Contact Us

Phone: (202) 885-2940 Graduate programs: spagrad@american.edu Undergraduate programs: aupublicaffairs@american.edu

Kerwin Hall

Public Affairs, School of 4400 Massachusetts Avenue NW Washington, DC 20016 United States

Back to top

Meet Seth Gershenson Education Policy and Economics Scholar

American University School of Public Affairs Associate Professor Seth Gershenson is an expert in the economics of education, education policy, labor economics and time use. In addition to teaching courses on quantitative methods, managerial economics, and economics for policy analysis to graduate students, he also coordinates the SPA Analytics and Management Institute.

As coordinator of the SPA Analytics and Management Institute, he leverages his networks in Washington to bring leading experts outside of academia to teach students practical skills sought after by policy shops, think tanks, federal agencies, and consultancies. This includes technical skills, such as using cutting edge mapping software and acquiring the legal know-how to comply with regulations, as well as creative skills of writing, blogging, data visualization, and engaging with the media. The program is aimed at providing graduate students with tangible skills that complement their academic coursework and increase their marketability.

Gershenson was a visiting scholar at the Institute for Health & Social Policy at Johns Hopkins University in the spring of 2014 and has been selected to be visiting scholar in the Center for Education Policy Analysis at Stanford University and the Education Research Section at Princeton University during his sabbatical in the spring of 2018.

Prior to joining AU's School of Public Affairs, Gershenson received an Outstanding Teaching award from Michigan State University and was honored with the New Scholar Award by the Association for Education Finance & Policy. He earned his bachelor's degree in economics from Drexel University and his doctorate in economics with a focus on education policy from MSU.

Since 2015, Gershenson has been an IZA Research Fellow for the Institute of Labor Economics. He serves on the editorial board of Educational Evaluation and Policy Analysis, SAGE Open, Journal of Education for Students Placed at Risk and is a technical advisor for the Johns Hopkins Institute for Education Policy.

He won the Thomas A. Downes Award in 2016 for the best article published in Education Finance & Policy. Gershenson has been honored with the Outstanding Teaching Award by AU in the School of Public Affairs in 2014 and was named the Emerging Education Policy Scholar by AEI & Thomas B. Fordham Institute in 2014.

Building Reliable Research Models

Using his research expertise, Gershenson will focus on some cutting-edge tools of statistical analysis in his course, "Friends Don't Let Friends Lie with Statistics: A Short Course on Causal Inference." He will cover the tools and methodology of the so-called "credibility revolution" in modern micro-econometrics and program evaluation. There will be discussion of the workhorse experimental and quasi-experimental methods for estimating and identifying causal effects, even in the absence of a true randomized experiment. This course introduces methods for identifying and estimating causal effects from experimental and non-experimental (observational) data, of both the cross-sectional and panel (longitudinal) variety. Specific topics include the Rubin causal model, regression discontinuity designs, matching estimators, difference-in-difference models, the synthetic control method, instrumental variable estimation, machine learning, and quantile and heterogeneous treatment effects.

"This course is really important because identifying the causal impacts of programs is crucial to conducting valid cost-benefit analyses and making evidence-based policy - it is quite literally the evidence - but in the absence of randomized controlled trials, teasing out causal relationships is tricky," said Gershenson. "This course teaches the tricks necessary to uncover causal relationships in observational data."