Semiparametric Stochastic Frontier Estimation Using Generalized Additive Models

Abstract

This article specified a semiparametric stochastic frontier function using generalized additive models that accounts for random noise in the sample data. We estimated the parameters of the model by applying the generalized spline-smoothing approach to measure technical efficiency scores of Wisconsin dairy producers between 1993 and 1998. Results showed that the sample dairy producers did not use resources efficiently, as the estimated mean technical efficiency score was found to be 0.778. Unlike precedent studies, we found no correlation between the estimated technical efficiency scores and four farm-specific characteristics, such as operation type, milk system, barn type, and milk frequency.