On the economic value of the agronomic effects of crop diversification for farmers: estimation based on farm cost accounting data[Valeur économique des effets agronomiques de la diversification des cultures pour les agriculteurs : estimation à partir de données de comptabilité analytique des exploitations agricoles]

Despite many benefits provided by diversified cropping systems, there is a dearth of empirical evidence on the economic relevance of their effects, mainly due to lack of information on the dynamics of farmers' crop acreages. Our article contributes to fill this gap and, thereby, to shed light on a pair of apparently contradictory facts. European farmers tend to stick to specialized crop acreages despite agronomic experiments tending to show that crop diversification could reduce chemical input uses while maintaining or even enhancing arable crop yield levels We provide estimates of the effects of previous crops and crop acreage diversity on yield and chemical input use levels based on a sample of 769 arable crop producers covering the Marne département in France from 2008 to 2014. Our farm level dataset combines cost accounting data, information on crop sequences as well as detailed soil and weather data. Our estimation approach relies on yield functions and input use models defined as systems of simultaneous equations. These models feature farm specific random parameters for accounting for unobserved heterogeneity across farms and farmers as well as for accommodating input use endogeneity in the considered empirical crop yield functions. We estimate pre crop and crop acreage diversity effects for four major crops in the area. Pre crops effects on yields are estimated relatively accurately and are generally consistent with the rankings provided by crop production experts. Estimated pre crop effects on input uses are small and insignificant from a statistical viewpoint despite our large sample, suggesting that pre crops don't impact much chemical input requirements or/and that farmers tend to downplay these effects when deciding their chemical input use levels. Our results also show that crop acreage diversity positively impacts yield levels and tend to induce reductions in pesticide uses, herbicide uses in particular. Overall, our results demonstrate statistically significant though economically limited effects of pre crops and crop acreage diversity on crop gross margins. They also suggest that policy measures aimed to foster crop diversification are unlikely to significantly reduce chemical input uses on major crops if they are not supplemented by measures specifically aimed to reduce the uses of these inputs.