difference in differences

Introduction This blog post explores how Double Machine Learning (DML) extends to conditional Difference-in-Differences (DiD), focusing on doubly robust estimators. The key insight is that conditional DiD can be understood through the lens of cross-sectional ATT estimation.

0. Motivation Staggered‐adoption policies break the canonical two-period / two-group DiD model. It has been shown that the traditional two-way fixed-effects (TWFE) regression can assign negative weights to treatment effects, thereby obscuring their dynamic and heterogeneous patterns.

Motivation 1. Data feature – only one treatment treatment is assigned at an aggregate level, to group of individuals example: policy interventions often take place at an aggregate level, and affect aggregate entities, such as schools, or geographic or administrative areas only one or a few groups of treated units, and many more control units