My reply (beyond that it's funny to see a reference to an economics paper by someone named "Dollar") is that it's a good idea to model panel data using three error terms, at the unit, time, and unit*time levels. You can (and should) also have predictors at all three levels, as appropriate. Also I'm not a fan of the overloaded term "fixed effects," but that's another story
Related;
A Primer for Panel Data Analysis
14.32 Econometrics (MIT OpenCourseWare)
Macro Aid Effectiveness Research: A Guide for the Perplexed
Through the Looking-Glass, and What OLS Found There: On Growth, Foreign Aid, and Reverse Causality
No comments:
Post a Comment