Feasible Estimation of Linear Models with N-Fixed Effects

Levy Economics Institute of Bard College Working Paper No. 782

18 Pages Posted: 14 Dec 2013

See all articles by Fernando Rios‐Avila

Fernando Rios‐Avila

Bard College - The Levy Economics Institute

Date Written: December 12, 2013

Abstract

In this paper an alternative approach for the estimation of higher-order linear fixed-effects models is described. The strategy relies on the transformation of the data prior to calculating estimations of the model. While the approach is computationally intensive, the hardware requirements for the estimation process are minimal, allowing for the estimation of models with more than two high-order fixed effects for large datasets. An illustration of the implementation is presented using the US Census Bureau Current Population Survey data with four fixed effects.

Keywords: Fixed-Effects Models, Two-Step Estimation, Stata

JEL Classification: C13, C23

Suggested Citation

Rios‐Avila, Fernando, Feasible Estimation of Linear Models with N-Fixed Effects (December 12, 2013). Levy Economics Institute of Bard College Working Paper No. 782, Available at SSRN: https://ssrn.com/abstract=2366943 or http://dx.doi.org/10.2139/ssrn.2366943

Fernando Rios‐Avila (Contact Author)

Bard College - The Levy Economics Institute ( email )

Blithewood
Annandale-on-Hudson, NY 12504-5000
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
47
Abstract Views
401
PlumX Metrics