Abstract:
Comparisons of unit’s efficiencies operate under different technologies remains an important issue in economic studies. This paper tries to shed light on the sensitivity of empirical results to the selection of the estimation method. The primary avenue of comparison in this analysis will be to assess the sensitivity of technical efficiency predictions to the choice of estimation method. Real data was used to compare metafrontier estimation methods. Three methods are compared in this study: stochastic frontier using all data, stochastic metafrontier and Data Envelopment Analysis (DEA) metafrontier. Data used to compare methods were collected in the framework of the Project SYPROBIO (SYstèmes de PROduction BIOlogique diversifiés) during the agricultural campaign 2012–2013.To handle performance of MF, the percentage of underestimated firm’s efficiency was calculate using poled data SF, SMF and DEA metafrontier. The paired-t.test and Spearman’s rank order correlation were used to compare efficiency and TGR derived from the approaches. Results showed that the stochastic frontier using pooled data did not return the optimum output set. For both the stochastic metafrontier production function and the data envelopment analysis metafrontier, there are very large differences in the technical efficiencies. While the predicted technical efficiencies vary widely across farms for both estimations, the variations, across the two estimation methods, are significant. But no methods can be considered as significantly better than other. Difference between two methods depends only on linear programming function’s used to obtain metafrontier, technologies gap ratio and the fact that the DEA frontier is not stochastic.
Keywords: DEA metafrontier, stochastic metafrontier, technical efficiency, technologies gap ratio.
Language:
English
Date of publication:
2016
Country:
Region Focus:
West Africa
University/affiliation:
Collection:
RUFORUM Theses and Dissertations
Agris Subject Categories:
Additional keywords:
Access restriction:
Supervisor:
Romain Glele Kakai
Form:
Web resource
ISSN:
E_ISSN:
Edition:
Extent:
viii, 64