Gauging the impact of Ethiopia’s productive safety net programme on agriculture: Application of targeted maximum likelihood estimation approach

The Productive Safety Net Programme (PSNP) is Ethiopia's poverty reduction strategy that forms the most important pillar of the country's agricultural transformation into a more productive and competitive sector. However, the extent to which the PSNP is linked to agriculture is unclear. This paper evaluates the impact of the PSNP on a range of agricultural outcomes. We use data from the Living Standard Measurement Study—Integrated Survey on Agriculture and apply the targeted maximum likelihood estimation method. We find no evidence that PSNP participation improved technology adoption, time spent in agriculture, household‐level access to agricultural services, or women's control over agricultural assets. However, PSNP participation increased access to credit, the share of non‐farm income, hours spent on casual work, community access to irrigation water, ownership of agricultural tools, community service in crop and livestock production, natural resource management, and access to credit. We also observe that PSNP households have a lower level of endowments compared to non‐PSNP households. Given the observed lack of impact on household‐level agricultural outcomes, we recommend integrating household‐level interventions, such as increasing the transfer size and provision of productive assets, that could lift household endowments above an asset threshold that would allow the productive use of community assets. This may boost the productive impact of the PSNP at the household level, facilitate agricultural development and economic growth. To generate additional insights, we recommend further research with sufficient data on the causal pathways between safety nets and agriculture.