or dead, in the process of ending spontaneously, or viable or nonviable,” miscarriage care, ectopic and molar pregnancy treatment, and abortions for nonviable fetuses arguably fall within the medical definition of abortion.⁴³ Thus, if treatments for these conditions were not excluded from the definition, they may qualify as an abortion and be subject to a ban. Abortion bans and fetal personhood efforts have also called into question whether certain types of birth control and fertility care could be banned,⁴⁴ leading some states to try to explicitly protect it through abortion definitions. Notably, most states with abortion bans altered their definitions of abortion since Dobbs to exclude more types of care, likely in response to the barrage of pregnancy complication stories demonstrating the overbreadth of abortion bans.⁴⁵

2. Ban Exceptions

Unlike statutory definitional exclusions, which identify types of care the state does not consider an abortion, ban exceptions permit care that otherwise meets the definition of abortion.⁴⁶ The most common health-related exceptions to abortion bans are those to save the life of the pregnant person (life exceptions) and those to avoid severe health risks to the pregnant person (health exceptions).⁴⁷ All states with abortion bans have a life exception,⁴⁸ something courts and commentators of all political stripes have suggested is constitutionally necessary.⁴⁹ Most states with an abortion ban also have a narrow health exception.⁵⁰ Less common exceptions include fatal fetal anomaly, accidental fetal death, birth control, molar pregnancy, ectopic pregnancy, and removal of a dead fetus.⁵¹

Life exceptions are written simply: Abortions necessary to prevent the death of a pregnant person or to save their life are allowed.⁵² Health exceptions involve a little more variation, but all states with these exceptions use some variant of “necessary . . . to prevent serious risk of substantial and irreversible impairment of a major bodily function.”⁵³ However, most health exceptions specifically exclude threatened self-harm, mental health, or psychological reasons for abortion.⁵⁴ And nearly half require that the abortion be performed in a way that gives the fetus the best chance at life. This specification typically requires a more invasive abortion through labor induction or cesarean section over a less invasive abortion procedure that first kills the fetus, even when the fetus will die regardless.⁵⁵ As to fetal anomaly exceptions, all but one state with such an exception requires the conditions affecting the fetus to be “fatal,” “lethal,” “incompatible with life,” or “nonviable.”⁵⁶ North Carolina is more vague, only requiring “a life-limiting anomaly.”⁵⁷

These definitional exclusions and ban exceptions expose a patchwork of abortion restrictions across states. As we argue later, abortion bans combine two dangerous elements: steep penalties for violations and vague language in the exclusions and exceptions. If they make a mistake, doctors face decades to life in prison, the loss of a medical license, and crippling fines; many providers will avoid or delay any abortion care until they are 100% sure it satisfies an exception or exclusion.⁵⁸ This “skew[ing of] the penalties in one direction” has led to hesitant medicine that has chilled critical reproductive healthcare, endangered lives, and even killed pregnant people.⁵⁹

B. Legal Challenges

The post-Dobbs litigation strategy has been diverse and forceful. Using a mix of old and new legal theories, litigators have challenged bans under state constitutions and have won injunctions in many states, at least temporarily.⁶⁰ Litigators have sued to expand access to abortion medication and defended against lawsuits trying to restrict it.⁶¹ But the cases garnering perhaps the most attention have concerned pregnancy complications. This Section explores two primary efforts and where they stand.

1. Clarifying Abortion Exceptions

Some of the most prominent post-Dobbs litigation efforts have been attempts to seek judicial clarification of abortion exceptions. The lead plaintiff in the first such lawsuit, Amanda Zurawski, was profiled in the introduction of this Article.⁶² The case eventually included twenty patients and two doctors as named plaintiffs.⁶³ The Texas plaintiffs had different stories—some, like Amanda, had their health threatened by a pregnancy; others experienced severe fetal anomalies; and some experienced both, where a fetal anomaly also complicated the woman’s health.⁶⁴ Though the case was successful in the trial court, earning an injunction preventing enforcement of the bans against a physician who, based on “good faith judgment,” concluded that the patient had “a physical emergent medical condition,”⁶⁵ the Texas Supreme Court ultimately vacated the injunction.⁶⁶

While that case was pending, a pregnant woman, Kate Cox, also in Texas, experienced complications and sued because she was unable to receive an abortion. Unlike the plaintiffs in Zurawski, who were suing after the fact, Cox was still pregnant.⁶⁷ Not only was she carrying a fetus with Trisomy 18—a fetal anomaly often characterized as incompatible with life—but she was also experiencing pregnancy risks. She was leaking fluid, at elevated risk for gestational hypertension and diabetes, and facing a high-risk birth, given that she had already had two prior C-sections.⁶⁸ A trial court entered a temporary restraining order, enjoining the enforcement of Texas’s abortion laws against anyone who assisted in Cox’s abortion.⁶⁹ But almost immediately, the Texas Attorney General threatened prosecution against any hospital that participated in Cox’s abortion.⁷⁰ Cox ultimately had to travel out of state to obtain abortion care.⁷¹

On appeal, the Texas Supreme Court held that Cox did not have a “life-threatening physical condition” and therefore did not meet the exception to the ban.⁷² When looking at Cox’s health risks alone, the court found them unavailing: “No one disputes that Ms. Cox’s pregnancy has been extremely complicated. . . . Some difficulties in pregnancy, however, even serious ones, do not pose the heightened risks to the mother the exception encompasses.”⁷³ The Texas Supreme Court’s later decision in Zurawski echoed this sentiment. There, the court noted that “[a]ll pregnancies carry risks,” and that “ordinary risks” are not sufficient to meet the ban’s exception.⁷⁴ It criticized the trial court for “open[ing] the door to permit abortion to address any pregnancy risk.”⁷⁵ Taken together, the Texas Supreme Court expressed skepticism about health exceptions, clearly indicating that Texas’s life exception was meant to be narrow and did not include fetal anomalies, no matter how severe.⁷⁶ The court did, however, offer some guidance for future cases: To justify abortion under the state’s ban, the risk of death must exist but need not be imminent or to have already caused injury,⁷⁷ and to successfully prosecute a provider, the state purportedly⁷⁸ bears the burden of showing that no reasonable provider would have concluded that the exception was met.⁷⁹

Though these cases were not traditionally successful, they nevertheless dominated headlines and changed the conversation about abortion.⁸⁰ In that way, they represented a victory in the court of public opinion and spurred other cases in their wake. Two similar lawsuits were filed in Idaho and Tennessee—Idaho’s case concluded after trial and resulted in a partial victory.⁸¹

2. EMTALA

Another prominent strategy related to medical emergencies has been to use a federal law, the Emergency Medical Treatment and Active Labor Act (EMTALA), to partially preempt state abortion bans. EMTALA requires nearly all hospitals with emergency rooms to screen patients for medical emergencies and, in the event a medical emergency is detected, to provide stabilizing treatment.⁸² In The New Abortion Battleground, the two of us along with Rachel Rebouché described how EMTALA could save pregnant people’s lives in the aftermath of Dobbs by essentially creating a federally mandated health exception to state abortion bans that lack one.⁸³ An EMTALA floor might have limited reach given its narrow definition of emergency medical condition,⁸⁴ but it would be particularly useful in the six states that allow only lifesaving abortions.

Shortly after Dobbs, President Biden’s Department of Health and Human Services (HHS) issued a guidance document noting that EMTALA preempts “state laws or mandates that apply to specific procedures,” like abortion bans.⁸⁵ It clarified that emergency medical conditions could include ectopic pregnancy, complications from pregnancy loss, and emergent pregnancy complications like hypertension or preeclampsia.⁸⁶ And it stated that the treatments to stabilize those emergencies might include abortion procedures, like dilation and curettage.⁸⁷

Quickly after HHS issued this guidance, Texas sued the Biden Administration, arguing that the federal government was attempting “to use federal law to transform every emergency room in the country into a walk-in abortion clinic.”⁸⁸ And shortly thereafter, the federal government filed its own suit against Idaho, arguing that the lack of a health exception in the state’s ban violated EMTALA.⁸⁹ The district courts in Texas and Idaho reached conflicting results,⁹⁰ and it looked like the Fifth and Ninth Circuits were about to split too,⁹¹ when the Supreme Court granted certiorari on an emergency appeal in the Idaho case.⁹² However, after full briefing and argument, the Supreme Court did not reach a decision on the merits, instead dismissing the case as improvidently granted.⁹³ Of concern to many justices was whether EMTALA could be read to require abortions on the basis of mental health.⁹⁴

Though the case was remanded to the Ninth Circuit, the Trump Administration moved to voluntarily dismiss it in March 2025.⁹⁵Anticipating this development, a private hospital in Idaho sued the state on the same theory in January 2025, earning a temporary restraining order to continue the injunction in Idaho after the original lawsuit was dismissed.⁹⁶ This hospital was forced to airlift pregnant patients out of state for emergency care every other week when the Supreme Court stayed Idaho’s injunction, compared to only once per year when the injunction was in place.⁹⁷ The issue will likely return to the Supreme Court in coming years.

C. Risks of Focusing on Medical Exceptions

Stories of medically indicated abortions have certainly played a large role in the post-Dobbs law and policy narratives, and there is some evidence that they are changing hearts and minds.⁹⁸ Tresa Undem—a prominent abortion pollster—has found that these stories are increasingly breaking through to voters.⁹⁹ And “74% of those who had heard stories about pregnant women almost dying strongly agree that Republicans’ views are too extreme—compared to 29% of those who haven’t heard stories about this.”¹⁰⁰ Further, 84% of respondents believe that “[l]aws on abortion can’t account for every situation where an abortion might be needed.”¹⁰¹

But focusing so intensely on pregnancy complication stories poses risks.¹⁰² First, pregnancy complication cases might only generate narrow remedies, affecting so few people that the resolution won’t fully justify the effort. For instance, the requested relief in Zurawski was not striking down the whole Texas abortion ban, but rather broadening the life exception to prevent a small number of future tragedies.¹⁰³ The EMTALA litigation has a similar, narrow aim, especially once the Solicitor General asserted that EMTALA does not protect mental health.¹⁰⁴ In fact, the remedy sought was so narrow that justices at oral argument were confused whether any conflict truly existed between the two parties’ positions.¹⁰⁵ The same is true of legislative victories, where antiabortion states have tweaked their statutes to add more exclusions or slightly broaden exceptions.¹⁰⁶ Though these changes could be lifesaving for the people they impact, they will nevertheless help a tiny number of people.

The second risk is that focusing on pregnancy complications stigmatizes everyone who obtains an abortion for other reasons. There is a vast literature about abortion stigma and the role it plays in limiting and politicizing abortion.¹⁰⁷ Focusing on abortions for people who have wanted pregnancies but are suffering serious complications has the potential to send a message that other abortions are less worthy, even wrong.¹⁰⁸ This concern is particularly noteworthy given some history in the abortion rights movement of centering the abortion stories of white, privileged women choosing abortion for medical reasons.¹⁰⁹

We understand these concerns and take them seriously. However, this Article presents a way to convert this risk into an opportunity—to use pregnancy complications cases to reveal that all abortions are health saving, reducing stigma for all abortion seekers. Notably, unlike prior generations of pregnancy complication storytellers, many viral storytellers today have been women of color, poor women, single moms, and pregnant teenagers.¹¹⁰ Our Article suggests that using these challenges as a wedge—in litigation and in public opinion—can make abortion more accessible for everyone. Below, we present multiple strategies to do just that.

II. Dismantling Abortion Bans Through Medical Exceptions

The first wave of pregnancy complication cases has been a critical tool to demonstrate how abortion bans harm reproductive healthcare and threaten maternal health. But the remedies sought were narrow. In this Part, we outline three legal theories—vagueness, religious liberty, and rational basis review—that can use pregnancy complications to generate broader and more ambitious remedies.

We argue that (1) pregnancy complication exceptions are impermissibly vague and cannot be cured without broad language that the antiabortion movement has been unwilling to accept; (2) the presence of secular exceptions demands religious exceptions that can be used by religiously motivated patients to receive abortions and religiously motivated providers to continue their abortion practice; and (3) pregnancy complications open up arguments that abortion bans fail to establish a rational connection to a legitimate government interest. There is no doubt these theories will be met with much resistance, especially by the current Supreme Court. However, now is the time to build the foundation for the future, while enticing other legal actors—such as litigators and other courts and judges—to explore these theories while we wait for a more hospitable Supreme Court to emerge.¹¹¹

A. Unconstitutional Vagueness

It is a fundamental principle of due process that “[n]o one may be required at peril of life, liberty or property to speculate as to the meaning of penal statutes.”¹¹² A statute is unconstitutionally vague when it “fails to give ordinary people fair notice of the conduct it punishes, or [is] so standardless that it invites arbitrary enforcement.”¹¹³ According to the Supreme Court:

[T]he void for vagueness doctrine addresses at least two connected but discrete due process concerns: first, that regulated parties should know what is required of them so they may act accordingly; second, precision and guidance are necessary so that those enforcing the law do not act in an arbitrary or discriminatory way.¹¹⁴

In addition to due process, the doctrine also implicates separation of powers—it is the sole job of the legislature to say what the law is, and when laws are crafted too vaguely, the legislature is unconstitutionally delegating that role to judges or executive officers.¹¹⁵ The requirements under the vagueness doctrine are more stringent when the statute’s penalties are criminal and severe¹¹⁶ and when the statute impacts constitutionally protected rights, such as the pregnant person’s right to life.¹¹⁷ The closeness between banned conduct and protected conduct matters because “[u]ncertain meanings inevitably lead citizens to ‘“steer far wider of the unlawful zone” . . . than if the boundaries of the forbidden areas were clearly marked.’”¹¹⁸

Roe v. Wade ultimately housed the right to abortion within substantive due process, but in the years leading up to Roe, vagueness was one of the primary theories that courts used to invalidate abortion bans. From 1969 through 1971, courts in Texas, New Jersey, California, Illinois, Georgia, Florida, and DC found abortion bans unconstitutionally vague.¹¹⁹ The vagueness claims were all based on the inherent ambiguity of the bans’ life and health exceptions, and precedent (including Supreme Court precedent) supports a revival of this theory.¹²⁰ The early vagueness cases sound shockingly current despite being over a half-century old. For instance, the district court in Roe v. Wade invalidated Texas’s abortion ban on two theories: vagueness and the Ninth Amendment.¹²¹ As to vagueness of the law’s life exception, the court asked—and was unable to answer—a series of questions¹²² that are arising today—almost verbatim—in court cases, media outlets, and physician accounts.¹²³

The most developed vagueness analysis to invalidate an abortion ban before Roe came from the Supreme Court of California in People v. Belous.¹²⁴ There, the jury convicted a physician under the state’s abortion ban after referring a patient to a trusted abortion provider in Chula Vista, California.¹²⁵ He believed that the patient “would, in fact, do anything to terminate the pregnancy, which might involve butchery in Tijuana or self-mutilation,” putting the patient’s “very life . . . in danger.”¹²⁶

The Belous court reviewed California’s abortion ban, which contained one exception: “unless . . . necessary to preserve [the patient’s] life.”¹²⁷ It found the words “necessary” and “preserve” unconstitutionally vague in this context.¹²⁸ The court went through many possible definitions of this phrase, finding none of them adequate. The state, seeking to cure its vagueness problem, asked the court to adopt a definition that was draconian: requiring “certainty or immediacy of death.”¹²⁹ The court determined that this interpretation would violate the woman’s constitutional right to life: “[T]he state may not require that degree of risk involved in respondent’s definition, which would prohibit an abortion, where death from childbirth although not medically certain, would be substantially certain or more likely than not.”¹³⁰ In the post-Dobbs environment, courts have been similarly quick to reject this narrow interpretation or strike down laws requiring it on the same grounds.¹³¹

Next, the Belous court rejected broad interpretations of the life exception that would render the statute “virtually meaningless.”¹³² These included interpretations permitting abortions whenever a person’s risk exceeded baseline pregnancy risks or whenever abortion was safer than pregnancy.¹³³ The court thought these interpretations would violate the statutory purpose by capturing nearly all abortions. Ruling out these definitions that were clear, but inapposite, the Belous court was left with a definition that permitted abortions when they were “‘substantially or reasonably necessary’ to preserve the life of the mother.”¹³⁴ The court then held that “[i]n the present context[,] those terms are not sufficiently precise and would be subject to such different interpretations as to add little or nothing to ‘necessary.’”¹³⁵

The court noted that some jurors, judges, doctors, and prosecutors would use that language to require near certainty of death, while others would allow abortion any time it was safer than childbirth—i.e., always.¹³⁶ Finally, the court emphasized that “[t]he problem caused by the vagueness of the statute is accentuated because” the state has “in effect, delegated the duty to determine whether a pregnant woman has the right to an abortion” to the physician, who “acts at his peril if he determines that the woman is entitled to an abortion.”¹³⁷ The fact that the “physician has a ‘direct, personal, substantial, pecuniary interest in reaching a conclusion’ that the woman should not have an abortion” creates a chill that may “as effectively condemn[ women] to death as if the law flatly prohibited all abortions.”¹³⁸

In the years immediately before Roe, vagueness challenges like Belous were having such an impact that the Supreme Court took an abortion vagueness case on direct appeal from the D.C. District Court.¹³⁹ The Supreme Court noted in Vuitch that the varying results in vagueness cases meant that “[a] refusal to accept jurisdiction here would only compound confusion for doctors, their patients, and law enforcement officials.”¹⁴⁰ The Vuitch Court considered a D.C. ban that excepted abortions “done as necessary for the preservation of the mother’s life or health.”¹⁴¹ The crucial question was whether the statute allowed abortions to preserve the patient’s mental health or wellbeing.¹⁴² A three-judge plurality¹⁴³ interpreted the word health to “include[] psychological as well as physical well-being,” finding that when “properly construed” in this way, “the District of Columbia abortion law is not unconstitutionally vague.”¹⁴⁴

Nevertheless, courts continued to invalidate abortion statutes on this ground, and litigants did not think Vuitch settled the issue.¹⁴⁵ When briefing Roe v. Wade, appellants still pushed vagueness, arguing that the “confusion spoken of in Vuitch has not subsided.”¹⁴⁶ Roe’s holding, based instead on substantive due process, mooted the vagueness issue.¹⁴⁷

But Roe’s companion case, Doe v. Bolton, considered the question directly.¹⁴⁸ That case concerned Georgia’s abortion ban, which prohibited abortion unless a physician used “his best clinical judgment” to determine “that an abortion was necessary.”¹⁴⁹ The plaintiff, Mary Doe, had sought an abortion on the basis of poverty and mental health, but a hospital review committee denied her request. ¹⁵⁰ The Court relied on Vuitch to find the statute was not unconstitutionally vague when construed broadly enough to allow Doe’s abortion. Indeed, the Court broadened the types of factors relevant to “health” further: “[M]edical judgment may be exercised in the light of all factors—physical, emotional, psychological, familial, and the woman’s age—relevant to the well-being of the patient. All these factors may relate to health. This allows the attending physician the room he needs to make his best medical judgment.”¹⁵¹

Though Vuitch and Doe refused to invalidate bans on vagueness grounds, these cases do not control modern abortion bans—the vast majority of which either lack a health exception or have a narrow health exception that explicitly prohibits the broad definition of health that Vuitch and Doe adopted. Six states lack any health exception,¹⁵² and ten more have health exceptions that specifically exclude things like a “claim or diagnosis that the patient would engage in conduct that would result in the patient’s death,”¹⁵³ “psychological conditions,”¹⁵⁴ a physician’s “belie[f] that the woman may or will take action to harm herself,”¹⁵⁵ and “psychological conditions, emotional conditions, familial conditions, or the woman’s age.”¹⁵⁶ And no abortion ban allows abortion because of concerns about the patient’s family, age, or general wellbeing. Vuitch and Doe avoided the core vagueness issue by interpreting abortion bans in a way that is not possible today: These modern bans now beg all the questions that the Belous court raised. ¹⁵⁷

Indeed, the Roe opinion’s first draft invalidated the Texas abortion ban on vagueness grounds because it only contained a life, not health, exception.¹⁵⁸ Blackmun’s first draft, published in Bernard Schwartz’s book, The Unpublished Opinions of the Burger Court, noted: “The vagueness claim in Vuitch focused only on the word ‘health’ . . . . We are concerned, in contrast, with a procedure that is exempt from criminality only if it is ‘for the purpose of saving the life of the mother.’”¹⁵⁹ The first Doe draft was substantially the same, also distinguishing Roe on this ground.¹⁶⁰ In other words, the Court saw Roe and Doe differently on vagueness because the Texas ban did not allow a broad definition of health, such as the one Doe adopted.

The first draft of Roe asked a series of questions underscoring the vagueness issue, reminiscent of those that we hear routinely today:

Does [saving the life] mean that he may procure an abortion only when, without it, the patient will surely die? Or only when the odds are greater than even that she will die? Or when there is a mere possibility that she will not survive? . . . Further, who is to exercise that judgment—the physician alone in the light of his training and experience, or a group or committee of his peers, or a medical association, or a hospital review committee? And when is the saving of a life to be measured in the time scale? Must death be imminent? Or is it enough if life is prolonged for a year, a month, a few days, overnight? Is a mother’s life “saved” if a post-rape or post-incest or “fourteenth-child” abortion preserves, or tends to preserve, her mental health?¹⁶¹

Blackmun’s first Roe draft concluded that the statute is “insufficiently informative to the physician . . . who must measure its indefinite meaning at the risk of his liberty.”¹⁶²

After Justice Blackmun circulated the draft, Justice Brennan and others pressed him to alter the reasoning—not out of disagreement with the vagueness conclusion, but because Brennan believed there was “no point in delaying longer our confrontation with the core issue on which there appears to be a majority and which would make the reading the vagueness issue unnecessary.”¹⁶³ The vagueness strategy did not cease after Roe and Doe, however. Courts, including the Supreme Court, continued to invalidate abortion regulations for vagueness in ways that have resonance today.¹⁶⁴

With Dobbs overturning the substantive due process right, this alternative justification for Roe remains powerful. Two years after Dobbs, we have clear evidence that doctors—the criminalized parties—do not understand what constitutes an illegal abortion or when conditions trigger life and health exceptions.¹⁶⁵ For instance, one survey demonstrated that 70% of doctors “were unclear about what constitutes a ‘life-threatening emergency’ allowing a physician to legally perform an abortion in states where the procedure is otherwise banned.”¹⁶⁶ We also know this confusion is chilling reproductive healthcare. In another survey, 68% of OB/GYNs in states with abortion bans reported that Dobbs “worsened their ability to manage pregnancy-related emergencies,” and 64% “believe that the Dobbs decision has worsened pregnancy-related mortality.”¹⁶⁷ A different study found that 93% of OB/GYNs in ban states “reported situations in which they or their colleagues could not follow clinical standards due to legal constraints.”¹⁶⁸ The emerging data goes beyond perception. In Missouri, for instance, the number of hospital-based abortions—abortions that are almost always done for medical reasons—dropped 72% (from 102 abortions per year to only 29) in the year after Dobbs.¹⁶⁹

For a variety of reasons, legislative solutions adding or attempting to clarify exclusions and exceptions have failed. First, some of the exclusions, like the miscarriage exclusion, are inherently vague and difficult to apply. Nearly every state with a previability abortion ban uses the language “removal of a dead fetus” to protect miscarriage care.¹⁷⁰ Though it seems clear on its face, fetal death is, in reality, hard to diagnose—particularly early in pregnancy when most miscarriages happen.¹⁷¹ Miscarriage symptoms, like bleeding and cramping, are not diagnostic as they occur in a substantial number of healthy pregnancies.¹⁷² And though ultrasounds and bloodwork can diagnose miscarriage, they can only do so over time. For instance, the Society of Radiologists recommends ultrasounds over ten to fourteen days to diagnose a miscarriage conclusively when the pregnancy is too small to visualize fetal cardiac activity.¹⁷³ Though the American College of Obstetricians and Gynecologists (ACOG) recommends a more patient-centered approach that does not require certainty of fetal death to initiate miscarriage management,¹⁷⁴ patients in states with abortion bans are regularly unable to get miscarriage care if there is any uncertainty about fetal demise, leading to countless dangerous delays in care.¹⁷⁵ As mentioned in the introduction, this miscarriage uncertainty was likely one of the factors that delayed Porsha’s care and ultimately led to her death.¹⁷⁶ Two other post-Dobbs deaths in Texas involve similar facts.¹⁷⁷

The same problem exists with life and health exceptions, the vagueness of which led to many of the horror stories now flooding the media.¹⁷⁸ All pregnancies involve life-threatening risks, but exceptions provide no guidance on when those life-threatening risks satisfy the statute.¹⁷⁹ How high must the risk of death be to trigger the life exception? All pregnancies involve the risk of death,¹⁸⁰ and many pregnant patients have additional health complications that increase those risks.

Health exceptions, when they exist, do not provide much more protection and involve nearly identical line drawing problems. They are intentionally written to be narrow, and they use language that is inherently vague. If the classic health exception requires a “serious risk of substantial and irreversible impairment of a major bodily function,” then there are no less than three words that are open to significant interpretation.¹⁸¹ How high or imminent must the risk be to qualify as “serious”? When does an impairment become “substantial”? Does the loss of a fallopian tube or a uterus count as a “impairment of a major bodily function”? And more broadly, is a physician’s good faith, subjective belief of the emergency sufficient, or will the provider be judged objectively in hindsight?¹⁸² What happens when a state’s many abortion laws conflict or use different standards? This last point has been particularly challenging given that some states, like Texas, have different intent standards in different abortion laws.¹⁸³

But it is not just the health and life exceptions that have proved so complicated. The typical fetal-anomaly-exception’s requirement of lethality or fatality has confounded many as well. If fatal “means 100 percent certainty of death in infancy despite aggressive treatment, then almost no condition meets this standard. But once more wiggle room is introduced to establish fatality,” then the statute will inevitably be too vague to apply or involve arbitrary standards.¹⁸⁴ The one state that more leniently excepts “life-limiting anomal[ies]” provides no guidance on what this phrase means.¹⁸⁵ And as medical advances continue, even the most prototypically “lethal” conditions include some survivors that reach adolescence.¹⁸⁶ Already, stories abound of people unable to obtain abortions even when providers believed the fetal condition was lethal in states that purport to except these situations.¹⁸⁷

Compared to the pre-Roe era, the chill from these vague exclusions and exceptions is greater today for a few reasons. First, the penalties in abortion bans are higher now: Violating Texas’s pre-Roe ban risked two to five years in prison,¹⁸⁸ whereas now violators risk life in prison.¹⁸⁹ Second, medicine has become a hyperregulated, public practice. Before Roe, “[m]ost physicians were in small or solo practices,” making “quiet lawbreaking . . . more plausible.” ¹⁹⁰ Today, however, “most doctors are employed by healthcare systems or work within larger group practices” where other employees can and do report perceived indiscretions, increasing the risk that anything even resembling an illegal abortion will be subject to investigation.¹⁹¹ Third, physicians historically had discretion to use their good faith judgment in assessing medical exceptions; today, however, the law holds them to objective standards.¹⁹² And fourth, most states legally require physicians to report abortions to the state—inviting an investigation into their choices.¹⁹³

All these changes make physicians more risk averse. Yvette Lindgren and Michelle Oberman have called this hesitance “yellow light” medicine—i.e., where “clinicians know their patients need abortion-related care, but hesitate or refuse to provide it because it seems legally risky.”¹⁹⁴ In these situations, “a clinician’s willingness to provide care no longer turns exclusively on the largely national professional standard of care,” but “on a doctor’s personal risk tolerance.”¹⁹⁵ They demonstrate that, because physicians “cannot avoid criminal risks without increasing the risks that they will harm their patients,” abortion bans will inevitably chill reproductive healthcare.¹⁹⁶

At this point, the risks are not theoretical. We know that physician uncertainty is causing the chilling effect that the Belous court feared: “[C]ondemn[ing patients] to death as if the law flatly prohibited all abortions.”¹⁹⁷ ProPublica has already reported on five women who died because of physician reluctance to treat pregnancy complications.¹⁹⁸ Hundreds of other patients have come forward with their harrowing stories of near-death experiences, significant harms, or unnecessary trauma because of bans.¹⁹⁹ These stories likely represent a drop in the bucket of people affected. There is no question, therefore, that bans are chilling the protection of other constitutional rights, like the pregnant person’s right to life.

Abortion bans also raise vagueness concerns beyond how sick a pregnant person must be to trigger an exception. For instance, there are also questions about whether the health exceptions must be met through an objective or subjective standard.²⁰⁰ This claim is also not new. In the 1979 case Colautti v. Franklin, the Supreme Court invalidated as impermissibly vague portions of the Pennsylvania Abortion Control Act, which limited the way abortions could be performed after viability.²⁰¹ The Court found the statute too vague because it failed to clearly establish whether viability must be shown through an objective or subjective standard. ²⁰² The Court found that this lack of clarity, “in conjunction with a statute imposing strict civil and criminal liability for an erroneous determination of viability, could have a profound chilling effect on the willingness of physicians to perform abortions near the point of viability in the manner indicated by their best medical judgment.”²⁰³ Rather, citing Doe, the Court found that a statute can only be constitutional if it clearly allows the doctor discretion.²⁰⁴

Recently, a North Dakota court invalidated the state’s abortion ban on vagueness grounds, mirroring the analysis in Colautti.²⁰⁵ The court found problematic “the dual standard of both subjective and objective elements currently required in the [health] exception.”²⁰⁶ The court noted that “a North Dakota physician may provide an abortion with the subjective intent to prevent death or a serious health risk, yet still be held criminally liable if, after the fact, other physicians deem that the abortion was not necessary or was not a reasonable medical judgment.”²⁰⁷ Ultimately, the court concluded that “[a]s written, [the ban] can have a profound chilling effect on the willingness of physicians to perform abortions, even where the North Dakota Supreme Court has already said there is a fundamental right to do so to preserve a woman’s life or health.”²⁰⁸

In late January 2025, the North Dakota Supreme Court refused to stay the injunction issued by the lower court, finding it likely that the plaintiffs will demonstrate the ban is unconstitutionally vague.²⁰⁹ The court agreed that the mixed subjective and objective standard was problematic but also added a variety of other claims, including that the law fails to adequately define “serious health risk” or “substantial physical impairment.”²¹⁰ The court noted that “the degree of specificity required here [is] very high” because of “[t]he harsh punishment the plaintiffs face if they fail to conform their medical practice” to the law.²¹¹

Colautti also found unconstitutionally vague a standard-of-care provision that required the physician after viability to try to “preserve the life and health of the fetus” by using the “the abortion technique . . . which would provide the best opportunity for the fetus to be aborted alive so long as a different technique would not be necessary in order to preserve the life or health of the mother.”²¹² The Court noted that this provision would require doctors to use abortion procedures that were more risky to the pregnant person, as the safest procedures were lethal for the fetus.²¹³ Moreover, citing Vuitch, the Court noted that the health exception within the provision had not been construed to mean that “all factors relevant to the welfare of the woman may be taken into account by the physician in making his decision.”²¹⁴ “Consequently, it is uncertain whether the statute permits the physician to consider his duty to the patient to be paramount to his duty to the fetus, or whether it requires the physician to make a ‘trade-off’ between the woman’s health and additional percentage points of fetal survival.”²¹⁵ The Court held that “where conflicting duties of this magnitude are involved, the State, at the least, must proceed with greater precision before it may subject a physician to possible criminal sanctions.”²¹⁶

Importantly, six states have life or health exceptions with similar language today, requiring the abortion be performed in a way that gives the fetus the best chance at life.²¹⁷ So, in addition to ambiguity as to when the health exception is triggered, there is also ambiguity as to which abortion method is permissible to complete the medically necessary abortion. Reports persist of people being forced to endure C-sections and labor inductions over safer and less invasive traditional abortion procedures, even when the fetus has no chance at life.²¹⁸

Reflecting these issues, plaintiffs and courts have expressed a renewed interest in vagueness challenges in the post-Dobbs environment.²¹⁹ A court in North Carolina has similarly invalidated for vagueness part of the North Carolina abortion ban that required physicians to document “probable existence of an intrauterine pregnancy.” ²²⁰ New lawsuits in Wyoming and South Carolina allege that those states’ bans are unconstitutionally vague based on the bans’ health and, in South Carolina’s case, fetal anomaly exceptions.²²¹ Plaintiffs have also historically brought challenges based on an impermissibly vague definition of abortion.²²²

There are multiple possible remedies for unconstitutionally vague exceptions: eliminate, clarify, or broaden. However, all these remedies will fail for different reasons. First, eliminating exceptions entirely, though clear, is widely agreed to be an unconstitutional infringement of the pregnant person’s right to life²²³ and would put antiabortion legislators in the position of proposing a law that would shock the conscience of most Americans. ²²⁴ After all, there is a reason every state with an abortion ban has, at a minimum, a life exception.²²⁵

On the other end of the spectrum, legislators could create a broad health exception that allows abortions to be provided on a good faith, subjective belief that the abortion was necessary to protect the abortion seeker’s health—broadly defined and including mental health. This broad exception would ensure compliance with the Vuitch and Doe holdings. But this cure would fail for a political reason—the antiabortion movement would never allow it.²²⁶ Indeed, in the EMTALA litigation, the conservative justices accepted the argument that a mental health exception would effectively dismantle abortion bans.²²⁷ Antiabortion courts and activists have even been unwilling to loosen physical health exceptions because they fear doing so would open the door to abortions for any high-risk pregnancy.²²⁸ In the words of at least one antiabortion activist, these abortions are “quasi-elective” because the medical emergency is not immediate and urgent.²²⁹ It is certainly true that abortion providers who see abortion as a human right and understand the risks pregnancy poses to all patients will use broad health exceptions to help as many patients as they can, so the concern is not invalid.

Between these two extremes, however, it is hard to imagine how legislatures could clarify exceptions in a way that solves the problem²³⁰ or how courts could interpret the statute to avoid the vagueness challenge.²³¹ Recent efforts to clarify exceptions have failed to stop the chill of abortion bans.²³² As just one example, Texas excluded ectopic pregnancy from its definition of abortion years before Dobbs and even added ectopic pregnancy as an affirmative defense for civil and criminal abortion liability in 2023.²³³ Nevertheless, Kelsie Norris-De La Cruz was sent home in 2024 even though doctors strongly suspected she had an ectopic pregnancy.²³⁴ The delay in her care threatened her life and caused her to lose most of her right fallopian tube.²³⁵ Independent physicians who reviewed her case for the Washington Post said that “[e]ctopic pregnancies . . . can be hard to diagnose on an ultrasound with 100 percent certainty . . . and if the diagnosis is wrong, a doctor might fear potential legal repercussions for terminating a viable pregnancy.”²³⁶

But ectopic pregnancy is probably the clearest exception possible. Other exclusions and exceptions—like removal of a dead fetus, health or life of the pregnant person, and lethal fetal anomaly—are innately ambiguous, as described above.²³⁷ Put simply, legislators are trying to regulate pregnancy as if it is black and white, but it is not and never will be. It will never be possible to implement a ban imposing such severe penalties for mistakes that draws clear enough lines to guide doctors because every pregnancy is both health- and life-threatening at baseline and involves unique and constantly changing risks.²³⁸ And as described in Section II.C, many of the lines drawn will also be irrational. As a result, attacking these exceptions under the vagueness theory could lead to broad remedies that far exceed the pregnancy complication cases.

B. Religious Liberty

The Roberts Court’s defining feature has perhaps been its jurisprudence around religious liberty. Almost every litigant asserting a religious exemption from a law that most non-religious people would have to follow has succeeded with this Court.²³⁹ From anti-discrimination laws to COVID restrictions to work schedules, religious plaintiffs have successfully used the First Amendment and related statutes to redraw the lines of otherwise generally applicable laws.²⁴⁰

Though religious liberty is often seen through a conservative, antiabortion lens,²⁴¹ there is a long history of religious arguments for abortion rights. For instance, in the years before Roe, when abortion was mostly illegal throughout the country, the Clergy Consultation Service assisted thousands of women in obtaining abortions by providing funding and transportation to get them to an abortion provider.²⁴² Since Roe, the Religious Community for Reproductive Choice (RCRC) has provided advocacy on behalf of and support for people of diverse faiths who believe in reproductive freedom.²⁴³ RCRC articulates the connection between faith and abortion as rooted in the belief that “decisions to become a parent, when, and under what circumstances are deeply personal. These matters are best left to an individual to discern in consultation with family, faith and others one may bring into the conversation.”²⁴⁴ Contrary to popular perception, some faiths, like Judaism, teach that personhood starts at birth, and some believe that abortion is not only permissible, but at times, obligatory when the pregnant person’s life or health (including mental health) is threatened.²⁴⁵

In fact, liberal religious support for abortion rights nearly derailed the passage of the Religious Freedom Restoration Act (RFRA).²⁴⁶ In 1992, when Congress was debating RFRA, one of the leading antiabortion movement lawyers, James Bopp, testified against the law to argue that it could threaten antiabortion legislation.²⁴⁷ As he saw it, “RFRA is not neutral about abortion; it is a very potent new legal weapon . . . to strike down protective abortion rights laws. It is intended to treat abortion as other religious claims, by subjecting those restrictions on abortion to a compelling interest.”²⁴⁸ Under that standard, he asserted, “very few restrictions on abortion would survive.”²⁴⁹

Despite such opposition from the antiabortion movement, RFRA passed. However, RFRA never became the threat Bopp feared. Soon after its passage, the Supreme Court held that state laws could not be subject to RFRA religious liberty challenges because Congress, in instructing courts on how to analyze such claims, exceeded its powers under the Fourteenth Amendment.²⁵⁰ Thus, state abortion bans are exempt from federal RFRA challenges.²⁵¹

That does not mean, however, that it is impossible to challenge state abortion bans as violations of religious liberty. While RFRA is unavailable to litigants, other vehicles remain. More than half of states have their own version of RFRA,²⁵² and those statutes could support the claims developed here.²⁵³ But more generally, states are bound by the First Amendment guarantee to the free exercise of religion.²⁵⁴ And now, the confluence of the current Supreme Court’s more muscular First Amendment jurisprudence on religious liberty and the post-Dobbs focus on medical exceptions could bring about, albeit under a different legal source, the exact fear that Bopp foresaw in his 1990s testimony.

There are two important shifts that make abortion rights claims rooted in religious liberty more promising now than ever before. First, despite groups like RCRC regularly filing amicus briefs with the Supreme Court and participating in litigation in other ways—prior to Dobbs—these claims were seen as unnecessary because Roe and Casey protected the right to abortion. Second, for decades the Supreme Court followed the rule of Employment Division v. Smith, a 1990 precedent holding that neutral and generally-applicable laws were not constitutionally problematic even if they burdened someone’s religious practice.²⁵⁵ Thus, people who believed an abortion restriction infringed upon their religious beliefs were unable to show that the law was religiously discriminatory under Smith, but could otherwise challenge the law under Roe and Casey.

Now, however, abortion is not a protected federal constitutional right, and the Supreme Court has drastically shifted its religion jurisprudence. Although it has not formally overruled Smith,²⁵⁶ the Court has significantly curbed its application. Perhaps most relevant here, in 2021’s Tandon v. Newsom, the Court ruled that the state of California could not apply its COVID-era prohibition on large gatherings to religious organizations because the state’s law had exceptions for some secular gatherings.²⁵⁷ The Court announced that “government regulations are not neutral and generally applicable, and therefore trigger strict scrutiny under the Free Exercise Clause, whenever they treat any comparable secular activity more favorably than religious exercise.”²⁵⁸

That same year, the Court extended this principle to the City of Philadelphia’s antidiscrimination law as it applied to foster care agencies. In Fulton v. City of Philadelphia, the Court found unconstitutional Philadelphia’s denial of a foster care contract to Catholic Social Services because the agency’s discrimination against same-sex couples violated the city’s antidiscrimination ordinance.²⁵⁹ Part of the Court’s rationale was that Philadelphia had a system of individual exemptions to its antidiscrimination mandate, exemptions that were available for secular reasons but unavailable to the Catholic agency for its religious reasons.²⁶⁰ Accordingly, the Court held that the antidiscrimination law is not “generally applicable” because the city may not refuse exemptions for religious reasons when it allows them for secular ones.²⁶¹ As Elizabeth Sepper explained the Court’s emerging doctrine in this area: “A single secular exception seems to require religious exemption.”²⁶²

In the wake of Dobbs, this line of cases opens the door to religious liberty claims on behalf of abortion rights because abortion bans have secular carveouts, including exceptions based on pregnancy complications, but no religious carveouts. Several cases have been filed on behalf of religious people claiming that abortion bans violated their right, grounded in their religious beliefs, to obtain an abortion.²⁶³ Most of these cases are still being litigated or have met procedural hurdles, but a case out of Indiana shows the impact of medical exceptions on this new line of religious liberty jurisprudence. In that case, five women and the organization Hoosier Jews for Choice challenged the state’s abortion ban, alleging they had “sincere religious beliefs . . . [that] direct them to seek pregnancy terminations” that the state’s law forbids.²⁶⁴ Hoosier Jews for Choice’s religious liberty rationale is rooted in a broad understanding of health exceptions. One of the individual plaintiffs in the case explained that her understanding of Jewish law was that “the life of a pregnant woman—including her physical and mental health and wellbeing—‘must take precedence over the potential for life embodied in a fetus’” and that, in accordance with her faith, “if her health or wellbeing—physical, mental, or emotional—were endangered by a pregnancy, pregnancy-related condition, or fetal abnormality, she must terminate the pregnancy.”²⁶⁵

The Indiana Court of Appeals affirmed the lower court’s preliminary injunction of the state’s abortion ban based on the ban’s secular exceptions:

The State has not shown that its claimed compelling interest in protecting the potential for life is satisfied by denying Plaintiffs’ religious-based exception that prioritizes a mother’s health over potential life, given that other exceptions are allowed based on the same prioritization—that is, the exceptions applicable when the pregnancy poses a “serious health risk” or termination would “save the pregnant woman’s life.”²⁶⁶

This win for religious liberty and abortion rights was based on secular exceptions, including the statute’s health and life exceptions; however, the victory was narrow. This preliminary injunction applies only to the plaintiffs and only for abortions that plaintiffs seek because of their sincere religious beliefs.²⁶⁷

A similar claim could have a broader impact if brought on behalf of abortion providers rather than abortion seekers. The claim in such a case would be that some abortion providers believe they have a religious obligation to provide medical care for people who come before them; therefore, if a patient comes to them seeking an abortion—for whatever reason—and the provider has the training and expertise to provide that care, the provider is required by their religion to provide it.²⁶⁸ The religious provider’s free exercise rights are deprived, like those of the houses of worship in Tandon or the Catholic adoption agency in Fulton, because the state law recognizes secular exceptions—including for medical reasons—but not the provider’s religious exceptions. Although this type of case has long been theorized by scholars,²⁶⁹ such a case had not been brought to the courts in the first two years after Dobbs.

However, in January 2025, five physicians in South Carolina brought this exact claim.²⁷⁰ The federal complaint alleges that these doctors “have long held sincere religious and conscientious beliefs commanding them to respect the inherent worth of every person, help people in critical need, and place others before themselves” which includes “using their medical training to honor a patient’s request to end her pregnancy when it threatens to profoundly harm her.”²⁷¹ The crux of the religious liberty claim in this case is based on a comparison to secular medical exceptions: “It is neither religiously neutral nor generally applicable for South Carolina to allow abortion under the Abortion Ban’s secular Exceptions while criminalizing abortion when Plaintiffs’ religious beliefs compel it in substantially similar circumstances.”²⁷² Rather, “South Carolina has criminalized religious conduct while allowing secular conduct that undermines its purported state interest in similar ways.”²⁷³

This litigation is in its early stages, so we do not know how it will be resolved.²⁷⁴ It is certainly possible that courts will not view secular exceptions based on life or health as equivalent to the claimed religious exceptions,²⁷⁵ or that courts will accept there is differential treatment but conclude the state can do so because the law nonetheless meets the high burden of strict scrutiny.²⁷⁶ But we do know one important thing: If the plaintiffs here obtain relief, it will be much broader than the relief obtained so far in the Indiana case filed on behalf of women seeking abortions. Here, success would result in these religious doctors being able to provide abortions consistent with their religious beliefs.²⁷⁷ If the religious belief is general—about supporting the health and welfare of all abortion patients, not limited only to extreme cases—then all patients who come before these doctors would fit within their religious provision of care.

Thus, in light of the Supreme Court’s newly developing jurisprudence on religious exceptions, another aspect of the broad power of medical exceptions becomes clear. Because there are secular exceptions for extreme pregnancy complications, there also must be religious exceptions for patients and providers motivated by their beliefs, exceptions that would benefit all patients who seek care from religious providers.

C. Rational Basis Review

Beyond vagueness and religious liberty, abortion bans are also susceptible to rational basis challenges based on pregnancy complication cases. In overruling Roe and Casey, Dobbs concluded that abortion is not a fundamental right and that the undue burden test is no longer the standard to review abortion restrictions and bans.²⁷⁸ That conclusion, however, does not mean that abortion laws are free from constitutional constraints. Rather, as Dobbs explained, abortion bans must be rationally related to a legitimate government interest.²⁷⁹ Under rational basis review, there is a “strong presumption of validity,” and courts are not supposed to substitute their beliefs for the judgment of legislatures.²⁸⁰

Importantly, although Dobbs was litigated as a case under the Due Process Clause, it concludes that both the due process and equal protection rational basis tests apply to abortion laws. The due process aspect of the Court’s conclusion is most obvious, as the case concerned (and rejected) abortion as a protected right under the substantive due process doctrine.²⁸¹ However, the Court also held that the equal protection rational basis test applies to abortion bans and restrictions.²⁸² Thus, Dobbs clearly adopts both rational basis tests as the applicable standards for reviewing abortion laws. The two tests employ the same language but differently frame the question reviewing courts must ask. Due process considers whether the restriction on liberty from an abortion ban is rationally related to a legitimate government interest; equal protection asks the same question about differential treatment within an abortion ban.²⁸³

Under both the due process and equal protection frameworks, conventional wisdom is that the rational basis test is so forgiving that the government almost always wins. Empirical study bears this out, and it is easy to see why when you examine the doctrine.²⁸⁴ Under equal protection rational basis, blackletter law says that almost anything is a legitimate purpose,²⁸⁵ the challenger has the burden of disproving all possible legitimate purposes,²⁸⁶ and vast over- and under-inclusivity is permitted.²⁸⁷ Under due process rational basis, blackletter law says that there is a heavy presumption that the government interest is rational,²⁸⁸ and the means used to accomplish that interest need not be logically sound.²⁸⁹ With such forgiving doctrine, it is no surprise that challengers of government action lose quite frequently.

However, challengers do not always lose under rational basis review, and that is an important lesson for abortion claims post-Dobbs. In multiple articles, Katie Eyer has attempted to resuscitate the possibilities of rational basis review and persuade people that it is a legitimate theory for challenging government action.²⁹⁰ Her work thoroughly catalogs a wide variety of cases in which challengers have prevailed in rational basis challenges. For instance, in the years immediately preceding the Court’s first announcement of the intermediate scrutiny standard for sex discrimination claims, multiple Supreme Court cases succeeded under the rational basis test.²⁹¹ Similarly, gay rights cases have also succeeded on rational basis challenges at the Supreme Court despite the Court having never announced any heightened scrutiny based on sexual orientation.²⁹²

A similar story can be told in the lower courts, both federal and state. As Eyer shows, many of the initial wins for gay couples challenging marriage bans succeeded on rational basis challenges.²⁹³ Plaintiffs challenging criminal justice policies that disparately impact people of color—claims that have faced difficult hurdles at the Supreme Court²⁹⁴—have also found some success under rational basis review.²⁹⁵ Advocates also had early success in the lower courts challenging laws that discriminated against pregnant women under rational basis review.²⁹⁶ Likewise, from the opposite side of the political spectrum, challenges to state and local occupational licensing requirements have been successful before some lower courts based on arguments that varying treatment under these codes fails rational basis review.²⁹⁷ The success of these cases defies conventional wisdom about rational basis review and is an important lesson that cannot be ignored: “[R]ational basis review—as deployed in the lower and state courts,” Eyer summarizes, “has often afforded one of the most plausible openings for social movements to create space for constitutional change.”²⁹⁸

Drawing on these cases, abortion rights litigation post-Dobbs should not preemptively concede the rationality of abortion bans. The Mississippi ban at issue in Dobbs contained life, health, and fetal anomaly exceptions, which were critical to the Court’s holding.²⁹⁹ Thus, the Court has not reached the question of whether abortion bans that include medically indicated abortions are constitutional under rational basis review.

Pregnancy complications are a great vehicle for developing this theory. The goal of this Section is not to provide every possible argument that can be made against abortion bans under the rational basis test nor to sketch out the detailed case theory of each claim described below. Rather, the goal here is to outline some of the plausible rational basis claims using pregnancy complication cases, and hopefully to encourage future development of this theory, bucking the conventional wisdom that the government always wins under rational basis review.

The theories explored here may achieve different remedies: Some could strike down a ban in its entirety, while others would expand exceptions.³⁰⁰ However, even in the latter scenario, the history of rational basis review litigation has shown that narrow remedies can lead to broader gains in the political or legislative arena.³⁰¹ Further, legislators trying to craft more “rational” policies than those struck down will face new lawsuits because there is simply no rational way to make distinctions in the inherently fluid and nuanced arena of pregnancy. Eventually, the entire ban may be found unconstitutional.

Starting with due process rational basis claims, scholars like Jill Wieber Lens and Meghan Boone have argued generally that abortion bans fail due process rational basis because the state has not shown that they are advancing a legitimate interest.³⁰² Other academics have argued more specifically that abortion bans fail the rational basis test because of their treatment of medical exceptions. For instance, William Aceves has argued that requiring people to be close to death before providing life-saving care is tantamount to torture, something that should fail the rational basis test.³⁰³ And Robyn Powell argues that abortion bans have a disproportionate impact on people with disabilities and that “[f]orcing them to forego medically necessary abortions until reaching the precipice of a life-threatening emergency lacks any legitimate rationale.”³⁰⁴

Other due process irrationality arguments based on medical exceptions are also possible. Such arguments could relate to the narrowness of life and health exceptions—that the lines they draw, even if not unconstitutionally vague, are irrational. Every pregnancy, even ones that appear healthy for long periods of time, creates a heightened risk of death and bodily injury. The United States’s maternal mortality rate is many times higher than our peer countries’ and startlingly high for Black women in particular.³⁰⁵ An abortion is fourteen times less deadly than childbirth here.³⁰⁶ Beyond that, pregnancy is disabling, at least in the short term, causing all sorts of risks and side-effects that abortion typically cures.³⁰⁷ Abortion is therefore always health- (and sometimes life-) saving. Historical abortion bans had exceptions that gave wide discretion to physicians.³⁰⁸ Reva Siegel and Mary Ziegler have shown that this discretion generally included the ability to provide life-saving abortions if a pregnancy resulted in conditions that would not be considered life-threatening today, like excessive vomiting or mental unfitness.³⁰⁹ Unlike the bans from prior generations, today’s bans force only some people to endure life and health threats. This line-drawing is inherently arbitrary and irrational; thus, it should fail the due process rational basis test.

More narrowly, patients carrying fetuses with life-threatening fetal anomalies could bring challenges in states that lack or have an irrationally narrow fetal anomaly exception. In 2022 a Wyoming court reasoned that an abortion ban with no exception for lethal fetal anomalies might violate the due process rational basis test. The court’s rationale focused on how fetal anomalies undermine the purported state interest: “When the potential life is found to have a diagnosable genetic defect that is incompatible with life, the Court could find that [the law] is beyond a reasonable doubt, not related to a legitimate government interest.”³¹⁰ What rational basis could there be to force pregnant patients to endure the risks and burdens of pregnancy only to birth a child that will die in their arms?

There could also be traditional equal protection rational basis claims based on differential treatment within the health exceptions. We have already seen such a ruling from a state court in the post-Dobbs landscape. Last year, a Georgia district court struck down the state’s ban on abortion at six weeks in part because the health exception excluded mental health. Plaintiffs used equal protection to challenge the ban as “treating pregnant women with mental health emergencies differently than women with physical health emergencies.”³¹¹ The court sided with the plaintiffs: “[T]here is no basis—rational, compelling, or sensical—to distinguish between diagnosed medical emergencies involving the brain . . . versus the heart or the lungs or the liver—all of which can result in serious bodily harm or death to the mother.”³¹² The court found it “patently unconstitutional” to “fate [a pregnant person] to death” simply because her life- or health-threatening condition is mental and not physical.³¹³ In early 2025, the Georgia Supreme Court vacated this ruling on other grounds, remanding the case for reconsideration of standing.³¹⁴

Abortion ban exceptions that exclude mental health are particularly irrational; evidence shows that mental health crises “account for more deaths during pregnancy or the year following it than any other health condition.”³¹⁵ Excluding mental health also endorses the “debunked theory of substance dualism, which treats the mental and physical as biologically and functionally distinct.”³¹⁶ As noted above, abortion case law around the time of Roe refused to make such distinctions between physical and mental health.³¹⁷ That fifty years later exceptions now exclude mental health, after decades of research underscoring its importance, is both ironic and irrational.

One could also use fetal anomaly cases to bring an equal protection challenge based on differential treatment in parental rights after birth compared to before birth. After birth, parents of critically-ill infants can refuse life-sustaining treatment and allow their infant to die, but abortion bans that lack exceptions for severe fetal anomaly prohibit the exercise of these rights before birth.³¹⁸ If anything, that a pregnant person’s body suffers to sustain a child who, once born, could be allowed to die should make it harder for the state to justify denying parents their right to make compassionate end-of-life choices.³¹⁹

Also under equal protection, a successful claim might be brought that denial of abortions for those with pregnancy complications is unconstitutional because it is based on animus. Under the Equal Protection Clause’s rational basis test, the Supreme Court has decided four cases striking down laws that might otherwise have been found to have a rational basis because they were based on hatred of a disfavored group of people.³²⁰ To establish animus, the impact of the law must be so extreme as to show that the true motivation was nothing other than a “bare . . . desire to harm a politically unpopular group.”³²¹ While these cases have been difficult to replicate because of the somewhat confusing doctrinal explanation for their holdings,³²² they remain good law and ripe for trying to wield against abortion bans.

Here, the claim would be that abortion bans are motivated by animus against a disfavored group, people who choose to terminate a pregnancy, and that the serious harm of pregnancy complications demonstrates that animus.³²³ In pregnancy complication cases, abortion bans cause significant pain, disability, and even potentially, death.³²⁴ These effects are surely as severe—if not more so—than those the Court has previously found sufficient under this line of cases, such as denied housing or food.³²⁵ That people are free to seek medically necessary, even lifesaving, healthcare in nearly all other contexts but abortion suggests that animus may be motivating the bans. Moreover, abortion seekers are certainly a disfavored group who face unique issues around stigma and shame.³²⁶ The “legitimate” rationales listed by the Dobbs majority³²⁷ cannot justify the harms imposed by abortion bans for people in medical crisis. Rather, the “practical effect of the law” is “to impose . . . a separate status, and so a stigma upon all who” seek health- or life-saving healthcare that would otherwise be legal were it not abortion.³²⁸

Finally, another less-known and less-developed theory of equal protection rational basis review also might have some purchase in pregnancy complication cases: the class of one.³²⁹ In Village of Willowbrook v. Olech, a property owner sued her local government for singling her out for special treatment after she sued the town.³³⁰ Even though the plaintiff did not allege discrimination based on membership in any class, the Court adopted a “class of one” theory of equal protection.³³¹ The Court held that plaintiffs can succeed under rational basis review if they allege that they have “been intentionally treated differently from others similarly situated and that there is no rational basis for the difference in treatment.”³³²

Even though this class of one equal protection theory has been met with criticism from academics,³³³ the Kate Cox case shows how it could be harnessed. There, after the trial court enjoined the state of Texas from enforcing the state’s abortion laws in Cox’s case, the Texas Attorney General sent a letter to three Texas hospitals threatening sanctions if anyone assisted in Cox’s abortion.³³⁴ This extraordinary step to stop medical care professionals from complying with a lawful court order (even if eventually overturned on appeal) would be powerful evidence that the state was treating Cox in an arbitrary manner with no rational basis, given that her fetus had a condition inconsistent with meaningful life. A court could logically deduce that the state was punishing this one person for challenging its authority.

Arguing that pregnancy complications demonstrate the unconstitutionality of abortion bans will no doubt be a heavy lift under the canonical understanding of rational basis review. However, as explained here, rational basis is a more complex doctrine than conventional wisdom suggests. Courts might find that abortion bans are irrational as a matter of due process or equal protection because of the complexity and risks pregnancy complication cases present; this stance could threaten abortion bans generally.

III. Challenging Dobbs as Unworkable

So far, we have described theories for attacking abortion bans post-Dobbs. But pregnancy complication cases can threaten Dobbs itself by proving the case unworkable. Though there is virtually no chance the current Supreme Court would agree with this argument, it must be an essential part of a long-term legal strategy for restoring reproductive freedom—a strategy that abortion rights advocates should invest in now.³³⁵

Workability is one of many factors the Supreme Court uses to determine if overturning precedent is appropriate.³³⁶ Workability evaluates whether a precedent is susceptible to reasonable and consistent application by lower court judges and others.³³⁷ Along with several other factors—the age of a decision, whether people have relied on a ruling, the quality of the precedent’s reasoning, and any changed factual or legal bases since the original ruling—workability is an essential part of the Court’s modern stare decisis jurisprudence.³³⁸

Along with Rachel Rebouché, we have argued that Dobbs should be overruled,³³⁹ as that is the only way to restore abortion rights short of national legislation upheld by the courts.³⁴⁰ The pregnancy complication cases discussed in this Article could form a key part of this strategy, as they can be leveraged to ultimately show that Dobbs is unworkable and should be overturned.

A. The Evolution of the Unworkability Doctrine

The Supreme Court has developed the concept of unworkability in several cases discussing stare decisis. Perhaps most relevant, the Dobbs majority discussed unworkability in depth. After arguing that both Roe and Casey utilized erroneous reasoning and poor-quality analysis, the Dobbs majority turned to workability. The Court explained that workability is “whether [the precedent’s rule] can be understood and applied in a consistent and predictable manner.”³⁴¹ Thus, the key to Dobbs’s long-term sustainability is how others, particularly other courts, think about and use the legal rule developed by the precedent. Notably, Dobbs did not analyze Roe’s strict scrutiny test or trimester framework, but instead analyzed the modified rule developed in Casey. ³⁴²

Dobbs found much unworkable in the Casey framework. First, Dobbs argued that many vague terms in Casey’s rule make it “inherently standardless.”³⁴³ Casey found that an abortion restriction was unconstitutional if it constituted an “undue burden” on a woman’s right to choose to terminate a pregnancy; an “undue burden” existed if the law placed a “substantial obstacle” in the path of someone choosing to obtain an abortion.³⁴⁴ Both these terms, according to the Dobbs majority, are too ambiguous. The difference between a “due” versus “undue” burden depends on judges’ own perceptions, and whether an obstacle is “substantial” involves “a wide gray area.”³⁴⁵ Casey discussed further that “unnecessary health regulations” could be unconstitutional.³⁴⁶ To the Dobbs majority, this language is yet a third unclear term because “necessary” can have various meanings, ranging from strict to less so, and “Casey did not explain the sense in which the term is used in this rule.”³⁴⁷ Finally, Dobbs faulted Casey’s “large fraction” test, which was used to determine how many women must find an abortion regulation a substantial obstacle before it is deemed unconstitutional.³⁴⁸ According to Dobbs, “there is obviously no clear line between a fraction that is ‘large’ and one that is not.”³⁴⁹

Beyond the undue-burden standard’s imprecise terms, Dobbs found fault with how that rule had been applied, both in Casey itself and by lower courts subsequently. The disagreement among the various factions in Casey evidenced, according to the Dobbs majority, the rule’s unworkability: While the plurality used the undue burden test to uphold many of Pennsylvania’s abortion restrictions, Justice Stevens’s separate opinion used the test to reason to the restrictions’ unconstitutionality.³⁵⁰ Dobbs also used justices’ disagreements about how to apply the Casey test in subsequent cases as evidence of Casey’s unworkability.³⁵¹

According to the Dobbs majority, lower courts also struggled to apply the Casey test, further indicating its unworkability. The decision highlighted several circuit splits on the application of the test: whether Casey amounts to a balancing test; how to apply the large fraction test; and the constitutionality of parental involvement laws, bans on particular procedures later during pregnancy, waiting periods, and abortion reason bans.³⁵² As a result, Dobbs concluded that the Casey standard is unworkable because it “perpetuate[s] give-it-a-try litigation before judges assigned an unwieldy and inappropriate task” while failing to advance “the evenhanded, predictable, and consistent development of legal principles.”³⁵³

Dobbs’s workability analysis drew on several other Supreme Court cases about stare decisis. In 2018’s Janus v. American Federation of State, County, and Municipal Employees, Council, the Court overturned the forty-one-year-old decision in Abood v. Detroit Board of Education.³⁵⁴ Abood held that a public employee union could require nonmembers who work for the employer to pay fees germane to the union’s collective-bargaining work but not to its other work, distinguishing between chargeable versus nonchargeable expenditures.³⁵⁵ Abood itself recognized that this distinction will present “difficult problems in drawing lines,”³⁵⁶ which subsequent developments bore out.³⁵⁷ Justice Alito’s Janus opinion found almost the same fault with Abood as he found with Casey’s test in Dobbs: malleable language and inconsistent application.³⁵⁸

But Justice Alito’s opinion in Janus also added another factor to the workability consideration: Whether individuals who have to act in accordance with the Court’s rule can do so in a non-burdensome and principled way.³⁵⁹ The Janus Court found Abood’s rule problematic because it would be hard for decisionmakers to apply it “without launching a legal challenge and retaining the services of attorneys and accountants[.] Indeed, even with such services, it would be a laborious and difficult task to check these figures.”³⁶⁰ A similar consideration appeared in South Dakota v. Wayfair, which reconsidered precedent that had required out-of-state businesses to have a physical presence in a state before that state could collect sales taxes from the business.³⁶¹ The Court concluded that states had not been able “to apply the physical presence rule to online retail sales,” finding it unworkable. ³⁶² In other words, Abood and Wayfair suggest that workability not only concerns whether courts can workably apply a rule, but also whether other actors can reliably and consistently interact with it.

The Supreme Court’s various conceptions of unworkability came in part from a concerted effort by antiabortion movement lawyers who understood the need to weaponize the concept as part of their long-term strategy to overturn Roe v. Wade. Mary Ziegler has detailed this strategy in her historical account of the workability factor.³⁶³ As her research shows, in the wake of Roe, antiabortion movement lawyers sought to broaden the definition of “unworkability” to pave the way for the Court to overturn Roe.³⁶⁴ She explains that “[t]o demonstrate the unworkability of Roe, pro-lifers would seek out [court] wins but settle for conflicting outcomes that would help to build the case against Roe.”³⁶⁵ In other words, creating a sense of chaos and confusion in abortion doctrine was a clear strategy, not an accidental byproduct, of decades of cases brought before a Court with changing personnel.³⁶⁶

Antiabortion lawyers experimented with other conceptions of unworkability beyond just inconsistent judicial application. As Ziegler catalogs, antiabortion lawyers pushed arguments that unworkability meant “unduly broad, unpopular, or likely to create undesirable consequences”; “the political divisiveness of an opinion”; “problems with the substance of an opinion”; the inability of “states to ban or restrict abortion”; difficulty for legislators to follow the decision; conflicts with public health data and scientific knowledge; “a mismatch between Roe’s rationale and the reality of abortion in American clinics”; and “when the meaning of a rule . . . was unclear or seemed to have changed.”³⁶⁷ She concluded, “[b]y giving the Court so many competing definitions of unworkability, antiabortion briefs hoped to convince the Court that Roe was not worth saving.”³⁶⁸

While the Supreme Court has not adopted all the approaches argued by antiabortion lawyers over the decades, this legal strategy has created a mix of considerations the Court analyzes to determine if a past rule is unworkable. Ziegler found that “[m]ost often” the Court looks to whether lower courts have interpreted a rule from the Court in inconsistent ways.³⁶⁹ However, it also looks to whether the previous rule is too vague and open-ended, whether subsequent rulings have created doctrinal uncertainty or incoherence, and whether there are real world effects that indicate the previous ruling produces too much chaos.³⁷⁰ This breadth of possible arguments is a direct result of antiabortion lawyers’ longstanding push to overturn Roe by expanding the understanding of workability at the Supreme Court. Now is the time for abortion rights litigators to take the same approach by developing workability arguments to overturn Dobbs.

B. The Rule of Dobbs

Pregnancy complication cases can help reveal that Dobbs’s rational basis test, when applied in the context of pregnancy, is vague and standardless, one of the key workability inquiries from Dobbs itself. From Dobbs, we learned Casey’s terms “undue burden,” “substantial obstacle,” and “unnecessary health regulations” were vague and standardless when applied to abortion regulations and bans.³⁷¹ The rational basis test from Dobbs, however, uses equally vague terms, requiring courts to determine whether the government is furthering “legitimate” state interests and whether the law has a “rational basis” for doing so.³⁷² A significant scholarly body has demonstrated inconsistent application in rational basis review, arguing that the standard lacks clarity.³⁷³ There is no doubt that these terms have a long pedigree in constitutional law,³⁷⁴ and that in other contexts these terms function without any workability concerns—just as the terms found unworkable in Dobbs permeate other aspects of the law and are not unworkable there.³⁷⁵ Rather, we suggest that, as applied in this particular context of abortion and rapidly developing pregnancy conditions, the rational basis test will, for many reasons, prove just as unworkable, if not more so, than the undue burden standard.³⁷⁶

The first reason the rational basis standard will prove unworkable as applied to pregnancy complications is that judges will inevitably reach varying results under the rational basis test given the enormously high stakes of pregnancy complication cases. Though abortion is no longer a fundamental right post-Dobbs, it nevertheless intersects with other rights, some of which are fundamental—particularly in the context of pregnancy complications—like the right to bodily autonomy, the right to parental autonomy, and the pregnant person’s right to life.³⁷⁷ Scholars have noted that “[r]ational basis with bite . . . appears to be strongly correlated with laws that burden what might be called a ‘significant right,’”—i.e., those interests that are “‘quasi-fundamental’ but [are] not a recognized fundamental right.”³⁷⁸ By contrast, rational basis review is traditionally used in contexts that are low stakes, such as selling alternative milk or fitting eyeglasses.³⁷⁹ There, if the legislature gets it wrong, the worst that happens is that people do not have access to things like an alternative milk or eyeglasses fit by someone with less training. Even in more high stakes areas where the Court has used the standard rational basis analysis, such as in evaluating bans on physician-assisted suicide, the result did not ban doctors from caring for a patient’s underlying medical condition. Rather, the Court’s decision to allow states to prohibit medical aid in dying still ensured doctors could treat a patient’s pain and acute medical condition.³⁸⁰ Pregnancy, on the other hand, risks major bodily harm and even death, and abortion bans infringe other significant rights.³⁸¹

In this context, the rational basis standard’s vagueness—combined with the life-and-death stakes of the outcome—will make it challenging to apply. The standard gives more than enough leeway for courts to reach different conclusions about whether a state’s nearly complete abortion ban is rational when considering a medical condition that is as complex, volatile, and body- and life-threatening as pregnancy.³⁸² The three justices in the Dobbs dissent noted this workability problem with the rational basis test in the context of pregnancy complications. They wrote:

Anyone concerned about workability should consider the majority’s substitute standard. . . . This Court will surely face critical questions about how that test applies. Must a state law allow abortions when necessary to protect a woman’s life and health? And if so, exactly when? How much risk to a woman’s life can a State force her to incur, before the Fourteenth Amendment’s protection of life kicks in? Suppose a patient with pulmonary hypertension has a 30-to-50 percent risk of dying with ongoing pregnancy; is that enough? And short of death, how much illness or injury can the State require her to accept, consistent with the Amendment’s protection of liberty and equality? Further, the Court may face questions about the application of abortion regulations to medical care most people view as quite different from abortion. What about the morning-after pill? IUDs? In vitro fertilization? And how about the use of dilation and evacuation or medication for miscarriage management?³⁸³

As these justices identified, any attempt to clarify what the “rational basis” for the law is in this particular context becomes vague and unworkable.

Second, pregnancy complications often involve urgent circumstances, where decisions must be rushed. Take Kate Cox’s case as an example.³⁸⁴ Her case required a state court judge (and subsequent appellate judges) to determine—with extremely limited time remaining in her pregnancy—whether her diagnosis qualified under the state’s abortion ban exception. It is hard enough to ask medical professionals to reach such a quick definitive conclusion; asking judges with no medical training to make this weighty decision in the midst of a rapidly devolving pregnancy is almost by definition unworkable. Had Cox brought a federal constitutional claim,³⁸⁵ the judge would have had to assess the vague rational basis factors discussed by the Dobbs dissent in a matter of days. Put simply, judges are not equipped to accomplish the task of assessing emergency pregnancy complications that Dobbs now demands.³⁸⁶

Third, given the strongly held views about abortion—which range from abortion is murder³⁸⁷ to abortion is a human right³⁸⁸—litigants will be motivated to press strategic litigation in both directions, and courts, judges, and juries will be motivated to exploit ambiguity within the rational basis test to reach the result they prefer. This dynamic will lead to inconsistent results—one of the hallmarks of unworkability—particularly in the context of pregnancy complications. As noted, antiabortion activists worked to create this inconsistent application of Casey and then used that to show the standard’s unworkability.³⁸⁹ If the abortion rights movement takes a similar approach, it could have a similar result. Very little time has passed since Dobbs, but we have already begun to see an emerging sense of inconsistency with respect to the rational basis test in the decision’s wake.

Most cases applying the rational basis test in the abortion context have found state bans constitutional,³⁹⁰ but not all. As discussed earlier, courts in Georgia and Wyoming have already found abortion bans in the context of pregnancy complications unconstitutional under rational basis tests.³⁹¹ More cases like these two will surely follow. Without doubt, many judges will reflexively apply the deferential rational basis test and allow all abortion restrictions to stand. However, other judges and courts will follow the path set by the courts in Georgia and Wyoming. Varying lower court decisions applying the rational basis test might help push a future Court to determine that the Dobbs test is unworkable.³⁹²

C. The Consequences of Dobbs

Beyond Dobbs’s rational basis test, the consequences of the case should contribute to the conclusion that its holding is unworkable. Aside from courts, those doing the work to implement Dobbs are also revealing its difficulty in application. In Abood and Wayfair, the Court considered whether entities other than courts could implement the precedent consistently.³⁹³ Since Dobbs returned abortion back to the states, physicians have been unable to assess the competing legal interests written into the law by state legislators, as detailed in Section II.A. Driven by individual or hospital physician risk aversion, this confusion has meant that patient care is now perilously inconsistent even within one state. Legislatures around the country have also struggled to implement the Dobbs ruling.³⁹⁴ Most antiabortion states have altered their definitions of abortion to try to exclude more types of care in the wake of Dobbs, but the tweaks have rarely helped.³⁹⁵ Even the courts have proved unable to understand and interpret their own states’ bans.³⁹⁶

Widening the lens even further, Dobbs is quickly creating (and will continue to cause) an unstable and incoherent legal landscape. As we predicted before Dobbs with our co-author Rachel Rebouché,³⁹⁷ Dobbs has raised complicated questions going to the heart of our constitutional republic. These include questions about how federal and state law intersect in this space, particularly in the context of medical exceptions; how state conflicts should be managed; and how abortion interacts with other constitutional rights. It is only because Dobbs eliminated the national floor for abortion, without offering any guardrails beyond rational basis review, that courts will have to grapple with the dizzying number of legal questions related to abortion.

To begin, in the two years since Dobbs, courts have already split over how EMTALA interacts with state abortion bans. Not only did lower courts reach different conclusions, but when the Supreme Court tried to decide the case, there was no majority opinion and separate opinions battled over the medicine of pregnancy complications.³⁹⁸ Justice Jackson’s separate opinion catalogued a host of conditions that she believed would require abortion under EMTALA even if banned by the state, while Justice Alito’s opinion concluded the opposite.³⁹⁹ Similar debates about complex medical conditions have appeared in lower court opinions struggling with the EMTALA conflict.⁴⁰⁰ This kind of practical difficulty for various courts trying to work through pregnancy complication cases in the wake of a Supreme Court decision is strong evidence of unworkability.⁴⁰¹

We also have already seen splits emerging on the vagueness challenges discussed in Section II.A. For instance, North Dakota’s abortion ban is currently enjoined on vagueness grounds related to the health exception, while the Idaho Supreme Court found that its similar abortion ban was not unconstitutionally vague (after construing the statute to permit good faith medical judgment and no imminent threat to life).⁴⁰² The current situation mirrors the pre-Roe landscape, during which there was also a large circuit split on vagueness. Appellant’s opening brief in Roe even included a chart that compared the language from different opinions on vagueness, showing that courts had interpreted the same language in widely different ways.⁴⁰³ As noted, those cases were never fully resolved; rather, Roe mooted them. The religion cases discussed in Section II.B are also starting to develop into a split.⁴⁰⁴ Other courts have dismissed these cases for lack of standing.⁴⁰⁵ As these challenges continue in the post-Dobbs environment, we can certainly expect inconsistent results.

Other questions related to pregnancy complications—and many more related to abortion generally—will also arise, creating further splits. For instance, can a state with an abortion ban prevent medical professionals from providing abortion counseling or out-of-state abortion referrals to patients facing an emergency medical complication during their pregnancy?⁴⁰⁶ Does the Comstock Act cover medically necessary abortions despite lacking any exceptions? Can a state government criminalize helping minors leave the state for an abortion due to mental health, or would that violate the right to travel? Does it violate free speech for a state to ban websites that give people information on how to obtain abortion pills? Can a state ban a drug the FDA has approved?

On all these issues, lower courts are inevitably going to diverge, reaching different conclusions on the same issues.⁴⁰⁷ And when these issues make it to the Supreme Court, there are going to be similar and predictable splits among the justices and challenges in reaching coherent majorities. In Dobbs, the negative and chaotic impact of Roe and Casey on other legal doctrines influenced the Court’s consideration of overturning those precedents.⁴⁰⁸ Going forward, the same will be true with respect to Dobbs. Almost surely, the dizzying result will make the application of the rules from Roe and Casey seem quite easy in comparison.

We recognize, as we must, that these arguments about workability mostly rely on future developments that may not come to pass. Of course, the current justices who decided Dobbs will not agree that the doctrine they announced in that case is unworkable.⁴⁰⁹ As with anyone writing for the future, we sketch what we feel is legitimate legal analysis of current doctrine. But it is also part blueprint for pushing the law in a particular direction in the future. Antiabortion movement lawyers, with an important assist from conservative judges in both the Supreme Court and lower federal courts, relentlessly pursued this tactic for decades in their efforts to overturn Roe.⁴¹⁰ It is time for abortion rights and justice lawyers to use these pregnancy complication cases to do the same. It is also time for receptive judges up and down the federal and state judiciaries to follow their lead by writing opinions creating and highlighting these conflicts.

Conclusion

Though pregnancy complication cases involve a small subset of abortions, this Article argues that they have the potential to liberate abortion for everyone. Through the doctrines of vagueness, religious liberty, and rational basis review, the abortion rights movement can begin a long-term strategy that harnesses the moral demands surrounding pregnancy complications to restore reproductive freedom, ultimately proving that Dobbs itself was unworkable and should be overruled. Moreover, even if these theories fail, storytelling and litigation around medical emergencies has the power to demonstrate how physical and mental health interact and how all abortions are health- and lifesaving. They are therefore an important part of changing hearts and minds, which may also lead to change at state houses and the ballot box.

Though the current Supreme Court is unlikely to hear these stories and cases sympathetically, the long-term battle to overturn Dobbs must be fought now, and medical exceptions provide an opportune vehicle to do so. Advocates in other countries have found success on this path. Ireland is a good example. In 1983, Ireland voted to add a constitutional amendment to protect fetal life, constitutionally enshrining a nationwide abortion ban.⁴¹¹ The original text of the amendment was edited during its debate to recognize the “equal right to life of the mother” in response to the highly publicized case of Sheila Hodgers.⁴¹² Sheila was diagnosed with breast cancer while pregnant but unable to obtain both abortion and breast cancer treatment due to potential fetal harm. She died two days after giving birth.⁴¹³ Her death was a tragedy; however, as Julie Suk has shown, her story and the “life of the mother” right it generated, “created an opening for a path to abortion legalization in Ireland, including the eventual repeal of the Eighth Amendment itself.”⁴¹⁴

In 1992, the Irish Supreme Court expanded the pregnant person’s right to life to include risk of suicide after Irish authorities tried to stop a 14-year-old rape victim with suicidal ideation from leaving the country to obtain an abortion.⁴¹⁵ Then, nearly two decades later, in 2010, the European Court of Human Rights heard a case involving another pregnant cancer patient, finding that Ireland had failed to create a workable framework for determining when abortion was warranted to save a person’s life.⁴¹⁶

In 2012, Savita Halappanavar’s death pushed the issue over the edge. A young Savita was unable to get proper miscarriage care because her dying fetus still had a heartbeat, leading her to develop a preventable infection that killed her.⁴¹⁷ Savita’s death made clear that “[e]ven with a legally recognized exception for risks to the mother’s life, the Irish abortion ban caused the deaths of pregnant women by chilling doctors from acting.”⁴¹⁸ Her death became an international story and a domestic rallying cry for Irish abortion rights,⁴¹⁹ and less than a year later, the Irish Parliament passed a law attempting to create a framework to determine if a person’s life was at risk due to a physical or mental health condition.⁴²⁰

The fight continued, and the strong support for abortion in the context of physical or mental health risks, broadly defined, and for socioeconomic reasons led to the full repeal of the Eighth Amendment in 2018.⁴²¹ Today in Ireland, abortion is available—and paid for by the government—for any reason throughout the first trimester and thereafter for health—including mental health—reasons.⁴²² About the entire progression, Suk concluded:

The experience of other countries shows that firmly establishing access in the most compelling cases can be a step towards more robust access to abortion in the full range of cases. Litigating this defense in compelling cases thoughtfully and strategically could lay the groundwork for expanding it over time to include all reasonable abortions, including those justified by socioeconomic hardships and physical and mental health consequences of being pregnant under financial stress.⁴²³

The fight in Ireland took decades of incremental change but led to victories for all pregnant people. In America, there is still a predominant belief that “[m]ost women have abortions because having a baby would be an inconvenience to them.”⁴²⁴ A massive shift in reproductive health policy will not only involve legal wins, but also a shift in public opinion to recognize the serious physical and emotional consequences of unwanted pregnancy. Litigation surrounding medical exceptions—and the public debates they generate—can achieve both at once.

*David S. Cohen is a Professor of Law at Drexel Kline School of Law; Greer Donley is the John B. Nicklas, Jr. Faculty Fellow & Professor of Law at the University of Pittsburgh School of Law. We would like to thank Maya Manian, Jill Wieber Lens, Yvette (Yvonne) Lindgren, Elizabeth Sepper, Meghan Boone, and Katie Eyer for their fantastic feedback on the draft; participants at a University of Virginia Law School workshop, especially Micah Schwartzman; our longtime co-author, Rachel Rebouché, for her continued partnership and brain trust; and Caitlin Williams for her amazing research support.