Steve Pudney

Personal homepage

1977 discussion paper “The estimation and testing of some error component models”

I get occasional requests for copies of an old LSE discussion paper (no. A13, 1977), The estimation and testing of some error components models, which was one of the first analyses of the dynamic fixed-effects regression model for short panels. It proposed a forerunner of the now-standard GMM estimator and discussed various tests and extensions of the panel regression model to the wider class of multi-equation and random coefficient models.