A new paper details the real health effects that the Tuskegee Study had on black people. The work from Marcella Alsan at the Stanford Medical School and Marianne Wanamaker at the University of Tennessee provides evidence for a strong claim: that by 1980, the public revelation of the Tuskegee Study in 1972 had reduced life expectancy among black men over 45 by over a year.
The study provides causal evidence in a body of research that stretches back decade, and it is a major revelation. As New York Magazine notes, the main conclusion of Alsan and Wanamaker’s work is that the Tuskegee Study was responsible for over a third of the life expectancy gap between older black men and white men in 1980. But the conclusions extend beyond that and fit with a body of anecdodal evidence from physicians and patients. Disclosure of the Tuskegee Study disrupted a slow convergence of black health outcomes with white health outcomes in the mid-20th century, accelerated an erosion of trust in doctors, and dampened health-seeking behavior and health-care utilization for black men. And it did so more for those who were more similar to the Tuskegee Study’s unwitting participants and those who were closely connected to the community of Tuskegee.
For Alsan and Wanamaker, the study spoke to longstanding interests, but required some methodological innovations. “From the medical side I had been sort of acutely aware of the Tuskegee experiment,” Alsan told me. “It would come up among the medical staff and even occasionally with patients. So I think that was one of the seeds that made me think that this could potentially be important.” He connected with Wanamaker, who had been studying racial disparities in education and labor markets, and both became convinced that this nefarious study might have actual quantifiable effects that could be uncovered.
One of the important contributions of their paper is the methodology it uses. It draws mostly on publicly accessible data on medical trust from the General Social Survey, data on health utilization from the National Health Interview Survey, Census measures of migration patterns, and data from the Centers for Disease Control and Prevention on mortality and morbidity. The innovation lies in how Alsan and Wanamaker used those data. Using a triple-difference model, the year 1972 was treated as the “exposure” itself since there aren’t real measures of how the news of the study was disseminated. As Wanamaker said, “we knew that there should be a sharp difference in 1972 and we knew there should be a difference between black men and any other peer group.”
There were two problems though. Black men and white men are already starkly different in health utilization and outcomes, and Alsan and Wanamaker needed some piece of data that could identify the Tuskegee Study itself as a factor in the divergence of those health outcomes. To accomplish this, they used distance from the epicenter of the Tuskegee Study, demographic similarity to the Tuskegee Study’s participants, and migration clusters, to identify if something that happened in Tuskegee in 1972 could be implicated. What they found is almost too perfect: outcomes, trust, and life expectancy diminished with proximity to Macon County; men who were more similar to Tuskegee’s uneducated, poor, medically underserved population were more likely to suffer diminished outcomes; and migrants from Alabama in the wake of Tuskegee were very likely to carry with them these diminished outcomes.