A nonparametric extension of generalized quadratic Box-Cox models in measuring technical efficiency

This article uses a nonparametric extension of estimating Generalized Quadratic Box-Cox (GQBC) models using the Additivity and Variance Stabilization (AVAS) algorithm. The new method accounts for random noise in the data and relaxes the sensitivity of technical efficiency scores to the choice of functional form. It also provides more flexible choices for estimating the parameter of the dependent variable. The model is specified to measure technical efficiency scores of New York dairy producers in the period 1990 to 2000. Results show that the sample producers did not use resources efficiently, as the estimated mean technical efficiency score was found to be 0.663.