Read on Twitter
I’ve been reading some neo-classical econometrics papers from 80-90s. Hansen-Singleton (Econometrica)
And Berry-Levinnshon-Pakes (Econometrica). They start out with stochastic economic models (to define structures/ Causal models). 1/n
And Berry-Levinnshon-Pakes (Econometrica). They start out with stochastic economic models (to define structures/ Causal models). 1/n
They make assumption that overspeed variables are generated to obey those models. At no point, there is any mention of Potential Outcomes or DAGs, the two super-combative approaches. Is this a deficiency or an advantage? 2/n
There are many papers in econometrics in this framework. You have Data ~ P. The economic model states that P = P (theta), where theta are model parameters. Given empirical observations, we make inference on theta. No DAGs, no POs. Did I miss anything? n/n
