The objective of this study was to evaluate the influence of technical and economic variables on the indices of technical efficiency of milk from Minas Gerais throughout distinct points of distribution of the efficiency indices by the technique of quantile regression. The technical efficiency indices were estimated based on a stochastic frontier model, using data from 875 milk producers in Minas Gerais state, Brazil, collected in 2005. The main results from production frontier showed the extensive use of the land factor. Overall, the variable percentage of lactating cows was the more relevant in explaining technical efficiency in all analyzed quantiles, whereas the percentage of household labor was important to explain only the lower levels of efficiency. Moreover, significant differences between the estimated coefficients of the quantiles were found in the study, which showed that the explanatory variables do not have the same impactation on increasing the efficiency at all points of distribution.
dairy farming; determinants of efficiency; stochastic frontier
