Difference in Difference Estimation of the Effects of Time Varying Treatment using structural nested mean models: Neyman seminar

Difference in Difference Estimation of the Effects of Time Varying Treatment using structural nested mean models: Neyman seminar

Neyman Seminar
May 4, 2022, 04:00 PM - 05:00 PM | Evans 1011. Zoom id: 97648161149. No passcode. Evans Hall | Happening As Scheduled
James Robins, Harvard University

Abstract: Economists and other social scientists have developed multiple methods that exploit longitudinal panel data to estimate causal effects, often using Difference-in-Differences (DID) when sequential ignorability is implausible. We have recently generalized DID methods using Structural Nested Mean Models. In this talk I will show that in conjunction with an assumption of no effect...