Wealth Effects of an Asset-Building Intervention Among Rural Households in Masindi, Uganda

Background and Purpose: Asset development is a key strategy to promote economic and social development in Sub-Saharan Africa (SSA). Research has found associations between asset ownership and household well-being. However, to date there has been little rigorous research on impacts of asset-building interventions for families in SSA in general and Uganda in particular. In this study, we analyze wealth outcomes of a matched savings intervention among rural households in Masindi, Uganda. Using propensity score optimal matching and matching estimators, we investigate the impact of the intervention on financial assets.

Methods: Data in this study are from AssetsAfrica, a research initiative designed to test asset-building innovations. The project used a quasi-experimental design, comparing across intervention and comparison villages. The study sample consisted of 393 individuals assigned to the intervention (n = 203) and comparison (n =190). The intervention was a structured, matched savings account offered to the intervention group for a three-year period. In addition, financial education, and asset management training was offered. Participants who successfully reached their savings goals received match funds at a 1:1 ratio. Productive assets (i.e., those that would generate income) including livestock and means of transportation, were eligible for match funds.  Data was collected pre and post the intervention.

The study uses propensity score optimal matching to address selection bias in the sample due to the quasi-experimental design.  We performed Hodges-Lehmann aligned rank test, (Hodges & Lehmann, 1962) regression of difference scores with covariance control (Rubin, 1979; Rosenbaum, 2002), and matching estimators (Abadie & Imbens, 2002) to estimate treatment effects. We used imbalance indexes (Haviland, et al., 2007; Guo; 2008) to check covariate imbalance before and after optimal matching

Results:  Both propensity score methods used in the study indicate a positive impact of the intervention on financial assets. The Hodges-Lehmann test indicates a statistically significant positive impact of the intervention on financial assets, that is, individuals in intervention had $39 more in financial assets than non-treated individuals. Matching estimators results indicate a statistically significant larger treatment effect on the treated. However, the impact of the intervention on ownership of productive assets is less conclusive.

Conclusions and Implications: Overall, results of this study show that asset-building interventions have potential utility as a policy solution for improving the economic well-being of poor households in Sub-Saharan Africa. As an incentive to encourage the poor to save, many asset-building interventions provide an opportunity for matched saving. Social workers can advocate for poor people for governments and international organizations to use savings accounts of individuals and families to channel resources for development directly to poor families. Because the resources reach the poor families directly, expenditures for overhead costs by nongovernmental organizations are eliminated or greatly reduced, which gives poor people more resources to improve their economic well-being.