Public-sector maize research locations and spatial heterogeneity in maize productivity: Insights from four African countries on the roles of agroclimatic similarity

Agricultural research and development (R&D) is one component of public investments in the agricultural sector toward food system transformation. Enhancing the effectiveness of agricultural R&D remains critical, given increasingly scarce public resources. Exploring spatial spillover potentials has been one way to enhance the effectiveness of agricultural R&D. Geographical locations of National Agricultural Research Systems (NARS) research activities are recognized as an important factor affecting such spatial spillover potentials. However, evidence is generally limited in Africa south of Sahara (SSA) as to the spillover potentials of NARS-developed technologies. This paper partly aims to fill this knowledge gap by obtaining insights for maize, one of the most commonly grown crops in SSA, using nationally representative farm household data for Ethiopia, Ghana, Nigeria, and Uganda, and spatial agroclimatic data. Building on recent literature, this study proxies spillover potentials by “agroclimatic similarity†(AS) indicators between locations where agricultural R&D for maize is conducted by NARS (research locations) and where each farm household is located (farm locations). Results of the analyses suggest that an indicator of the total factor productivity of maize growing farm households, the land productivity of maize, and the use of improved maize varieties are generally higher in farm locations that share similar agroclimatic conditions with maize research locations of NARS. These patterns hold for all four countries studied, even after controlling for the physical proximity to maize research stations and other farm household characteristics. The findings contribute to better understanding of how geographic locations of public investments affect their overall effectiveness as well as returns in maize production and the agricultural sector in general.