Multi-Directional Non-Parametric Analysis of Agricultural Efficiency: The Case of Lithuanian Family Farms

PhD defence

Tomas Balezentis


This thesis seeks to develop methodologies for assessment of agricultural efficiency and employ them to Lithuanian family farms. In particular, we focus on three particular objectives throughout the research:
(i) to perform a fully non-parametric analysis of efficiency effects,
(ii) to extend the Multi-Directional Efficiency Analysis approach,
(iii) to account for uncertainties via the use of probabilistic and fuzzy measures.
Therefore, the thesis encompass six papers dedicated to (the combinations of) these objectives.
One of the main contributions of this thesis is a number of extensions to the Multi-Directional Efficiency Analysis approach when the proposed models were employed to analyse empirical data of Lithuanian family farm performance, we saw substantial differences in efficiencies associated with different inputs. In particular, assets appeared to be the least efficiently used input relative to labour, intermediate consumption and land (in some cases land was not treated as a discretionary input). These findings call for further research on relationships among financial structure, investment decisions, and efficiency in Lithuanian family farms.

Application of different techniques yielded somewhat different patterns of the determinants of efficiency in Lithuanian family farms. In the conditional framework, the negative effects of production subsidies were evident for both radial and input-specific models. Asset use intensity had a positive effect upon land and labour use efficiency. As for analysis under the assumption of no separability among environmental effects, we observed a positive influence of farm size on efficiency as well as a negative influence of crop share in the total output.

The use of fuzzy clustering and fuzzy Free Disposal Hull enabled us to identify the degree of stochasticity associated with Lithuanian family farm performance. The former technique showed that the farms differed in terms of the mean values and variance of the efficiency scores over time with some clear patterns prevailing throughout the whole research period. The fuzzy Free Disposal Hull showed that the lower boundary of the fuzzy efficiency scores for livestock farms was more stable than those for other farming types. However, under increasing uncertainty, crop farms featured extremely high upper bounds of the fuzzy efficiencies.

The differences among the underlying technologies were analysed by the means of Multi-Directional Efficiency Analysis in the program efficiency framework, the bias-corrected Malmquist indices, and Malmquist indices based on Multi-Directional Efficiency Analysis. The proposed techniques, thus, can be successfully employed to disentangle the underlying patterns in efficiency and productivity change avoiding the loss of information due to the use of the radial measures.


Mette Asmild, Professor, IFRO, University of Copenhagen, Denmark

Assesment Committee

Arne Henningsen, Associate professor, IFRO, University of Copenhagen, Denmark

Niels Christian Petersen, Professor, University of Southern Denmark, Denmark

Alfons Oude Lansink, Professor, Alfons Oude Lansink, Wageningen University, Netherlands