The paper presents the non-parametric benchmarking technique, viz. Free Disposal Hull (FDH), along with its probabilistic extensions. Indeed, the frontier methods are rather sensitive to outliers. The probabilistic methodology, therefore, is useful when handling the practical problems of benchmarking.
The paper aimed at analysing the patterns of efficiency of the Lithuanian family farms with respect to the uncertain data. The latter aim was achieved by the virtue of the probabilistic production functions. The sensitivity of the efficiency scores estimated for the Lithuanian family farms was analysed by manipulating the numbers of randomly drawn benchmark observations estimations and thus constructing respective order-m frontiers. The livestock farms appeared to be most efficient, or even super-efficient, independently of the model orientation or the order of the frontier. The application of the order-alpha frontier did also confirm these results.
Keywords: efficiency; family farms; partial frontier; activity analysis.
Baležentis, T.; Baležentis, A. 2014. Probabilistic productive technology and partial production frontiers: an application to Lithuanian agriculture, Jaunųjų mokslininkų darbai = Journal of Young Scientists 1(41): 141-146. ISSN 1648-8776 [CEEOL nuo 2005, IndexCopernicus nuo 2008].
