Human capital is widely seen as critical to growth and poverty alleviation. School quality affects human capital both directly and indirectly, by affecting the return to parental investments in education, and the incentive to invest. This project examines two strategies for improving education that have recently attracted considerable attention: providing school choice, and improving preparation for school through preschool programs. An innovative feature of both studies is their reliance on random assignment to compare treatment and control groups, and the fact that they are set in less developed countries. The first project examines a voucher program designed to allow poor students to attend private secondary schools in Colombia, and the second project looks at an early childhood education program that provides training and attendance incentives for teachers in locally organized preschools in Kenya. These programs may affect educational outcomes through their impact on decisions by parents and educators. The project will examine how the program affects: 1) educators' decisions regarding pedagogical techniques and school organization; 2) household decisions, such as assignment of household tasks to children and monetary investments in schooling; and 3) educational outcomes, such as enrollment, attendance, grade repetition, and test scores. The projects also examine whether the effects differ by students' sex, socio-economic status, and academic performance. Both evaluations take advantage of random assignment to address the issues of endogenous program participation that plague many analyses of educational programs. The voucher program in Colombia will be assessed by comparing outcomes between voucher applicants who won and lost a lottery used to allocate the vouchers. The project in Kenya was designed as a prospective randomized evaluation, and we hope it will set a precedent for collaboration between researchers and nonprofit organizations in conducting prospective randomized studies.
