Simple Diagnostics for Two-Way Fixed Effects

TL;DR

This paper reviews the bias in two-way fixed effects difference-in-differences estimates when treatment effects vary over time, and proposes simple diagnostics to assess the severity of this bias, illustrated through a case study.

Contribution

It introduces straightforward diagnostic tools to evaluate bias in two-way fixed effects models with time-varying treatment effects.

Findings

Diagnostics effectively identify potential bias in estimates.

Case study demonstrates practical application of the diagnostics.

Highlights importance of checking assumptions in difference-in-differences analysis.

Abstract

Difference-in-differences estimation is a widely used method of program evaluation. When treatment is implemented in different places at different times, researchers often use two-way fixed effects to control for location-specific and period-specific shocks. Such estimates can be severely biased when treatment effects change over time within treated units. I review the sources of this bias and propose several simple diagnostics for assessing its likely severity. I illustrate these tools through a case study of free primary education in Sub-Saharan Africa.