Artificial neural networks approach to examine the impact of agri-environmental subsidies on agriculture eco-efficiency: the EU case study
Date Issued | Volume | Issue | Start Page | End Page |
---|---|---|---|---|
2024 | 10 | 1 | 15 | 15 |
The agricultural eco-efficiency is one property that may summarize the economic and environmental factors of agriculture when we aim at producing larger quantity of higher quality products, using fewer resources, and polluting less the environment. The agri- environmental subsidies (AES) are regarded as being a significant impact tool for encouraging sustainable farming and simultaneously addressing environmental issues arising from the agricultural activities. The aim of this research is to examine the impact of AES on the eco- efficiency of arable farms in the EU member states over the period 2017 - 2021. Two-stage analysis was implemented. At the first stage the eco-efficiency of farms was computed by using Data Envelopment Analysis. At the second stage additional data with new variables were used for also studying their impacts on the eco-efficiency. The most common approach for such analysis adopts regression methods. The present research applies artificial neural networks instead. After standardising original variable values for using them in the similar scales, our neural network models were distinctly better than the corresponding regression models. In this context we noticed that the AES was the most important factor influencing on the eco- efficiency scores of the arable farms in the EU member states. Comparisons of AES impact between different eco-efficiency groups revealed that AES was still the most influencial factor in lower efficiency group, whereas in the high and full-efficient groups the most influencial factors were subsidies on investment in agriculture. There were also identified significant differences in factors affecting eco-efficiency between those two groups. This study shows that the importance of AES in achieving sustainable agricultural development should not be underestimated in both supporting management decisions and CAP policy planning.