The Effect of Child Savings Accounts on Family Material Hardship: Findings from a Randomized Trial in Uganda

Background: In developing countries, child savings accounts are increasingly promoted as a strategy for financial inclusion among low-income households.  Such accounts allow families to begin the process of long-term asset accumulation for their children and have demonstrated positive secondary outcomes such as increases in educational attainment, psycho-social functioning, and health behavior.  However, little is known about the consumption patterns of savers as they actively deposit into these accounts.   Of interest to researchers and policy makers, is whether families, particularly those living in resource limited settings such as rural Uganda, could be restricting immediate consumption for children in order to save.

Methods: Using data from the National Institute of Child and Human Development (NICHD) funded Bridges to the Future, multi-level regression models were used to assess if household participation in a formal savings program, specifically a Child Development Account (CDA), was significantly associated with caregiver restrictions on expenditures for basic needs, as measured by various indicators of material hardship among Ugandan children in grades 5 and 6 (N=1410) and compared across a control group and a two-armed intervention with differing incentives (1:1 and 2:1 matched savings, “Bridges” and “Bridges PLUS”).   

Results: Children in control and treatment groups did not differ in their likelihood of having existing savings at baseline.  However, children in Bridges (Odds Ratio (OR): 4.71, 95% CI: 2.93-7.58) and Bridges PLUS (OR: 8.28, CI: 5.17-13.25) were more likely than the control to have savings 24-months post intervention initiation. Households in the Bridges PLUS group, who were provided a 2:1 incentivized match rate, yielded higher average savings than children who received a 1:1 match.  Although children in both treatment groups demonstrated a higher probability to save when compared to the control, there was no indication that such savings behavior resulted in family material hardship.  Conversely, there were signs of decline in the likelihood of experiencing material hardships for children in the two treatment groups. Children in both treatment arms were less likely to report overall food scarcity, limited clothing, or lack of blankets. There was no difference in reported access to specific food items like meat, fish, and sugar, where access constraints to these specific consumption items were documented equally across the three study arms. In addition, families of children in the Bridges PLUS group were more likely to start a small business after the intervention (OR: 1.89, CI: 1.20-2.98) relative to children in the Bridges group or the control condition.

Conclusions:Gaining insight on the possible risks of micro-savings is important in order to avoid unintended and potentially detrimental consequences to household members, including children.  Our findings suggest that savers were not restricting expenditures on basic necessities but rather appear to be generating new resources to meet the material needs of children.  Such results offer support for child savings accounts in low resource settings as beneficial for initiating savings behavior and promoting financial inclusion while not compromising the material needs of vulnerable children.