Exploring the dynamics of farms’ economic and environmental performance

Abstract

This paper presents a novel approach to analyse the dynamics of farm economic and environmental performance employing latent Markov (LM) models. The approach is grounded in a conceptual framework linking farm performance dynamics to structural constraints, farmers’ adaptive behaviour and sustainability transitions. LM models improve on traditional methods—such as static typologies, dynamic panels, and stochastic frontier analyses—by explicitly accounting for path-dependent adjustments shaped by both external factors and internal characteristics. The method is applied to Spanish rainfed field crop farms, using economic and environmental performance indicators from the Spanish Farm Accountancy Data Network for 2016–2021. Results reveal substantial heterogeneity in performance patterns and show that, while many farms maintained consistent performance year-on-year, a significant share transitioned between performance states. Although the environmental indicators are proxies, limiting the scope of policy inference, the findings demonstrate the capacity of LM models to capture performance dynamics in heterogeneous agricultural systems.