Investigating the dynamic eco-efficiency in agriculture sector of the European Union countries

One of the main aims of European Common Agricultural Policy (CAP) has been established on a considerable reduction of greenhouse gas (GHG) emissions without any decline in productivity by 2030. This introduces some challenges in the assessment of agricultural technical efficiency. In particular, we can formulate at least two problems that are worth to be addressed: (i) possibility of treating undesirable outputs and (ii) panel data analysis of agricultural technical efficiency enriched by eco-efficiency assessment. In this study, we applied a Window Slack-Based Measurement Data Envelopment Analysis (W-SBM-DEA) model in the presence of undesirable outputs to evaluate the performance of agriculture sector of EU-27 countries during the period from 2008 to 2017. The country-level cultivated agricultural area, labor, specific costs, overheads, and depreciation were considered as the model inputs, and the gross value of country-level crop and livestock products represented the desirable outputs. The agricultural total GHG emissions at the country-level were accounted for as the undesirable output in the model. Having compared the results of different models, the substantial difference in the performance of the EU countries was registered mainly due to incorporating undesirable outputs and window analysis when using SBM-DEA model. The results of eco-efficiency assessment indicated that the Netherlands, Belgium, Italy, and Malta are the most eco-efficient countries with a score higher than 0.90, while the lowest eco-efficiency scores were reported for Slovakia, Latvia, and Estonia. The highest average eco-efficiency scores of all EU members were registered in 2011 (0.82), 2012 (0.83), and 2017 (0.84). A comparison of eco-efficiency performance between the old and new EU members indicated that the greater scores were obtained by old EU members. The variability assessment results showed a low variability and subsequently a high stability in the European agricultural sector, particularly in the Netherlands, Italy, and Malta. Based on our findings, it was concluded that for assessing agricultural technical efficiency in European agriculture in the context of eco-efficiency, application of a W-DEA model with undesirable output for a specified period of time reduces the impacts of temporary changes and provides more realistic results when comparing to models without undesirable output. These more realistic assessments of technical efficiency could help policy-makers to make more precise decisions.