Improving Precision of Benefit Measures with Ridge Estimators: An Application to Economic Values of Air Quality
2005/06/07
Date : 2005-06-07
Time : 15:00
Speaker : Prof. Huang, Ju-Chin
Venue : B棟110室
Background : Prof. Huang, Ju-Chin 為Ph.D. in Economics and Statistics，North Carolina State University (1994)。
現為Associate Professor，University of New Hampshire。
其主要研究領域為Environmental Valuation及Applied Econometrics。
Abstract : This study examines alternative estimators to derive more precise benefit measures of air quality. Two ridge estimators, proposed respectively by Huang (Economics Letters, 1999) and Hoerl, Kennard, and Baldwin (HKB, Communications in Statistics, 1975), are employed and compared with the traditional OLS estimator. The advantage of the ridge estimators is that they are potentially more precise. Furthermore, they are easy to compute based on standard OLS results. In this paper, formulae to recover the Huang and HKB estimates and their MSEs from OLS results are provided. The marginal effects of air pollution on property values, presented in sixteen published property value/air pollution studies, are re-computed using the ridge estimators. Results show that for the coefficient of interest, namely the marginal effect of air pollution, the ridge estimates are superior to the OLS estimates under the mean squared error criterion in all sixteen studies. This study demonstrates that without conducting new estimation or collecting new data, the existing benefit measures of environmental quality can be re-presented with higher precision.
