Choosing the Number of Instruments

By Stephen G. Donald, Whitney K. Newey. 2001

403 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/2334e92f62f8e00c187ac57d7dfd96166c0d8435

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

Citations in the corpus

Improved JIVE Estimators for Overidentified Linear Models with and without Heteroskedasticity (2009, background): “First note that with homoskedasticity, (16) is zero for both CI J I V E and CN AG AR.”
Performance of Conditional Wald Tests in IV Regression with Weak Instruments (2007, background): “of k for these estimators are (see Donald and Newey (2001) ):”
The optimal choice of moments in dynamic panel data models (2009, methodology): “Both methods are based on minimization of the asymptotic MSE.”
Constructing Optimal Instruments by First-Stage Prediction Averaging (2010, methodology): “define the residuals ˜ = y X ˜ . As pointed out in Donald and Newey (2001) , it is important”
Sparse Models and Methods for Optimal Instruments With an Application to Eminent Domain (2012, methodology): “Our use of Lasso as a variable selection technique does not require any a priori knowledge about the identity of the most relevant instruments and so provides a useful complement to Donald and Newey (2001) and Okui (2010).”
Efficient estimation of general dynamic models with a continuum of moment conditions (2007, methodology): “The first method consists in computing the MSE analytically using a second order expansion as in Donald and Newey (2001).”
Finite sample evidence of IV estimators under weak instruments (2007, methodology): “Two other -class estimators have been considered for use under the presence of weak instruments: the ‘modified LIML’ estimator by Fuller (1977), which was expressly developed to have finite sample moments and is obtained when D ̨ T n where is the same as for LIML, and ̨ > 0 is a constant to be cho”
On the Asymptotic Optimality of the LIML Estimator with Possibly Many Instruments (2010, background): “Donald and Newey (2001) (in their Lemma A.6) has 9 investigated the asymptotic properties of the LIML estimator when K2n/n −→ 0.”
Instrumental variable estimation based on grouped data (2005, background): “DONALD and NEWEY (2001), who choose the instrument set by minimizing mean-square error criteria, consider a case where the number of instruments grows with, but at a smaller rate than the sample size.”
A regularization approach to the many instruments problem (2012, background): “Finally, Donald and Newey (2001) propose to select the number of instruments, L, that minimizes the mean square error (MSE) of the estimator.”
The Hausman Test and Weak Instruments (2011, background): “�2 ";z in ( 6 ): It turns out that H b �2";z � ; which is comparable to H3, has desirable asymptotic”
Further results on projection-based inference in IV regressions with weak, collinear or missing instruments ∗ (2007, background): “For other results relevant to the instrument selection, the reader may consult Cragg and Donald (1993), Hall et al. (1996), Shea (1997), Chao and Swanson (2000), Donald and Newey (2001) , Hall and Peixe (2003), Hahn and Hausman (2002a, 2002b), and Stock and Yogo (2002).”
Estimation and Testing of Forecast Rationality under Flexible Loss (2005, methodology): “Outside this situation, one could attempt to use data-based methods for selection of moment conditions using criteria such as those proposed by Donald and Newey (2001), replacing their MSE loss with our loss L in (1) evaluated at ( p 0 , ˆ α T ), wherê α T is a consistent estimate of α 0 .”