Inconsistency of naive GMM estimation for QR models with endogenous regressors
A naive GMM approach to estimating QR (logit and probit) models with endogenous explanatory variables can lead to inconsistent estimators. This result was previously shown by Dagenais via simulation. In this paper, a special case is presented for which an analytical proof is possible; it turns out that the estimator is indeed inconsistent, but the framework analysed here can be useful for hypothesis testing. Un approccio GMM `naive' per la stima di modelli QR con regressori endogeni porta a stimatori inconsistenti. Questo risultato, ottenuto via simulazione da Dagenais, viene qui provato analiticamente in un caso particolare. Si ha che lo stimatore in effetti inconsitente, ma pu essere di una qualche utilit per la prova di ipotesi.
(This abstract was borrowed from another version of this item.)
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
- Lewbel, Arthur, 2000.
"Semiparametric qualitative response model estimation with unknown heteroscedasticity or instrumental variables,"
Journal of Econometrics,
Elsevier, vol. 97(1), pages 145-177, July.
- Arthur Lewbel, 1999. "Semiparametric Qualitative Response Model Estimation with Unknown Heteroskedasticity or Instrumental Variables," Boston College Working Papers in Economics 454, Boston College Department of Economics.
- Grogger, Jeffrey, 1990. "A simple test for exogeneity in probit, logit, and poisson regression models," Economics Letters, Elsevier, vol. 33(4), pages 329-332, August.
- Dagenais, Marcel G., 1999. "Inconsistency of a proposed nonlinear instrumental variables estimator for probit and logit models with endogenous regressors," Economics Letters, Elsevier, vol. 63(1), pages 19-21, April. Full references (including those not matched with items on IDEAS)