Improving Parenting Practices for Early Child Development: Experimental Evidence from Rwanda

Abstract

This paper investigates the short- and medium-term impact of a randomized group-based early child development program targeting parents of children aged 6–24 months in a poor, rural district of Rwanda. This low-intensity, short-duration, and low-cost program engaged parents through sessions that included a novel radio show and facilitated discussions during 17 weekly village-level meetings. The intervention included two treatment arms, with different components. Twelve months after baseline, children’s communication, problem-solving, and personal social skills improved in treated groups and persisted almost 3 years later in the full treatment arm. Positive effects on maternal time investments, attitudes, and beliefs, as well as investments in play resources, also persisted over time. A mediation analysis shows that the positive impact observed in child development can be attributed to positive changes in parental and home environment inputs, particularly in the longer term. This study offers important insights for the design and delivery at scale of early child development interventions among some of the most disadvantaged communities in the world.

Group Meetings and Boosters to Sustain Early Impacts on Child Development: Experimental Evidence from Kenya

“…We include these 25 dyads in our final sample, though results are equivalent if we exclude them from our analysis. This results in an overall three-year attrition rate of 18 percent, which is consistent with rates found in similar ECD programs in Sub-Saharan Africa (e.g., Justino et al 2022). Most of those families who had attrited at either survey wave had moved away, at least temporarily.…”

“…Finally, with or without boosters, we also demonstrate significant cost-effectiveness of the program using group-based delivery that was much cheaper than the one-on-one home visiting approaches that have been used previously. Our cost per child of $119 in 2020 US dollars terms is higher than the $76 per child of the First Steps intervention in Rwanda (Justino et al 2022), but our costs become comparable after adjusting for inflation and converting to purchasing power parity (PPP) terms. 20 Both studies suggest the potential feasibility of scaling group-based parenting interventions in contexts of weak institutions and low levels of human capital.…”

“…Our study demonstrates the sustained effectiveness of a low-cost and groupbased ECD parenting intervention two years after the end of the intervention in a low-resource, rural setting, which is rare in the literature. The closest study to ours from an LMIC setting is the First Steps parenting program from rural Rwanda (Justino et al 2022), which utilizes a radio show to communicate program messages to groups of parents, and finds that early impacts of 0.38 SD decline to 0.2 SD two years later. However, they assess child development using the ASQ III, a low-quality measure based on caregiver report.…”

“…Studies from high-income countries have found larger sustained gains in children's cognitive outcomes for lower-educated households (Carneiro et al 2023;Doyle 2020) and in socioemotional outcomes for girls versus boys (Heckman et al 2017). To our knowledge, only one study of sustained outcomes from a LMIC setting examined heterogenous effects and found no differential impacts across any child or household characteristic two years after the end of their program (Justino et al 2022).…”

Early Childhood Human Capital Formation at Scale

“…Third, we contribute to the growing literature studying at-scale delivery of ECD interventions in developing countries. Our experiment conducts and evaluates an ECD intervention at a scale larger than many of the existing studies (Attanasio et al, 2014;Yousafzai et al, 2014;Sylvia et al, 2020;Grantham-McGregor et al, 2020;Justino et al, 2022). 3 For example, Araujo et al (2021a) study a home-visiting intervention implemented by paid para-professionals in Peru that targeted over 67,000 children in rural areas.…”