lukestein’s Twitter Archive—№ 2,039

@Osea82 @ryanbedwards Coauthor and I have found Pischke’s notes on measurement error helpful when thinking about how much is required to (by itself) explain a given level of 𝛽ols/𝛽iv. econ.lse.ac.uk/staff/spischke/ec524/Merr_new.pdf