On Optimal Inference in the Linear IV Model

This paper considers tests and confidence sets (CS's) concerning the coefficient on the endogenous variable in the linear IV regression model with homoskedastic normal errors and one right-hand side endogenous variable. The paper derives a finite-sample lower bound function for the probability that a CS constructed using a two-sided invariant similar test has infinite length and shows numerically that the conditional likelihood ratio (CLR) CS of Moreira (2003) is not always very close to this lower bound function. This implies that the CLR test is not always very close to the two-sided asymptotically-efficient (AE) power envelope for invariant similar tests of Andrews, Moreira, and Stock (2006) (AMS). On the other hand, the paper establishes the finite-sample optimality of the CLR test when the correlation between the structural and reduced-form errors, or between the two reduced-form errors, goes to 1 or -1 and other parameters are held constant, where optimality means achievement of the two-sided AE power envelope of AMS. These results cover the full range of (non-zero) IV strength. The paper investigates in detail scenarios in which the CLR test is not on the two-sided AE power envelope of AMS. Also, the paper shows via theory and numerical work that the CLR test is close to having greatest average power, where the average is over a grid of concentration parameter values and over pairs alternative hypothesis values of the parameter of interest, uniformly over pairs of alternative hypothesis values and uniformly over the correlation between the structural and reduced-form errors.

[1]  Marcelo J. Moreira,et al.  Optimal Two-Sided Tests for Instrumental Variables Regression with Heteroskedastic and Autocorrelated Errors , 2015, Journal of Econometrics.

[2]  D. Andrews,et al.  Identification- and Singularity-Robust Inference for Moment Condition Models , 2015, SSRN Electronic Journal.

[3]  Soumendu Sundar Mukherjee,et al.  Weak convergence and empirical processes , 2019 .

[4]  D. Andrews,et al.  Supplement to “ On optimal inference in the linear IV model ” : Online Supplementary Material , 2019 .

[5]  D. Andrews,et al.  ASYMPTOTIC SIZE OF KLEIBERGEN’S LM AND CONDITIONAL LR TESTS FOR MOMENT CONDITION MODELS , 2017, Econometric Theory.

[6]  Marcelo J. Moreira,et al.  Optimal Invariant Tests in an Instrumental Variables Regression With Heteroskedastic and Autocorrelated Errors , 2017, 1705.00231.

[7]  Isaiah Andrews.  Conditional Linear Combination Tests for Weakly Identified Models , 2016 .

[8]  Mark W. Watson,et al.  NEARLY OPTIMAL TESTS WHEN A NUISANCE PARAMETER IS PRESENT UNDER THE NULL HYPOTHESIS , 2015 .

[9]  D. Andrews,et al.  Asymptotic Size of Kleibergen's LM and Conditional LR Tests for Moment Condition Models , 2014 .

[10]  Marcelo J. Moreira,et al.  Tests based on t-statistics for IV regression with weak instruments , 2014 .

[11]  Anna Mikusheva,et al.  Conditional Inference with a Functional Nuisance Parameter , 2014, 1409.6337.

[12]  Marcelo J. Moreira,et al.  Contributions to the Theory of Optimal Tests , 2013 .

[13]  Marcelo J. Moreira.  Tests with correct size when instruments can be arbitrarily weak , 2009 .

[14]  G. Hillier.  EXACT PROPERTIES OF THE CONDITIONAL LIKELIHOOD RATIO TEST IN AN IV REGRESSION MODEL , 2009, Econometric Theory.

[15]  V. Chernozhukov,et al.  ADMISSIBLE INVARIANT SIMILAR TESTS FOR INSTRUMENTAL VARIABLES REGRESSION , 2007, Econometric Theory.

[16]  Marcelo J. Moreira,et al.  Efficient two-sided nonsimilar invariant tests in IV regression with weak instruments , 2008 .

[17]  J. MacKinnon,et al.  Bootstrap Inference in a Linear Equation Estimated by Instrumental Variables , 2008 .

[18]  G. Chamberlain.  Decision Theory Applied to an Instrumental Variables Model , 2007 .

[19]  Donald W. K. Andrews,et al.  Optimal Two‐Sided Invariant Similar Tests for Instrumental Variables Regression , 2006 .

[20]  Anna Mikusheva,et al.  Robust Confidence Sets in the Presence of Weak Instruments , 2005 .

[21]  Marcelo J. Moreira,et al.  Optimal Invariant Similar Tests for Instrumental Variables Regression , 2004 .

[22]  Jean-Marie Dufour,et al.  Projection-Based Statistical Inference in Linear Structural Models with Possibly Weak Instruments , 2005 .

[23]  Marcelo J. Moreira.  A Conditional Likelihood Ratio Test for Structural Models , 2003 .

[24]  Frank Kleibergen,et al.  Pivotal statistics for testing structural parameters in instrumental variables regression , 2002 .

[25]  D. Andrews.  Hypothesis testing with a restricted parameter space , 1998 .

[26]  Jean-Marie Dufour,et al.  Some Impossibility Theorems in Econometrics with Applications to Structural and Dynamic Models , 1997 .

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

[28]  D. Andrews,et al.  Optimal Tests When a Nuisance Parameter Is Present Only Under the Alternative , 1992 .

[29]  T. W. Anderson,et al.  Estimation of the Parameters of a Single Equation in a Complete System of Stochastic Equations , 1949 .

[30]  T. W. Anderson.  The Non-Central Wishart Distribution and Certain Problems of Multivariate Statistics , 1946 .

[31]  A. Wald.  Tests of statistical hypotheses concerning several parameters when the number of observations is large , 1943 .