UC Berkeley

Education is an important input to human capital accumulation that has substantial positive effects on economic, health, and social outcomes. While large gains have been made in improving access to education in low-income countries in recent decades, there are several key areas for improvement. First, while primary school enrollment has improved, secondary school enrollment and completion continues to lag behind. Second, low-income and vulnerable households still face access barriers to educational resources. Third, disparities in educational access have worsened due to the recent worldwide school closures associated with the Covid-19 pandemic. In my dissertation, I study these areas by conducting field experiments in low-income settings. I shed light on access challenges in three different settings: secondary school transition, early childhood education, and Covid-19 related school closures.

My first chapter is titled "Information, Student-Parent Communication, and Secondary School Choice: Experimental Evidence from Kenya" and examines challenges associated with the transition from primary school to secondary school, and secondary school completion in Kenya. Secondary school dropout rates are high in low-income countries, and information gaps about school characteristics may be an important contributing factor. If school choices are made with imperfect information, households may choose schools that are too expensive, not a good fit academically, or too costly to commute to, increasing the likelihood of the students dropping out. These information gaps may be further exacerbated when students and parents fail to communicate before making high stakes schooling decisions. I study the importance of these information and communication gaps in the transition from primary to secondary school using a field experiment with 3,000 Kenyan students and their parents. The intervention consisted of an informational meeting for 8th graders before they applied to secondary school, and randomly varied whether the parent participated in the meeting for a facilitated conversation with the student. I find that informational meetings with students led them to apply to more commutable schools without compromising school quality. Moreover, including the parents in these meetings improved parental knowledge about costs and led to better alignment of school preferences between the students and their parents. This ultimately led students to enroll in lower cost schools, generating meaningful savings, particularly for low income households.

My second chapter, "Promoting Child Reading in Kenya: Estimating the Demand for Storybooks" is coauthored with Edward Miguel, Joan Hamory, and Michael Walker, and examines demand for early childhood reading materials in Kenya. Reading with young children is believed to be an especially important investment in human capital; however, many households in western Kenya lack reading materials for young children, and are unaware of the benefits of early reading. In this paper, we estimate the demand for storybooks amongst 1,000 Kenyan households with children aged 3 to 6 years old. The intervention involved highlighting the benefits of early childhood reading to parents and offering them the chance to buy children’s reading books at a subsidized price, where the subsidized price was randomized across three levels. We found that demand for reading materials was very high (97% overall), and downward sloping with price. The willingness to pay for storybooks was higher in urban areas, but did not differ by household income levels or gender of child. These findings highlight that lack of access to educational materials may play an important role in explaining low early-childhood human capital investment in Kenya.

Finally, my third chapter, titled "Learning from Home during the Covid-19 School Closures: Evidence from Kenya" examines how access to schooling in low-income contexts has been impacted during the Covid-19 pandemic, and the consequences for educational outcomes. The pandemic-related school closures have affected 1.5 billion students across almost 170 countries. Students in low-income countries were particularly impacted since schools were less likely to be able to shift to remote teaching technologies. I surveyed a sample of 3,000 Kenyan primary school students to measure student learning at home and parent involvement in their learning to shed light on differential effects of school closures by several demographic characteristics. I find that students from more educated and higher income households received significantly more help from parents with their schoolwork and were more likely to be exempted from chores so that they could study. However, there is no evidence suggesting differences in at-home learning environment across child gender. Taken together, my findings suggest that while gender gaps in schooling access may be closing at the primary school level, there may still be significant gaps in equality by socio-economic status that are particularly exacerbated during out of school periods such as Covid-19.