The Importance of Estimating Potential Outcomes

A person I had always wanted to get to know Dr. Elizabeth Stuart, a professor at Johns Hopkins in their biostatistics department. I knew about her for a long time before I met her because of her expansive work on a variety of issues in the area of “matching” and unconfoundedness. She did her PhD, as it turned out, at Harvard at the end of the 1990s and early 2000s around the time when Guido Imbens was still there in the economics department, and Don Rubin in the statistics department. At Harvard she worked with people like Don Rubin, her dissertation adviser, as well as Gary King, one of her collaborators and someone else I’ve interviewed on the podcast, and so I wanted to talk to her to try and piece together more of the progression of causal inference throughout the social sciences in the late 20th and early 21st century, not just through writing, but maybe even moreso through students and faculty placements at departments around the world.

But these big ideas are in many ways just the “hook”, as I have said, to build a mental map of why I select certain people for the podcast. Dr. Stuart is an important scholar in her own right. She has spent a career being driven by questions about health and selected into statistics as a way of enhancing her own ability to contribute fruitfully to large and important policy questions regarding health. After graduating from Harvard in 2004, she went to Mathematica before then moving to Johns Hopkins school of public health where she steadily moved forward through tenure to associate then full professor. She is now a professor in the department of mental health, the department of biostatistics, and the department of health policy and management at Johns Hopkins. And she is now leading up pioneering new curriculum options for students there as well as moving into a new administrative position within the university.

I learned things I didn’t know, such as her brief flirtation with going to Princeton’s economics program (the economics students, though, seemed miserable so she opted against it). Since I’ve been also obsessed with trying to better understand Princeton’s economics program throughout the 1970s to 1990s, I was surprised to again realize what a small world it was that Dr. Stuart herself skipped over that like a stone over water before landing at the center of the causal inference universe itself — Harvard’s statistics department. So this was a fun interview. And I hope you enjoy learning more about Dr. Stuart’s life.

If you enjoy this podcast interview, or any of the others, please share it, as well as follow, like and even consider subscribing! The substack goes to subsidizing the cost of paying “my guy” who turns the raw interviews into usable podcast and YouTube videos. Thank you again for all your support!

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.