Abortion and Economic Policy, Part II: “The Key Economic Decision Many Women Will Make in Their Lifetimes”

In this series for Fireside Stacks, Roosevelt’s Hannah Groch-Begley interviews experts who study the economics of reproductive rights. You can find the first entry in the series here.

---

Back in 2021, the US Supreme Court heard arguments and read briefs relating to the Dobbs v. Jackson Women’s Health Organization case. To refresh our memories: The state of Mississippi, and its health officer, Thomas Dobbs, banned most abortion procedures after 15 weeks’ gestation. The only abortion clinic in the state, Jackson Women’s Health Organization, sued, arguing it had a constitutional right to perform the procedure for those who needed it beyond 15 weeks. Ultimately, in a June 2022 decision the court found in favor of Mississippi—and, in doing so, overturned decades of precedent. Previous cases, including Roe v. Wade (1973) and Planned Parenthood v. Casey (1992), had found that there was a constitutional right to abortion. After Dobbs, there was not.

One of the arguments the state of Mississippi presented to the court was that there was no reliable way to study the impact that the Roe v. Wade decision had on our society, particularly “on the ability of women to participate equally in the economic and social life of the Nation.” In fact, they went so far as to argue that there was “no good reason to believe that decades of advances for women rest on Roe” at all and thus that there would be no negative result from its undoing. Somehow, Roe had both “inflicted significant damage” and done nothing we could measure.

In response, Middlebury College Professor of Economics Caitlin Myers organized over 150 respected economists and researchers to push back on this claim in a formal amicus brief to the court.

Myers sat down with us to talk about the experience of putting together the brief, what economists know about legal abortion and “the ability of women to participate equally” in economic life, and the pain of watching the court actively ignore the hard data.

This interview has been edited and condensed for clarity.

Hannah Groch-Begley: What was it like writing a brief for the Supreme Court? What did you learn from that process?

Caitlin Myers: It was fascinating. I never would have imagined myself leading an amicus brief. I am an academic. I really like sitting at my desk and running regressions. I like publishing papers with tables of numbers, and I do my best to explain to the broader public what they mean. But I’m not an advocate. I'm not an activist. So when I saw Dobbs coming down the pike, I thought, I’ll probably not have much to do with this. I very naively assumed that the primary issue would be about constitutional law, and I am not an expert in constitutional law.

What I didn’t understand was that Mississippi would make one of the pillars of its argument this notion that there was no societal reliance on abortion, that Roe could be overturned and it wouldn’t affect anything for people. When I saw that, my jaw dropped, and I thought, Look, whatever you think about the ethics of this issue, whatever you think about constitutional law, this just isn’t right, and we have a large, robust body of scientific evidence that shows otherwise. I thought, There’s a role for me to play here.

Economists had not led a brief in any of the prior landmark abortion cases, probably because back in the ’70s and even the early ’90s the “credibility revolution” hadn’t really begun and we hadn’t yet done a lot of work on the effects of abortion policy. But it’s wrong to suggest there is no evidence today. So I talked to the Center for Reproductive Rights about this and got help from pro bono attorneys, who were incredible. What was really important to me was that the brief be all science. I outlined all of the literature and the points that we would make, and then I shared it with six or seven other people in the field who’ve written a lot on this. And I said, “I want to make sure that everyone agrees on every single thing here, that everybody agrees on the facts.”

It was already fall, I’m teaching students regression analysis, and so I just started saying to other economists, “I’m putting together this brief. If you’re interested in signing, let me know.” And I could not have predicted the outpouring of interest. So many people wanted to sign on to assert, “These are the facts.” And so we had 150 signatories very, very quickly. I’m sure we could have gotten many times that if I hadn’t been in a hurry. But it was really rewarding to hear people's feedback, to piece all the research together. It was a great opportunity to synthesize and reflect.

Hannah: You mentioned the credibility revolution. What’s happened to economics as a discipline in recent decades that allowed researchers to perform the kinds of studies that were at the heart of your brief to the Supreme Court?

Caitlin: In economics and many social science fields, a revolution began in the ’90s that some people call the credibility revolution. And this revolution is marked by a rise of new statistical techniques that let us go beyond just throwing up our hands and saying, “Well, correlation isn't necessarily causation, so we never know” to actually saying, “Sometimes correlation is causation; how do we figure out when it is?” The gold standard of this field—known by many as causal inference—is a randomized controlled trial. If a drug manufacturer wants to evaluate the safety and efficacy of a new drug, they take a large group of individuals, randomly assign some of them to receive the new drug, some of them to receive a placebo, follow them over time, and see what their outcomes look like. And because you randomly assigned the drug, you infer that if there’s any differences in outcomes for the two groups, it’s because it’s the causal effect of the drug.

But a lot of the questions we have in social science don’t lend themselves to that particular methodology. For instance, I’m really interested in the causal effects of the legalization of abortion in the early 1970s. We don’t have time machines. Even if we did have that technology, it’s not really ethical to come in and say, “We’re going to provide this care to this group of pregnant women but not to this group.”

What we look for in those situations instead are natural experiments. A natural experiment is a situation where researchers can credibly argue that some external force beyond their control randomly assigned the treatment or intervention. And so that’s the credibility revolution in the ’90s: the rise of the use of both controlled experiments and natural experiments to isolate and measure causal effects.

Hannah: That moment in 1973 when Roe takes effect. What does that natural experiment show?

Caitlin: The Roe era offers one of my favorite examples of a natural experiment, one that I use when I’m teaching courses on causal inference, because I think it’s a really compelling and intuitive example. Roe v. Wade was decided in January of 1973, and it had the effect of invalidating state bans on abortions. But that actually didn’t legalize abortion everywhere, because abortion was already legal in some states. This is something not everybody knows, but five states—California, New York, Washington state, Hawaii, Alaska—plus the District of Columbia relaxed their abortion regulations, either through legislative action or judicial action, beginning in late 1969.

The shock of legalization in those states starting in 1969 may have actually been greater than the shock of Roe v. Wade, because not only did their residents start to obtain abortions but people started to travel from nearby states. New York and DC in particular were destinations.

If you look at those early legalizing states, you can see birth rates fall in all of them. And you might think, Okay, that could be because they legalized abortion, and people who otherwise would have carried unwanted pregnancies to term are now getting abortions they wouldn't have gotten. Or you could think, Man, there’s a lot going on in the whole country, right? Vietnam War, civil rights movement, lots of other forces at work . . . maybe birth rates were going to fall anyway.

But what you can do is go look at the rest of the country in 1969 where abortion wasn’t yet legalized, and you’ll see that, while birth rates were declining, they did not decline nearly as steeply as in these early legalizing states. Initial analyses from economist Phil Levine and coauthors that rely on those comparisons suggest that the early legalization of abortion resulted in at least a 4 percent decline in births.

And that’s actually an underestimate, because the control group, these other states that you’re looking at, they’re not completely unaffected. People are traveling out of them for the procedure. If you exclude states neighboring the early repeal states from the controls, the estimated effect increases to 11 percent! So there’s this whole literature that’s developed rigorously examining the data from the era. What we conclude from all of these studies is that it’s clear the legalization of abortion allowed people who otherwise would have carried pregnancies to term to obtain an abortion, which might seem obvious to folks, but this is not a trivial point.

You’ll hear people argue, “The only thing that happens when abortion is legalized is people get safer abortions. People are going to get the same number of abortions no matter what.” The data actually doesn’t support that. Sure, plenty of people were finding a way, but there were also people—and they were particularly likely to be poor people, women of color, and young women—who weren’t finding a way. It’s clear that when the service became legal, it allowed more people to get abortions.

Hannah: You described it as the “shock” of legalization. Is it just the large shift in birth rates that suggests that language, or are there other effects you’re also seeing?

Caitlin: Yeah, that’s a great question. Shock is a word we often use when there’s a sudden application of a major treatment, a major policy change. There’s no doubt the legalization of abortion was a major policy change.

I want to emphasize that the data suggests that even when abortion wasn’t legal, relatively affluent, relatively more educated, white women were often finding a way to obtain an abortion anyway in this era. The populations that were the most impacted by the change in access were women of color and young women. If you think about teenagers, this is an age when women are—to use an economics term—acquiring human capital. In other words, they’re finishing their education, right? They’re developing experience in the labor force.

And my own work has shown that the legalization of abortion reduced the probability that a woman coming of age in the 1970s would become a teen mom by more than a third. It reduced the probability that she married as a teenager by about 20 percent. This has huge effects on their abilities to control how old they are when and under what circumstances they become mothers. And that is the key economic decision many women will make in their lifetimes.

The legalization of abortion allowed women to complete their education. It increased high school graduation rates, it increased college initiation, it increased college graduation rates. It then increased the probability that women entered higher-status occupations that paid better, and it lowered the probability that they and the children they eventually had would live in poverty.

I think it’s key to point out the clearest evidence of all of those effects is for Black women. These were populations who were particularly likely to be unable to access abortion when it wasn’t legal. Researchers have also looked at maternal mortality. Work by economist Sherajum Monira Farin and others shows that the legalization of abortion reduced Black maternal mortality by somewhere between 30 and 50 percent. Huge, absolutely huge effects.

And the policy change happened quite suddenly. It was preceded by reforms, so it’s not as though it came out of nowhere. In some states, it slowly became a bit easier, a bit more possible, to obtain an abortion under certain restrictive circumstances. But what we can see in the numbers is that legalization was the policy change that suddenly started allowing far more people than before to obtain abortions. And I think shock is a pretty fair word.

Hannah: Roe was in place for nearly 50 years. So does the data show significant change over time?

Caitlin: Absolutely—when you see such huge effects from the legalization of abortion on women in the ’70s, the obvious next question is: Does it still matter? And you can tell stories about why it might not. In fact, if you read Mississippi’s brief in the Dobbs case, they tell these stories. They say even if the legalization of abortion had societal impacts in the ’70s, so much has changed. We have much greater access to effective contraception, we have antidiscrimination laws that cover pregnancy, we have paid family leave, we have subsidized childcare.

I find those arguments silly. They’re almost glibly saying everybody can now effortlessly balance parenthood and work. You just have to know a parent, or maybe even just spend a moment thinking about parents. I say this having four kids and as somebody who has a lot of resources to care for them. Childcare is enormously expensive. It’s about 75 percent of the household income of a low-income family to have high-quality, center-based childcare. Subsidies are inadequate. Most people don’t have them. So the idea that we’ve just policy-advanced ourselves out of a situation where childbearing impacts people’s economic lives is not true, at least not yet.

Those arguments really overlook the fact that not everybody has access to the most effective forms of contraception, particularly in states that haven’t expanded Medicaid. In Mississippi, about a quarter of the population of people in their 20s lacks health insurance. Unintended pregnancies are still very common. They’re almost half of all pregnancies. Most people don’t have paid leave, and particularly low-income populations that are most likely to experience unintended pregnancy—most likely to seek abortions—are also most likely to be working in shift work, which makes it even less likely that they’re going to be able to arrange time off work.

Access to legal abortion absolutely made a difference for those women.

Hannah: So you collect all this research into a single brief, get 150 signatures signing off on it, and present it to the court. What happened then?

Caitlin: I care deeply about undergraduate education, and when I was writing the brief, I was also teaching 70 students in a second-year-level statistics class. I told them I was writing the brief and I was bringing these methodologies in. And so oral arguments in Dobbs were on the day I had back-to-back labs for students to practice. In between labs, I was listening to the oral arguments, and a lot of my students had gathered around to listen as well. It just so happened that in that moment between labs, Justice John Roberts asked the attorney for the Center for Reproductive Rights, representing Jackson Women’s Health, something like, “Is there any reason to believe there is societal reliance on abortion? Is there any, you know, evidence on the causal effects of abortion policy?” [Editor’s note: you can see this exchange on pages 52–53 of the transcript of the Dobbs oral arguments.]

And the attorney responded, “I would refer, Your Honor, to the economists’ amicus brief.” And she began to talk about the brief, about the methodology, about the things that I’ve been talking about with you. And my students were all around listening, and they’re learning these methodologies. And I have to say that there were roughly 15 seconds in which I thought, Here is this coming together of all my research and all my teaching, and students are seeing how the methodologies are used and discussed and described, and it was really an absolutely amazing moment. Until Justice Roberts interrupted and said, “Let’s set the data aside.”

Hannah: Oh my god.

Caitlin: I felt as though it were just all crashing down around me. You know, I’m sitting here with all of these young people learning statistics and giving them examples of evidence-based policy, and then we’re just going to set the evidence aside. I felt like the writing was on the wall for Roe. I never had any kind of grandiose idea that I would change everything about this. But I did imagine that the evidence somehow would be relevant, that people would listen, that people would care about the facts. I have a cold, dark economist heart, but that kind of broke it.

Hannah: I hope that the lesson isn’t that this doesn’t matter. I hope that the lesson is that you’ve got to keep trying, and the evidence still matters, and your students still need to learn all this.

Caitlin: In some ways, it's even more incentive for me to be an effective teacher, because fundamentally, what I’m teaching students is how to be critically engaged citizens, how to evaluate statistical evidence and really understand it, really look at it, really ask the right questions about the methodologies. And so to some extent, I couldn’t help but wonder, Is part of the problem here insufficient education in statistics and quantitative methods? People ought to be interested in the evidence and be able to listen to the evidence and the basis for the evidence and evaluate it.

We still care about the data here at Roosevelt. The second part of our conversation with Professor Myers will dive into the evidence on what has changed in the three years since Dobbs—and what we still need time to understand.

If you ask Eleanor

“The emphasis of the planned-parenthood movement, as far as I know, has never been on having fewer children, but on planning for the birth of every child so that it will be born when the mother is well and strong and the family able to provide for it adequately.”

- Eleanor Roosevelt, If You Ask Me (April 1945)

Fireside Stacks is a weekly newsletter from Roosevelt Forward about progressive politics, policy, and economics. If you enjoyed this installment, consider sharing it with your friends.