The Asymptotic Properties of Estimates of the Parameters of a Single Equation in a Complete System of Stochastic Equations

By Theodore W. Anderson, H. Rubin. 1950

Subfields: Economics & Operations Research

How researchers see this paper

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

In Economics & Operations Research

1. A. Rosen
2. M. Arellano
3. G. Bascle
4. Albert Maydeu-Olivares
5. Yves Rosseel
6. Dexin Shi
7. Bo E. Honoré
8. Luojia Hu

Citations in the corpus

• Concrete Problems in AI Safety (2016, methodology): “ 69]. The econometrics literature has in fact developed a large family of tools for handling partial specication, which also includes limitedinformation maximum likelihood and instrumental variables [10, 11, 133, 132]. Returning to machine learning, the method of moments has recently seen a great de”
• Weak Identification of Forward-looking Models in Monetary Economics (2004, unknown): “When p > 1, the hypothesis of under-identification can be tested using the reduced rank regression technique, developed by Anderson and Rubin (1950).”
• Bayesian Model Averaging in the Instrumental Variable Regression Model (2012, unknown): “In practice, researchers typically try different specifications until a set of restrictions (i.e. a particular choice of instruments, exogenous and endogenous regressors) passes a battery of misspecification tests (e.g. Anderson and Rubin, 1949, 1950; Hausman, 1983; Sargan, 1958).”
• The Impact of Microcredit on the Poor in Bangladesh: Revisiting the Evidence (2014, unknown): “…normal errors, but is consistent under substantial violations of that assumption: errors need not be normal, and they need only be uncorrelated with the instruments, not independent (Anderson & Rubin, 1950).9 The nonlinearities in the PK estimator turn out to make it less robust to such violations.”
• Controlling for endogeneity with instrumental variables in strategic management research (2008, methodology): “The LIML estimation, pioneered by Anderson and Rubin (1949, 1950), is an IV method based on a maximum likelihood estimator (for more details, see Greene, 2003; Murray, 2006b).21 The advantage of LIML estimation is that it is median unbiased: the median of its sampling distribution is generally…”
• On the Asymptotic Optimality of the LIML Estimator with Possibly Many Instruments (2010, background): “The LIML and TSLS estimators and their properties in the general case were originally developed by Anderson and Rubin (1949, 1950).”
• On the Performance of Some Robust Instrumental Variables Estimators (2004, methodology): “This basic approach was proposed by Andersen and Rubin (1950), and the LIML estimator can be interpreted as minimizing such a test statistic.”