Performance of a Novel AI-Based Proxy Means Test in Comparison to Other Methods for Targeting Pro-Poor Water Subsidies in Ghana

Access to safe drinking water is still very low among the poorest households in sub-Saharan Africa, and economic shocks can make water access even more difficult for poor consumers. Water subsidies can be a solution to enhance access to safe water services, but they are often ineffective as they regularly fail to reach the very poor. In this study, we developed a new Artificial Intelligence-based proxy means test (AI-based PMT) and assessed whether it performed better than four other methods (the Demographic and Health Survey (DHS) wealth index, the Poverty Probability Index (PPI), Community Based Targeting (CBT) and Ghana’s government LEAP (Livelihood Empowerment Against Poverty program) at identifying the poorest households.We first developed our new AI-based PMT by applying machine learning techniques to the nationally-representative 2016-2017 Ghana Living Standards Survey (GLSS7) and compared its performance with an already existing PMT (the PPI). We then evaluated this new method in rural towns of southwestern Ghana against four other approaches, with respect to their performance, their scalability (ease of implementation and affordability), and their acceptability to local stakeholders.In our field-evaluation we found that our new AI-based PMT performed better than the other approaches (e.g., it led to fewer inclusion errors) but it had higher implementation costs than CBT and LEAP. Local government considered CBT to be more transparent than the PMTs, while community members found the PMTs fairer.By highlighting the strengths and the weaknesses of five different targeting methods, this study provides guidance to practitioners in choosing the most appropriate method to target poor households eligible for water subsidies in rural Ghana