Abstract

This paper discusses, in a panel data context, a class of estimators for a slope coefficient in a regression equation with a two-components disturbance. The class of estimators is constructed by a weighting of second order moments based on the ‘within’ variation in the data set at the individual and at the time specific level. It contains as special cases the ordinary and the generalized least squares estimators, various ‘within’ and ‘between’ estimators, and estimators which can be given instrumental variable interpretations. We illustrate and interpret the relationship between the estimators by considering the estimation of marginal budget shares in Norwegian household consumption from a two-wave panel data set.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call