Difference-in-differences Designs with Attrition Bias Correction

• In Event: Causal Inference with Time-Series - Cross-Sectional Data

Brief Overview

The goal of this paper is to extend the parallel trends assumption in a difference-in-differences design to impute missing outcome variables resulting from attrition and to identify a causal estimand while correcting for attrition bias.

Author

• Sooahn Shin, Harvard University