On B-robust Instrumental Variable Estimation of the Linear Model with Panel Data
Journal of Econometrics, 2002, 106, 2, 297-324
WALDMANN, Robert, WAGENVOORT, Rien J.L.M., On B-robust Instrumental Variable Estimation of the Linear Model with Panel Data, Journal of Econometrics, 2002, 106, 2, 297-324 - https://hdl.handle.net/1814/23694
Retrieved from Cadmus, EUI Research Repository
The aim of this paper is to demonstrate how to acquire robust consistent estimates of the linear model when the fundamental orthogonality condition is not fulfilled. With this end in view, we develop two estimation procedures: Two stage generalized M (2SGM) and robust generalized method of moments (RGMM). Both estimators are B-robust, i.e. their associated influence function is bounded, consistent and asymptotic normally distributed. Our simulation results indicate that the relatively efficient RGMM estimator (in regressions with heteroskedastic and/or autocorrelated errors) provides accurate parameter estimates of a panel data model with all variables subject to measurement errors, even if a substantial portion of the data is contaminated with aberrant observations. The traditional estimation techniques such as 2SLS and GMM break down when outliers corrupt the data.
(Partially based on Chapter 3 of Wagenvoort's EUI PhD thesis, 1998.)
Cadmus permanent link: https://hdl.handle.net/1814/23694
Full-text via DOI: 10.1016/S0304-4076(01)00102-6
ISSN: 1872-6895; 0304-4076
Files associated with this item
There are no files associated with this item.