AILSA CHANG, HOST:
A hundred fifty-four economists are wading into the abortion debate. In a brief to the Supreme Court, they wrote that access to legal abortion here has led to women attaining higher levels of education and professional occupation and lower rates of children in poverty. But how can they be so sure? Well, our colleagues at The Indicator From Planet Money, Adrian Ma and Wailin Wong, explain it's based on something called causal inference.
WAILIN WONG, BYLINE: Causal inference is all about using statistical tools to figure out how much one thing caused another thing. Caitlin Myers is an econ professor at Middlebury College.
CAITLIN MYERS: For an empirical social scientist, it is, like, our daily bread-and-butter lives.
ADRIAN MA, BYLINE: Now, under ideal conditions, figuring out causation is relatively straightforward. The gold standard would be a randomized controlled trial. Like, take the COVID vaccine.
MYERS: People who were randomly assigned the vaccine were much less likely to get COVID or to develop severe symptoms than people who weren't randomly assigned to receive the vaccine. It's a correlation. But the random assignment in a randomized controlled trial gives us a lot of confidence that we can interpret it as causal.
MA: Yeah, but it's not always possible or ethical to set up a randomized controlled trial. And so in those situations, researchers lean on natural experiments.
WONG: In the case of policies around abortion, Caitlin says the first natural experiment occurred even before Roe was decided. In the late 1960s, abortion was restricted across the country, but then a few states decided to loosen or repeal those restrictions. Caitlin says this is a good approximation for a randomized trial.
MA: And when researchers compared the experiences of women in those states to women in the rest of the country, they found some pretty stark differences.
MYERS: Legal access to abortion reduced the fraction of women who became teen moms by one-third, and it reduced the fraction of women who got married as teenagers by about a fifth. And I'll emphasize we're talking about causation, not just correlation.
MA: Another key tool in the causal inference toolkit that is often combined with natural experiments is something called multiple regression analysis. Basically, we're talking about very fancy math involving calculus and statistics, and you can think about it like this.
Do you remember ever seeing those things in, like, kids' magazines where they have two pictures and you're like, spot the difference?
MYERS: I like that. I like that metaphor. I think it works.
WONG: Multiple regression analysis allows economists to focus on the differences - you know, like access to abortion in certain states - while controlling for the similarities; think trends that were happening across the country like better access to contraception. And in the decades since Roe, a lot of economic studies have found access to abortion was the difference that shaped many women's financial lives.
MA: For example, there's a study looking at patients who were denied an abortion because of how many weeks pregnant they were and compared them to others who did get an abortion before the cutoff.
MYERS: These researchers saw that the group that got turned away was dramatically more likely to experience financial instability. They were about 80% more likely to have an adverse credit outcome, like a bankruptcy, than the other group.
WONG: And Caitlin says causal inference doesn't just allow economists to understand what happened in the past, it also helps them predict what will likely happen if Roe is overturned and about half the states end up banning abortion. Based on her research, she estimates in the first year post-Roe, around 100,000 women who want an abortion would be unable to reach a provider.
Wailin Wong.
MA: Adrian Ma, NPR News. Transcript provided by NPR, Copyright NPR.
