Improved JIVE Estimators for Overidentified Linear Models with and without Heteroskedasticity
作者： Paul J. Devereux , Daniel A. Ackerberg
DOI:
关键词: Statistics 、 Simple (abstract algebra) 、 Mathematics 、 Small sample 、 Jackknife resampling 、 Estimator 、 Instrumental variable 、 Linear model 、 Set (abstract data type) 、 Heteroscedasticity
摘要: We introduce two simple new variants of the Jackknife Instrumental Variables (JIVE) estimator for overidentified linear models and show that they are superior to existing JIVE estimator, significantly improving on its small sample bias properties. also compare our estimators Nagar (1959) type estimators. that, in with heteroskedasticity, have properties both related B2SLS suggested Donald Newey (2001). These theoretical results verified a set Monte-Carlo experiments then applied estimating returns schooling using actual data.
参考文章(5)
John C. Chao, Norman R. Swanson, Consistent Estimation with a Large Number of Weak Instruments Social Science Research Network. ,(2003)
Douglas O Staiger, James H Stock, None, Instrumental Variables Regression with Weak Instruments Social Science Research Network. ,(1994)
Michael J. Stutzer, Yuichi Kitamura, An Information-Theoretic Alternative to Generalized Method of Moments Estimation Social Science Research Network. ,(1998)
Whitney K. Newey, Whitney K. Newey, Stephen G. Donald, Choosing The Number Of Instruments Social Science Research Network. ,(1999)
Joshua D. Angrist, Joshua D. Angrist, Joshua D. Angrist, Guido W. Imbens, Alan B. Krueger, Alan B. Krueger, Alan B. Krueger, Jackknife Instrumental Variables Estimation Social Science Research Network. ,(1995)