Optimizing Public and Private Seed Systems: A Marginal Treatment Effect Analysis for Targeting Biofortified Bean Diffusion in Zimbabwe

ABSTRACT

Understanding the full benefits of scaling the production and consumption of biofortified crops requires moving beyond estimates of average adoption to understand how returns vary across the heterogeneous population of growers of such varieties. We used a marginal treatment effect framework to optimize seed system targeting by analyzing the heterogeneous marginal benefits of adopting NUA45, biofortified beans in Zimbabwe. Using nationally representative data, we move beyond the population average to estimate the marginal returns for subgroups of farm households differentiated by their observed characteristics and latent resistance to adoption. Results indicate that scaling NUA45 generates positive returns for all marginal adopters, with benefits 1.3 percentage points higher among actual adopters than non‐adopters. However, the analysis reveals that farm households with socioeconomic characteristics making them more likely to adopt, particularly those near the adoption threshold, gain disproportionately compared to those least likely to adopt. These findings provide an evidence‐based strategy for optimizing public and private seed systems, where targeted diffusion strategies can maximize social returns by prioritizing farmers with a moderate‐to‐high propensity to adopt. Meanwhile, households least likely to adopt may require complementary support to achieve an equitable impact. This marginal treatment effect approach provides a novel blueprint for efficiently targeting the diffusion of biofortified seeds to maximize both scale and welfare.