Amid the battle over abortion rights, a failure to agree on how to define abortion

Every year, Lisa Campo-Engelstein tells her medical ethics class the story of Isabel: A fictional character who arrives at a health clinic seeking an abortion. Doctors determine that Isabel is 37 weeks pregnant and, what’s more, she’s suffering from high blood pressure that endangers the life of the fetus.

Thirty-seven weeks is just three shy of an average full-length pregnancy, so instead of an abortion, the clinic’s doctors recommend that Isabel have an emergency C-section to maximize the chance of a live birth. Isabel refuses. “I don’t want to get cut open to save a baby I didn’t even want in the first place,” she says. By refusing the C-section, is she having an abortion?

In Campo-Engelstein’s view, the answer is no — Isabel isn’t taking action to prevent a live birth (although that may be what ends up happening). But often her students disagree. “It always surprises me,” said the bioethicist from the University of Texas Medical Branch. ”A lot of the students will say, well, that’s an abortion then because she is choosing to end the child’s life.”

Abortion rights are positioned to be a major factor in the upcoming presidential election, the first since the Dobbs decision overturned Roe v. Wade and led to tighter abortion restrictions in several states. Three in 10 surveyed voters will only vote for a candidate who shares their views on abortion, the health policy research group KFF found last year. Campo-Engelstein’s classroom is a microcosm of a major problem underlying the debate over who can access these procedures, for what reasons, and when: Across the country, there’s widespread disagreement — among doctors, lawmakers, and the public — about what an abortion is. 

“You have a medical definition of abortion; you have 50-some-odd legal definitions of abortion; and then you have a bazillion public perceptions of abortion,” said OB-GYN Cara Heuser, who practices in Utah, where abortion is banned after 18 weeks. And a study published last month shows that states’ definitions just keep getting murkier. So when people talk about banning abortion, which definition are they talking about?

Over the last two years, states have struggled to define these procedures, as law scholars Greer Donley from the University of Pittsburgh and Caroline Kelly from the University of Chicago describe in a study that will soon be published in Duke Law Journal. A whopping 20 states have changed their definitions of abortion since the Dobbs decision, as lawmakers moved to exclude certain circumstances. Exclusions are different from exceptions, however. “This is not the state saying, ‘This is an abortion, but it’s OK. It’s an exception.’ This is the state saying, ‘this is not an abortion at all,’” Donley said.

Many states now exclude removing dead fetuses or removing ectopic pregnancies, for example. Several states also exclude pregnancies not known to the doctor treating the patient. Even in states that are relatively permissive of abortion, such as Massachusetts, definitions have shifted in response to patients’ experiences.

“The findings are entirely consistent with what I expected after Dobbs,” reproductive rights and justice scholar Ederlina Co from the University of the Pacific McGeorge School of Law wrote in an email to STAT. But like Donley, she thinks the complexity of pregnancy loss means lawmakers will always fail when they attempt to separate situations that should be allowed from those that should not.

OB-GYN Rachel Flink, who practices in upstate New York, has seen firsthand how nuanced the circumstances surrounding abortions can be. For example, medical complications can compound with other stresses — perhaps related to finances or family dynamics — and force people to terminate pregnancies. “Are they terminating a pregnancy because they don’t want to be pregnant? Or are they terminating a pregnancy because there’s complications? It’s not just one or the other,” she said.

Medical technology has its limits, so sometimes even the physician performing a procedure is unsure whether they’re giving their patient an abortion or treating a pregnancy complication. For example, Flink remembers talking with a medical resident about a pregnancy she suspected was ectopic, but terminated before the point at which she could make that diagnosis definitively without using an invasive procedure. “If it was an intrauterine pregnancy, then gosh, I guess we just did an abortion. But you know, we’ll never know,” she said. Self-induced abortions are similarly tricky, as doctors often can’t distinguish between these situations and miscarriages — ambiguity some patients may use to avoid stigma or legal action.

Donley has many more examples of ambiguous situations: What happens when doctors know a person will inevitably lose a pregnancy — or they’re almost sure, or they’re mostly sure — but the person has not lost the pregnancy yet? What if someone appears to be miscarrying very early in pregnancy, at a point when it’s very difficult to prove that the fetus has died? Trying to come up with a definition of abortion that excludes every possible ambiguity is a “fool’s errand,” Donley said. “You cannot do it.” 

Many doctors are just as confused as state legislators about what should and should not be considered an abortion, Flink, Heuser, and their colleagues have found. Flink is most comfortable with a definition she’s adapted from the American College of Obstetricians and Gynecologists: “An intervention designed to end a pregnancy without guaranteeing a live birth.”

But over time, she’s realized that many OB-GYNs lean on a circumstantial definition — if the procedure is done for maternal health indications, then they don’t consider it to be an abortion. Rather, said Flink, many of her colleagues believe that  “abortion is, like, intentionally doing something against the fetus.” 

In a 2023 survey of 209 OB-GYNs, for example, Flink, Heuser, and their colleagues found that only 44 doctors called inducing labor an abortion if it was done to protect the health of the pregnant person before the fetus could survive independently — something that falls squarely under Flink’s definition. 

This is consistent with views held by some members of the general public. Last fall, sociologist Alicia VandeVusse from the Guttmacher Institute and her colleagues published the results of a survey they undertook to learn whether abortion frequency is underreported because people are unsure how to define these procedures. When they asked participants whether various procedures and circumstances should be considered abortions, the range of responses they elicited demonstrated extensive misunderstanding and stigma. “I think it’s considered abortion only when you personally know you can take care of a child and you’re being selfish,” one participant said. 

Defining abortion around the motivation of the pregnant person “imputes just a huge value judgment,” said Kelly Cleland, the executive director of the American Society for Emergency Contraception, who has also researched medication abortions. “I don’t think there are, like, good abortions and bad abortions, which I think is a little bit of the implication of that.” 

Responses to VandeVusse’s survey suggest that such value judgments often inform people’s views on the subject, however. Like in Campo-Engelstein’s classroom, they can make people quicker to label a procedure as an abortion if they think the pregnant person has a choice in the matter, but they can also have the opposite effect. For example, when presented with the seemingly unambiguous statement, “had a surgical abortion” 25% of participants said they did not consider that to be a description of receiving an abortion.

“I was very surprised,” VandeVusse said. “People are so uncomfortable speaking about abortion, and the experience is so stigmatized … maybe people are trying to sort of not count things as abortions, even when they pretty clearly are.”

VandeVusse also posed this question to her study participants: What if a person takes abortion pills without doing a test to confirm they’re pregnant? Respondents were almost evenly split, with around a third saying this act qualifies as an abortion, a third saying it might, and a third saying it does not.

In some parts of the world, people routinely take advantage of this ambiguity, said demographer Suzanne Bell from Johns Hopkins University. Using drugs or herbs to “bring back” one’s period, without first confirming a pregnancy, is called menstrual regulation, and many people view this procedure in a softer, more forgiving light than abortion.

In Bangladesh, for example, menstrual regulation is legal, even though abortion is not. “As long as it’s been less than 10 weeks since the beginning of your last menstrual period, and as long as you have not confirmed or taken a pregnancy test … you can have a menstrual regulation to bring back your period. If you were to walk into that same facility and say, ‘I’d like to have an abortion,’ they would turn you away,” Bell said.

Many in the U.S. would also appreciate the option to take “missed period pills,” a 2020 study found. That’s just another name for abortion pills, but one that may make some patients feel more comfortable and might prove legally protective in states with abortion restrictions. For people without access to pregnancy tests, it’s normal to be unsure whether they’re pregnant — possibly for months. Menstrual regulation simply “codifies the ambiguity,” Bell said.

Before the Dobbs ruling poked holes in the nationwide right to abortion, some U.S.-based doctors were also drawing on the ambiguity around abortion — consciously or subconsciously — to ease the emotional turmoil often associated with abortion. If they could tell themselves or their patients that a procedure qualified as an early delivery instead, “then it was just better for everybody,” Flink said, paraphrasing some survey respondents’ views. 

After Dobbs, some doctors in restrictive states were dismayed when they had to rethink their definitions. “People were like, ‘Wait, this means I can’t do this thing that I’ve always done, and that I never even thought was an abortion? Now I have to ask the lawyer?’” Flink said.

Emergency contraception occupies an enduring niche in the debate over what “counts” as an abortion, leading some people to wonder about the future these drugs have in America. Indeed, Donley and Kelly’s analysis revealed that since the Dobbs decision, several states have removed exclusions related to contraception from their definitions of abortion. 

The debate over emergency contraception centers around when pregnancy begins. According to the medical definition, implantation of a fertilized egg in the lining of the uterus marks the beginning of the journey. This is in contrast to a definition popularized by the Catholic Church, in which life begins when an egg is fertilized. 

Emergency contraceptive pills work by preventing sperm and egg from ever meeting. They don’t cause abortions by the medical definition, and they’re extremely unlikely to cause abortions using a definition in which pregnancy begins at fertilization. But when the emergency contraceptive pill Plan B was undergoing FDA approval in the 2000s, an anti-abortion science adviser used a fringe interpretation of the scientific evidence to convince the agency to list preventing implantation as one of the drug’s possible mechanisms of action, said philosopher of science Christopher ChoGlueck from New Mexico Tech. That claim remained on Plan B’s label until 2022.

Although the FDA has now clarified the situation, many people are under the impression that emergency contraceptive pills cause abortions, including 22% of VandeVusse’s survey respondents, 38% of U.S. residents surveyed by KFF, and 39% of surveyed physicians. 

A lot of people hold these views simply because they’re misinformed, Cleland has found. But for a small number of people, preventing a fertilized egg from implanting is “what they would consider murder,” ChoGlueck said. Because of the potentially drastic consequences, they’re extremely reluctant to accept that emergency contraception does not cause abortion. “What I would argue is that no amount of evidence is ever going to convince them,” ChoGlueck said.

Americans have very little appetite for charitable discourse about how this should influence the categorization of emergency contraception, Campo-Engelstein said. Abortion bans have already begun to limit access to these medications. Idaho banned the pills in school-based health clinics, and Texas has long excluded emergency contraception from its state-funded family planning program.

Recently, certain legislators have attempted — so far unsuccessfully — to limit funding for emergency contraception more broadly. Donley doubts the political will exists to start banning forms of contraception for whole states or nationwide, but not everyone is so sure. “What I’ve learned in the last two years is that anything can happen,” Cleland said.

This story is part of ongoing coverage of reproductive health care supported by a grant from the Commonwealth Fund.