[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: how to test endogeneity of regressors in panel for Hausman-Taylor

From   John Bunge <>
Subject   st: how to test endogeneity of regressors in panel for Hausman-Taylor
Date   Mon, 11 Feb 2008 15:31:18 +0100

Dear all,

my concern is to estimate a panel with the Hausman Taylor methodology. This estimator assumes some of the time-invariant and time-variant covariates to be endogenous.

I would like to test the endogeneity of some regressors, i.e. correlation with individual random effect (remember that a two-component error process it is usually assumed, where I label the 'individual random effect' what is the time-invariant component). I heard from a test proposed by Baltagi, Bresson and Pirotte (2003): "Fixed Effects, Random Effects or Hausman-Taylor? A Pretest Estimator", published in Economics Letters 79, pp. 361-369. (this test is based on the well-known Hausman Specification test). Is this test implemented in Stata? 

Are there other ways to test the endogeneity of regressors and by this to analyze whether Hausman-Taylor might be more appropriate than Random Effects?



John Bunge
Debt and Finance Analysis Unit
United Nations Conference on Trade and Development (UNCTAD)
Palais des Nations
1211 Genève 10
Office: +41 229175902
Mobile: +41 762901769

Der WEB.DE SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index