A Note On Numerical Estimation Of Sato's Two-level CES Production Function

A production function is the technical relationship between the output and the inputs used for turning out the produce manufactured by an efficient firm. A production function may take on various forms. The popular Cobb-Douglas production function assumes unitary and fixed elasticity of substitution between inputs and therefore is very unrealistic. The CES production function is more realistic, but, in view of the uzawa-McFadden impossibility theorems, it cannot be legitimately used to incorporate more than inputs. Sato's two-level (or multi-level) functions can suitably be used to circumvent these difficulties. However, Sato's functions are highly nonlinear and require very efficient methods of nonlinear estimation. This paper suggests such a method, which is based on evolutionary computation meta-heuristics such as the Particle Swarm and Differential Evolution optimization, and demonstrates that it works very well to fit the Sato's function.

View article page

Download PDF