Maternal Health and Child Development Programs in the United States and Rwanda: An Evaluation of Policies to Improve Quality and Efficiency

Abstract

This dissertation evaluates national programs in the United States and Rwanda that aim to strengthen maternal care and promote healthy child development, which remain significantly policy challenges globally. Chapter one analyzes the impact of Rwanda’s national pay-for-performance (P4P) program on rewarded maternal and child health services, multitasking and health outcomes. In a difference-in-differences analysis using the Rwanda Demographic and Health Surveys, we find that P4P significantly increased the output of several rewarded maternal health services, but had no significant impact on other rewarded services or health outcomes. There was little evidence of multitasking. We also find mixed effects of P4P by baseline facility quality. Our results highlight the opportunities and challenges in implementing P4P in resource-limited settings. Chapters two and three examine Head Start, the largest federally funded child development program in the US. In chapter two, using experimental data from the Head Start Impact Study (HSIS) I compare the effects of Head Start on child development relative to two distinct groups: children who received home care, and children who attended other non-Head Start centers. Relative to home care, I find that Head Start achieves larger gains across a range of developmental outcomes; relative to other centers, Head Start’s impacts are smaller and restricted to improved behavior and access to health care. Compared to either group, gains from Head Start largely dissipate within one year of leaving the program. These findings suggest that Head Start enhances child development in the short-term, particularly in comparison to home care – but also raise questions about longer-term impacts. Chapter three examines efficiency and equity trade-offs of targeting strategies that seek to increase Head Start’s average impact by prioritizing enrollment of children with larger expected treatment effects. I identify child and family characteristics that predict variations in Head Start’s impacts in the HSIS data, construct targeting strategies based on these characteristics, and simulate their effects on cumulative program impacts and equity measures. I find that prioritizing enrollment of 4 year-olds otherwise likely to receive home care achieves the largest efficiency gains, but increases racial disparities in access to Head Start. Further prioritizing the enrollment of children from the most disadvantaged households, however, can eliminate these disparities while maintaining efficiency gains.