(available only to the uiuc domain)
This will be a half semester introduction to semiparametric methods in econometrics. We will begin with a brief introduction to semiparametric efficiency, a la van der Vaart. I will then describe some relatively new methods for causal inference in semiparametric treatment effect models. Some recent developments in quantile regression will then be surveyed, and the course will conclude a review of some recent developments concerning mixture/fraility models. Students are expected to have a background equivalent to my 574 Econometric Theory course.
1. The quantile regression lecture slides are in the QR subdirectory. Just click on the Lectures link and then click on that directory.
2. I've added an unpublished paper by Keith Knight and Gib Bassett that describes some very curious asymptotics for estimators based on averaging quantile estimates. Just click on the Readings link and look for BassetKnight.pdf. There is another paper there (Gib.pdf) that provides a more applied viewpoint.
3. The first class will be March 15, and will cover the material in L1 and part of L2 as posted on the course website.
4. There may be Just in Time questions for class available from here. But there may not.