lukestein’s Twitter Archive—№ 7,621

Your regular reminder that if you plan to assess virtue using panel regressions, you really should be including time fixed effects @itaisher/1250852646262628352