Estimation with Weak Instruments: Accuracy of Higher-Order Bias and MSE Approximations

In this paper, we consider parameter estimation in a linear simultaneous equations model. It is well known that two-stage least squares (2SLS) estimators may perform poorly when the instruments are weak. In this case 2SLS tends to suffer from the substantial small sample biases. It is also known that LIML and Nagar-type estimators are less biased than 2SLS but suffer from large small sample variability. We construct a bias-corrected version of 2SLS based on the Jackknife principle. Using higher-order expansions we show that the MSE of our Jackknife 2SLS estimator is approximately the same as the MSE of the Nagar-type estimator. We also compare the Jackknife 2SLS with an estimator suggested by Fuller (Econometrica 45, 933--54) that significantly decreases the small sample variability of LIML. Monte Carlo simulations show that even in relatively large samples the MSE of LIML and Nagar can be substantially larger than for Jackknife 2SLS. The Jackknife 2SLS estimator and Fuller's estimator give the best overall performance. Based on our Monte Carlo experiments we conduct informal statistical tests of the accuracy of approximate bias and MSE formulas. We find that higher-order expansions traditionally used to rank LIML, 2SLS and other IV estimators are unreliable when identification of the model is weak. Overall, our results show that only estimators with well-defined finite sample moments should be used when identification of the model is weak. Copyright Royal Economic Socciety 2004

[1]  Liang-sheng Lu,et al.  [Expression of fusion proteins in beta(2)GP I gene-transfected HEp-2 cells and its clinical application]. , 2002, Zhonghua yi xue za zhi.

[2]  J. MacKinnon,et al.  Estimation and inference in econometrics , 1994 .

[3]  K. R. Kadiyala,et al.  Asymptotic probability concentrations and finite sample properties of modified LIML estimators for equations with more than two endogenous variables , 2000 .

[4]  T. Rothenberg.  ASYMPTOTIC PROPERTIES OF SOME ESTIMATORS IN STRUCTURAL MODELS , 1983 .

[5]  T. Sawa.  Finite-Sample Properties of the k-Class Estimators , 1972 .

[6]  Whitney K. Newey,et al.  Higher Order Properties of Gmm and Generalized Empirical Likelihood Estimators , 2003 .

[7]  THE CASE AGAINST JIVE , 2006 .

[8]  Thomas J. Rothenberg,et al.  Approximating the distributions of econometric estimators and test statistics , 1984 .

[9]  MEAN SQUARED ERROR REDUCTION FOR GMM ESTIMATORS OF LINEAR TIME SERIES MODELS , 2002 .

[10]  M. Chavance.  [Jackknife and bootstrap]. , 1992, Revue d'epidemiologie et de sante publique.

[11]  Takamitsu Sawa,et al.  DISTRIBUTIONS OF ESTIMATES OF COEFFICIENTS OF A SINGLE EQUATION IN A SIMULTANEOUS SYSTEM AND THEIR , 1973 .

[12]  Naoto Kunitomo,et al.  Evaluation of the Distribution Function of the Limited Information Maximum Likelihood Estimator , 1982 .

[13]  Stephen G. Donald,et al.  A jackknife interpretation of the continuous updating estimator , 2000 .

[14]  Wayne A. Fuller,et al.  Some Properties of a Modification of the Limited Information Estimator , 1977 .

[15]  K. Morimune.  Approximate Distributions of k-Class Estimators when the Degree of Overidentifiability is Large Compared with the Sample Size , 1983 .

[16]  A. L. Nagar.  The Bias and Moment Matrix of the General k-Class Estimators of the Parameters in Simultaneous Equations , 1959 .

[17]  Jinyong Hahn,et al.  A New Specification Test for the Validity of Instrumental Variables , 2000 .

[18]  P. Phillips,et al.  The Exact Distribution of Instrumental Variable Estimators in an Equation Containing n + 1 Endogenous Variables , 1980 .

[19]  M. Akahira.  ASYMPTOTIC DEFICIENCY OF THE JACKKNIFE ESTIMATOR , 1983 .

[20]  T. W. Anderson,et al.  The numerical values of some key parameters in econometric models , 1983 .

[21]  C Hale,et al.  The Bias of Instrumental Variable Estimators of Simultaneous Equation Systems , 1977 .

[22]  Stephen G. Donald,et al.  Choosing the Number of Instruments , 2001 .

[23]  Peter C. B. Phillips,et al.  A Saddlepoint Approximation to the Distribution of the k-Class Estimator of a Coefficient in a Simultaneous System , 1979 .

[24]  J. Angrist,et al.  Jackknife Instrumental Variables Estimation , 1995 .

[25]  Jinook Jeong,et al.  ON THE EXACT SMALL SAMPLE DISTRIBUTION OF THE INSTRUMENTAL VARIABLE ESTIMATOR , 1992 .

[26]  Jerry A. Hausman,et al.  Notes on bias in estimators for simultaneous equation models , 2002 .

[27]  Jerry A. Hausman,et al.  Specification and estimation of simultaneous equation models , 1983 .

[28]  J. Stock,et al.  Instrumental Variables Regression with Weak Instruments , 1994 .

[29]  Paul A. Bekker,et al.  ALTERNATIVE APPROXIMATIONS TO THE DISTRIBUTIONS OF INSTRUMENTAL VARIABLE ESTIMATORS , 1994 .

[30]  R. Mariano,et al.  The Exact Finite-Sample Distribution of the Limited-Information Maximum Likelihood Estimator in the Case of Two Included Endogenous Variables , 1972 .

[31]  Takamitsu Sawa,et al.  Evaluation of the Distribution Function of the Two-Stage Least Squares Estimate , 1979 .