Dynamic incentives for sustainable contract farming

Abstract

The rise of contract farming has transformed millions of farmers' lives. We study a new class of contract farming problems, where the farmer holds superior information and can invest effort to improve productivity over time. Despite their prevalence, the literature offers little guidance on how to manage such farmers with dynamic incentives. We build a game‐theoretic model that captures the dynamic incentives of learning and gaming, with hidden action and information. We characterize the optimal contract: it internalizes both the vertical and intertemporal externalities, with performance pay and deferred payment; the performance pay is to motivate the farmer to invest and improve the relationship‐specific productivity; the deferred payment is to ensure that the farmer is willing to share information and behave honestly over time. Even with random yield, the optimal contract can still have a simple implementation of a yield‐adjusted revenue‐sharing policy. Using real data, we show that the learning effect is significant. We then quantify when and how contract farming can improve smallholder farmers' productivity and income, creating shared value. We find when buyers have a long‐term perspective and can internalize the benefit of farmer improvement, they will pay higher prices to ensure farmers' long‐term viability. Our results inform the policy debate on contract farming: traditional procompetitive policies (based on spot transactions) can be counterproductive for modern agrifood value chains, hurting both buyers and farmers.