Matching Methods for Causal Inference with Time‐Series Cross‐Sectional Data

@KusanoKodai I think PanelMatch will be the future https://t.co/jQJUu9a4m6

I defined my causal estimand explicitly in terms of recent *change in treatment status*, as my focus was on the effects of a *recent* police stop at gunpoint. With 3 time points, I used matching methods for panel data developed by Imai et al. (2021) https:

Imai just shared this paper at #causalinference22 Great, 'so what' moment 🤯 Pulling apart the effects of democratization vs authoritarian reversal through matching #WhySabbaticalRocks

RT @DToshkov: Acemoglu et al. (2019) in JPE argue that democracy promotes economic growth (>1,000 citations already). A a new article, Kosu…