Valid Confidence Regions and Inference in the Presence of Weak Instruments
We investigate confidence intervals and inference for the instrumental variables model with weak instruments. Wald-based confidence intervals perform poorly in that the probability they reject the null is far greater than their nominal size. In the worst case, Wald-based confidence intervals always exclude the true paremeter value. Confidence intervals based on the LM, LR, and Anderson-Rubin statistics perform far better than the Wald. The Anderson-Rubin statistic always has the correct size, but LM and LR statistics have somewhat greater power. Performance of the LM and LR statistics is improved by a degrees-of-freedom correction in the overidentified ccase. We show that the practice of "pre- testing" by looking at the significance of the first -stage regression leads to extremely poor results when the instruments are very weak.
Year of publication: |
1996-02
|
---|---|
Authors: | Zivot, Eric ; Nelson, Charles R. ; Startz, Richard |
Institutions: | Department of Economics, University of Washington |
Saved in:
Saved in favorites
Similar items by person
-
Valid Confidence Regions and Inference in the Presence of Weak Instruments
Zivot, Eric, (1996)
-
Improved Inference for the Instrumental Variable Estimator
Startz, Richard, (1999)
-
Improved Inference for the Instrumental Variable Estimator
Startz, Richard, (1999)
- More ...