A spatially explicit approach for targeting resource-poor smallholders to improve their participation in agribusiness: a case of nyando and vihiga county in Western Kenya

The majority of smallholder farmers in Sub-Saharan Africa face myriad challenges to participating in agribusiness markets. However, how the spatially explicit factors interact to influence household decision choices at the local level is not well understood. This paper’s objective is to identify, map, and analyze spatial dependency and heterogeneity in factors that impede poor smallholders from participating in agribusiness markets. Using the researcher-administered survey questionnaires, we collected geo-referenced data from 392 households in Western Kenya. We used three spatial geostatistics methods in Geographic Information System to analyze data—Global Moran’s I, Cluster and Outliers Analysis, and geographically weighted regression. Results show that factors impeding smallholder farmers exhibited local spatial autocorrelation that was linked to the local context. We identified distinct local spatial clusters (hot spots and cold spots clusters) that were spatially and statistically significant. Results affirm that spatially explicit factors play a crucial role in influencing the farming decisions of smallholder households. The paper has demonstrated that geospatial analysis using geographically disaggregated data and methods could help in the identification of resource-poor households and neighborhoods. To improve poor smallholders’ participation in agribusiness, we recommend policymakers to design spatially targeted interventions that are embedded in the local context and informed by locally expressed needs.