Diff-in-diff papers and their impact

Guido Imbens and Susan Athey said a few years ago that synthetic control was the most important innovation in causal inference of the last 15 (now 20) years and as of a week ago, Andrew Goodman-Bacon’s 2021 article on difference-in-differences with differential timing has passed the original synthetic control article in total citations.

Bacon decomposition (left) and Abadie and Gardeazabal 2003 synth (right)

So, what I’m goin…