1st International Congress on Environmental Modelling and Software - Lugano, Switzerland - June 2002
Keywords
translog production function, nonparametric specification test, japanese manufacturing indutry
Start Date
1-7-2002 12:00 AM
Abstract
Beginning with a series of pioneering work by Cobb and Douglas, estimation of productionfunction has been one of the main issues in empirical economics and econometrics in various aspects. Thefunctional form they used is called Cobb-Douglas production function. It is later generalized to so-calledtranslog production function which has a more flexible form to describe the relation between the output andinput levels. It is also analytically convenient for obtaining factor demand functions or cost function so that ithas been widely used in both macroeconomic and microeconomic empirical studies. However, it is wellknown that statistical analysis leads to incorrect conclusions in general if the specified parametric model iswrong. This paper tests for the translog specification of production function for Japanese manufacturingindustry. We apply the nonparametric misspecification test by Hong and White [1995]. It is a consistent testhaving nontrivial power against local alternatives. We test the null hypothesis of translog specification usingcross section data of Japanese manufacturing firms listed in Tokyo Stock Exchange market. We also test theCobb-Douglas specification by the same method.
Nonparametric Test for Translog Specification of Production Function in Japanese Manufacturing Industry
Beginning with a series of pioneering work by Cobb and Douglas, estimation of productionfunction has been one of the main issues in empirical economics and econometrics in various aspects. Thefunctional form they used is called Cobb-Douglas production function. It is later generalized to so-calledtranslog production function which has a more flexible form to describe the relation between the output andinput levels. It is also analytically convenient for obtaining factor demand functions or cost function so that ithas been widely used in both macroeconomic and microeconomic empirical studies. However, it is wellknown that statistical analysis leads to incorrect conclusions in general if the specified parametric model iswrong. This paper tests for the translog specification of production function for Japanese manufacturingindustry. We apply the nonparametric misspecification test by Hong and White [1995]. It is a consistent testhaving nontrivial power against local alternatives. We test the null hypothesis of translog specification usingcross section data of Japanese manufacturing firms listed in Tokyo Stock Exchange market. We also test theCobb-Douglas specification by the same method.
