The accumulation of human capital has consistently been identified as a key determinant of economic growth and policies designed to expand access to schooling and improve the quality the education are top priorities in many countries. These public sector investments, however, must by made with an understanding of the constraints that households face as they make decisions about human capital investments in their children. The proposed research will inform these issues through three interrelated analyses of schooling choices and outcomes among children and adolescents in a low income setting. The first analysis will estimate the determinants of school enrollment and time use, and school choice, identifying the effect of individual characteristics, family background, local labor market conditions, and school availability and quality. The second component will estimate the determinants of age at school entry, grade repetition, and school attainment in a dynamic framework, with prospective longitudinal data on school behaviors. The third analysis will investigate the role of school quality, controlling for individual characteristics and family background, on school achievement, using both individual- and school-level data. Together these analyses will provide a richer picture of the determinants of early human capital formation and the role of public policies in enhancing education investments. In each analysis, we are particularly interested in the role of family background characteristics such as parental education, household economic resources, and measures of school access and quality. Since schooling choices cannot be made in isolation of the competing uses of children's time, we will explicitly examine the impact of local labor market opportunities on enrollment decisions and schooling outcomes. We will also investigate the effect of school access and school quality, at both the current grade level and future grade levels, on schooling decisions. The availability of data on siblings at a point in time and on individual children through time will allow us to control for both measured and unmeasured family and individual characteristics. Our analyses will employ appropriate statistical techniques to account for the possibility that public sector investments are non-randomly targeted at the community level of that households migrate in response to community-level infrastructure. These analyses are made possible by the extremely rich retrospective and longitudinal household, community, and school facility data available in the first and second Indonesian Family Life Surveys, as well as several other sources of individual- and community-level data in Indonesia.
Showing the most recent 10 out of 12 publications
