Are agro‐clusters pro‐poor? Evidence from Ethiopia

Governments and development agencies increasingly promote agro‐clusters as a pathway to improving smallholder incomes and ensuring inclusive rural development through mitigating production and market risks. However, there is very limited empirical evidence to support this promise. We use a large farm household survey of about 4000 smallholder farmers in Ethiopia growing cereals like teff, maize, wheat, maltbarley and sesame to examine the relationship between agro‐clusters and smallholder welfare and poverty. Using instrumental variable estimators, we establish a positive association between agro‐clusters, household income and per capita income. Agro‐clusters are also shown to reduce poverty and poverty gaps. Our results are robust over different agro‐cluster proxies and alternative estimators, such as the augmented inverse probability weighting estimator. We also show that our findings are unlikely to be driven by omitted variable bias. Moving beyond average effects and in the interest of understanding heterogeneous effects, we use quantile regressions at different income levels. We find that agro‐clusters are associated with welfare gains for all households. However, the most significant gains are observed for the wealthier households. Despite this regressive association, our findings suggest that agro‐clusters may be useful in making farming more profitable with significant welfare implications.