We’ve Been Underestimating Discrimination (Rose Jacobs, February 20, 2024, CBR)

The layered relationship over time between identity and opportunity make up the infrastructure of systemic discrimination, a phenomenon that social scientists have studied since the 1950s and that is increasingly acknowledged across American society, despite resistance from the Right. But in economics, practitioners have traditionally studied only direct discrimination, with projects that have a narrower scope. Take, for example, a study from three Federal Reserve economists—Neil Bhutta, Aurel Hizmo, and Daniel Ringo—that analyzes the extent to which lenders provided differential treatment by race, illegal under US fair-lending laws, in 2018 and 2019. The study establishes a steep decline in racial discrimination in mortgage issuance, as compared with research findings from a study of home loan applications in 1990, which is encouraging. But the researchers in both studies controlled for factors such as applicants’ credit scores and leverage, a standard economic approach but one that drew ridicule from journalist Michael Hobbes, who tweeted, “Yes[,] once you remove the influence of all of the other racist systems, racism doesn’t exist.”

The thinking among economists about how to account for such factors may be changing, however. University of Pennsylvania’s J. Aislinn Bohren, Brown’s Peter Hull, and Chicago Booth’s Alex Imas are among the economists who are proposing new approaches to measuring discrimination that take systemic factors into account. They are looking at the mechanisms by which historical discrimination continues to create unequal outcomes while also acknowledging the limits of economists’ traditional measurement tools and extending the tape measure—rethinking their models so that quantitative data can better illuminate whether the American dream is available to all. The research by Bohren, Hull, and Imas indicates that traditional estimates can undercount discrimination, and not by just a little: they sometimes miss the majority of the total.