Speaker: Wooyong Lee (University of Technology Sydney)

Host: Xi Wang, School of Economics

Time: 10:00-11:30 pm, March 31, 2023, GMT+8

Venue:  Zoom Meeting ID: 896 7292 2199 Passcode: 500394

Abstract:

This paper studies differences-in-differences (DID) models that allow the units to endogenously enter treatment in response to the variation in the time-varying unobservable covariates, which relaxes the standard parallel trend assumption and permits the well-known Ashenfelter's dip phenomenon (Ashenfelter, 1978). I show that, in this model, the average treatment effect on the treated (ATT) is identified only if there exists a group of exogenously untreated units, and that otherwise the data has no identifying power about the ATT. I then show that, even if there are no exogenously untreated units, the average treatment effect on the not yet treated (ATNT), which is the average effect of the treatment that the units would have experienced if they entered into treatment earlier than the actual treatment period, is partially identified under suitable additional assumptions. I propose an estimation and inference procedure for the ATT and the ATNT, whose finite-sample performance is examined by simulations.

Biography:

Wooyong Lee is currently a lecturer (assistant professor) of Economics Discipline Group, UTS Business School, University of Technology Sydney. He received PhD from the University of Chicago in June 2020.

Source: School of Economics