Sensitivity of the GME Estimates to Support Bounds
Preprint
- 1 August 2001
- preprint
- Published by Elsevier BV in SSRN Electronic Journal
Abstract
The claim has been made that the Generalized Maximum Entropy (GME) estimator of Golan, Judge and Miller is not sensitive to variations in the support bounds of either the parameters or the error terms. In this paper, we scrutinized this claim by means of Monte Carlo experiments and found that the parameter estimates are impacted in a substantial way by these changes. We also analyzed the famous data sample on the US manufacturing industry used by Cobb and Douglas in 1934 and found that the GME estimator is very sensitive to changes in support bounds. We conclude with a general result by Caputo and Paris according to which any support bound variation produces unexpected responses in the parameter estimates.Keywords
This publication has 9 references indexed in Scilit:
- Comparative statics of the generalized maximum entropy estimator of the general linear modelEuropean Journal of Operational Research, 2008
- Estimating Crop‐Specific Production Technologies in Chinese Agriculture: A Generalized Maximum Entropy ApproachAmerican Journal of Agricultural Economics, 2001
- Modeling Land Use Decisions with Aggregate DataAmerican Journal of Agricultural Economics, 1999
- Recovering Output‐Specific Inputs from Aggregate Input Data: A Generalized Cross‐Entropy ApproachAmerican Journal of Agricultural Economics, 1998
- An Analysis of Ill‐Posed Production Problems Using Maximum EntropyAmerican Journal of Agricultural Economics, 1998
- Estimation of multi-output production functions with incomplete data: A generalised maximum entropy approachEuropean Review of Agricultural Economics, 1998
- Field Notebook: USA 1998bPublished by Smithsonian Institution ,1998
- THE MAXIMUM ENTROPY APPROACH TO ESTIMATION AND INFERENCEAdvances in Econometrics, 1997
- GAMS, a user's guideACM SIGNUM Newsletter, 1988