Beliefs and Decision-Making in Education

Abstract

This dissertation consists of three essays that examine beliefs and decision-making in education. The first essay examines how a short-term preschool subsidy in rural Indonesia affects parents' beliefs and investments in preschool education. Using a difference-in-differences design, I show that the short-term subsidy had long-term effects on raising preschool enrollments but no significant impacts on preschool fees. I conduct a survey experiment to test different demand-side mechanisms. The results are most consistent with a learning effect model, where parents update their beliefs about the benefits of preschool from increased access to them as a result of the short-term subsidy. The second paper (co-authored with Felipe Barrera-Osorio, Paul Gertler, and Harry Patrinos) examines the impact of a national parental involvement program in Mexico. Using data from two randomized control trials, we separately estimate the effect of offering grants and information to parent associations. While grants to parent associations did not improve educational outcomes, information to parent associations reduced disciplinary actions in schools -- mainly by increasing parental involvement in schools and changing parenting behavior at home. We show that the divergent results from grants and information are partly explained by significant changes in perceptions of trust between parents and teachers. The third paper explores the mental models that policymakers use when integrating research evidence in their decisions, with a focus on those working in state and local education agencies in the United States. First, I examine policymakers' preferences for research evidence. Using a discrete choice experiment, I present policymakers with a series of research studies that vary along attributes of internal and external validity. I find that policymakers have preferences for larger studies and studies conducted in similar contexts as their own jurisdiction. However, they do not have a preference between experimental and observational studies. Second, I explore how much policymakers update their beliefs about the effectiveness of education policies using an information experiment. I show that policymakers update their beliefs in response to research evidence, but these effects are large and persistent only for those who were presented a brief, accessible explanation of how the evidence was generated.