Friday, November 28, 2014

Applied Econometrics

I really enjoy Marc Bellemare's applied econ posts, but I really enjoy his econometric related posts (for instance a while back he wrote some really nice posts related  linear probability models   here and here).

Recently he wrote a piece entitled "In defense of the cookbook  approach to econometrics." At one point he states:

"The problem is that there is often a wide gap between the theory and practice of econometrics. This is especially so given that the practice of econometrics is often the result of social norms..."

He goes on to make the point that 'cookbook' classes should be viewed as a complement, not a supplement to theoretical econometrics classes.

It is the gap between theory and practice that has given me a ton of grief in the last few years. After spending hours and hours in graduate school working through tons of theorems and proofs to basically restate everything I learned as an undergraduate in a more rigorous tone, I found that when it came to actually doing econometrics I wasn't much the better, or  sometimes wondered if maybe I even regressed. At every corner was a different challenge that seemed at odds with everything I learned in grad school.  Doing econometrics felt like running in water. As Peter Kennedy states in the applied econometrics chapter of his popular A Guide to Econometrcs, "econometrics is much easier without data".

As Angrist and Pischke state: "if applied econometrics were easy theorists would do it."

I was very lucky that my first job out of graduate school was at the same university where I attended as an undergraduate, and I had the benefit of my former professors to show me the ropes, or the 'social norms' as mentioned by Bellemare. The thing is, all along, I just thought that since I was an MS vs PhD graduate, maybe I didn't know these things because I just hadn't had that last theory course, or maybe the 'applied' econometrics course I took was too weak on theory. But as Kennedy points out:

"In virtually every econometric analysis there is a gap, usually a vast gulf, between the problem at hand and the closest scenario to which standard econometric theory is applicable....the issue here is that in their econometric theory courses students are taught standard solutions to standard problems, but in practice there are no standard problems...Applied econometricians are continually faced with awkward compromises..."

The hard part for the recent graduate that has not had a good applied econometrics course is figuring out how to compromise, or which sins are more forgivable or harmless than others.

Another issue with applied vs. theoretical econometrics is software implementation. Most economists I know seem to use STATA, but I have primarily worked in a SAS shop and have taught myself R. But most of my coursework econometrics was done on PAPER with some limited work in SPSS.  GRETL is also popular as a teaching tool. Statistical programming is a whole new world, and propensity score matching in SAS is not straight forward (although here  314-2012 is a really nice paper if you are interested). Speaking of which, if you don't have the luxury of someone showing you the ropes, maybe the best thing you can do is attend some conferences. While not strictly an academic conference, SAS Global Forum has been a great conference with proceedings replete with applied papers with software implementation. R bloggers also offer some good examples of applied work with software implementation.

See also:
Culture War: Classical Statistics vs. Machine Learning

Ambitious vs. Ambiguous Modeling

Mostly Harmless Econometrics as an off-road backwoods survival manual for practitioners.

1 comment:

  1. Re: software implementation: It's also worth looking into Python, especially for those new to applied work. There is a growing network of people using Python and it is a pretty easy language to pick up.


Note: Only a member of this blog may post a comment.