Is your econometrics textbook wrong?

This report surveys six influential econometric textbooks in terms of their mathematical treatment of causal concepts. It highlights conceptual and notational di fferences among the authors and points to areas where they deviate signi cantly from modern standards of causal analysis. We find that econonometric textbooks vary from complete denial to partial acceptance of the causal content of econometric equations and, uniformly, fail to provide coherent mathematical notation that distinguishes causal from statistical concepts. This survey also provides a panoramic view of the state of causal thinking in econometric education which, to the best of our knowledge, has not been surveyed before.

A new paper by Chen and Pearl.

I have only had time to glance and it seems you have to have read Pearl’s textbook to know more precisely what he is talking about. I’d be interested in hearing reader impressions of Pearl’s critique.

Hat tip to PM.

12 thoughts on “Is your econometrics textbook wrong?

  1. I have to say that this is one of Pearl’s weakest papers, but I guess if it is making it’s way through the blogosphere and thereby drawing attention to his other work then it has served its purpose.

  2. Having taught undergraduate econometrics for six years now using JM Woodridge’s text, I’m glad to see that Chen and Pearl have a relatively favorable review of its treatment of causality. As per their ‘ideal’ treatment, it distinguishes between the ‘population regression function’, i.e. structural model, and the ‘sample regression function’, i.e. the estimated regression equation. The parameters of the structural model are interpreted as having causal effects on the conditional expectation of Y only if the model is correctly and completely specified, assuring the ceteris paribus interpretation. So I would say no, ‘my’ econometrics textbook is not wrong – just misunderstood. But I agree with them that there is scope for improvement in the treatment of causality in the introductory texts.

  3. Pingback: Somewhere else, part 27 | Freakonometrics