Estimation and Testing of Regression Coefficients in Cobb-Douglas and Other Models
The Cobb-Douglas production function often leads to situations in which the reduced form solutions for the inputs have elasticities of each input with respect to a given exogenous variable which are the same for all reduced form equations. This paper exploits this property to develop estimators that are in the spirit of full information maximum likehood (FIML) and to develop test statistics that are of the Student sort, provided that the reduced form errors have normal properties often used for analysis of this sort.
KeywordsCovariance Income Assure Monopoly
Unable to display preview. Download preview PDF.
- Anderson T.W. , An Introduction to Multivariate Statistical Analysis (Wiley: New York),section 5.22, pp. 105–107Google Scholar
- Laitinen, Kenneth , “Why Is Demand Homogeneity So Often Rejected”,Economics Letters,pp. 187–191.Google Scholar
- Malinvaud, Edmund , Statistical Methods of Econometrics (North Holland - Amsterdam, New York, Oxford), Third Ed. pp 195–200 and pp. 230–231.Google Scholar
- McCarthy, Michael D. [ 1987 ], “Exact. Finite Sample Test Statistics for Restricted Reduced Forms: A Generalization of Student’s Result,” IER, pp. 259–269.Google Scholar
- Sargan, J.D., “Asymptotic Theory and Large Models, ”IER, pp. 75–91.Google Scholar