Files
Abstract
The issue of production function estimation has received recent attention, particularly in agricultural economics with the advent of precision farming. Yet, the evidence to date is far from unanimous on the
proper form of the production function. This paper reexamines the use of
the primal production function framework using nonparametric regression
techniques. Specifically, the paper demonstrates how a nonparametric regression based on a kernel density estimator can be used to estimate a production function using data on corn production from Illinois and Indiana.
Nonparametric results are compared to common parametric specifications
using the Nadaraya-Watson kernel regression estimator. The parametric
and nonparametric forms are also compared in terms of describing the
true technology of the firm by obtaining measures of the elasticity of
scale and the marginal physical product through nonparametric estimation of the gradient of the production surface. Finally, the elasticities of
substitution are compared between both parametric and nonparametric
representations.