THE IMPORTANCE OF FUNCTIONAL FORM IN THE ESTIMATION OF WELFARE

Abstract

Researchers have recognized the central role that the choice of functional form has on estimates of consumer surplus. The purpose of this paper is to quantify the magnitude of errors which might arise from the use of incorrect functional forms. It describes a simulation experiment where estimated consumer surplus, based on simulated data sets, is compared with consumer surplus computed directly from the simulated data. The errors resulting from the use of mismatching functional forms range from approximately 4% to 107%. For over two decades applied economists have been estimating environmental benefits using recreation demand models. The approaches used to estimate benefits have evolved from relatively simple, single-equation models of demand to more complex models incorporating, among other things, multiple sites, truncation problems, the opportunity cost of time, and environmental quality variables. Underlying this evolution has been a shift in emphasis from ad hoc specifications of the demand functions to demand functions which are consistent with the postulates of consumer theory. This paper explores the importance of one component of these models for the estimation of benefits: namely, the choice of functional form. Researchers have recognized the central role that the choice of functional form has on benefit estimates. However, the magnitude of error associated with incorrect functional form has not been measured. This paper presents a first attempt at quantifying these errors. To accomplish this objective, a simulation experiment is conducted wherein three sets of