An amazing fact I took away from job market papers: difference-in-difference is so CLEARLY the dominant research design of interest.
It's astonishing to me how much it dominates everything else.
@paulgp It also plays an important role in industry analytics, in antitrust, and in policy. Easy to understand and evaluate, do sensitivity, etc. Of course, always have to carefully consider assumptions!
@susanathey Funny comment from Stefan Hoderlein at the Interactions Econometrics conference this fall; he said he was blown away by how much all the analyses just use diff-in-diff (which I'm sure you know far better than most!) since so many of us outside of industry think they're all just running experiments!
@paulgp Shameless plug for our R package that makes nice visualizations! https://synth-inference.github.io/synthdid/