Jackknife instrumental variables estimation

✍ Scribed by J. D. Angrist; G. W. Imbens; A. B. Krueger

• Book ID: 101285650
• Publisher: John Wiley and Sons
• Year: 1999
• Tongue: English
• Weight: 119 KB
• Volume: 14
• Category: Article
• ISSN: 0883-7252
• DOI: 10.1002/(sici)1099-1255(199901/02)14:1<57::aid-jae501>3.0.co;2-g

✦ Synopsis

Two-stage-least-squares (2SLS) estimates are biased towards the probability limit of OLS estimates. This bias grows with the degree of over-identi®cation and can generate highly misleading results. In this paper we propose two simple alternatives to 2SLS and limited-information-maximum-likelihood (LIML) estimators for models with more instruments than endogenous regressors. These estimators can be interpreted as instrumental variables procedures using an instrument that is independent of disturbances even in ®nite samples. Independence is achieved by using a `leave-one-out' jackknife-type ®tted value in place of the usual ®rst stage equation. The new estimators are ®rst order equivalent to 2SLS but with ®nite-sample properties superior, in terms of bias and coverage rate of con®dence intervals, compared to those of 2SLS and similar to those of LIML, when there are many instruments.