From Cost Structure to ESG Sustainability : Structural and Cross-Country Evidence from the EU Livestock Sector

Abstract

The transition toward climate-neutral development in the European Union has intensified the need for structured and data-driven approaches to sustainability assessment in agriculture. Despite extensive research on environmental performance in livestock systems, empirical models explicitly linking managerial cost structure to ESG sustainability outcomes at a cross-country level remain limited. This study develops an integrated analytical framework to examine whether cost allocation patterns and energy structure systematically determine sustainability profiles in the EU livestock sector. The empirical analysis is based on a harmonized panel dataset covering Germany, France, the Netherlands, Spain, and Poland for the period 2000-2023. A multi-stage modelling approach combining k-means and fuzzy clustering, regression analysis, and Random Forest classification was implemented in MATLAB. The clustering stage identified three stable ESG structural profiles. Regression results indicate that energy share and environmental investment allocation significantly explain variation in GHG intensity (R² = 0.9173). Machine learning classification further confirms the structural importance of energy-related cost components in differentiating sustainability classes. The findings demonstrate that sustainability performance in livestock systems is not random but structurally embedded in managerial cost architecture. The study contributes to sustainability accounting by integrating ESG metrics with sector-level cost indicators and provides a scalable analytical framework for sustainability benchmarking and policy evaluation in the EU agri-food sector.