Using Farm Accountancy Data To Estimate Crop Rotation Effects
Résumé
Crop rotations are widely perceived as a key feature of sustainable crop systems. However real crop rotations made by farmers and their economic determinants are poorly documented. This article presents an original econometric approach to estimate them based on farm accountancy data, where farmers’ crop sequence choices are not observed. We develop a Bi-Level Programming model designed to estimate crop rotation effects on yields and input uses while simultaneously reconstructing farmers’ unobserved crop sequences. Our approach solves indeterminacy issues assuming that farmers are economically rational and using expert knowledge information. We apply our innovative approach on French data. Our statistical results show that crop rotations are beneficial in terms of reducing chemical input uses and improving crop yields. Our results also reveal a strong inertia in crop rotation sequences involving predominant crops. Accordingly significant long term economic incentives are needed to reap the potential environmental benefits provided by crop rotations.