In high-income countries, increases in educational attainment are strongly associated with a decline in mortality and morbidity, including obesity. The effect of increased education on obesity may, however, be strongly contextual and depend on the stage of the population in the nutritional transition. While initial increases in education in low- and middle-income countries may foster obesity, the link is expected to be negative thereafter. There is some evidence on a negative correlation between education and obesity in low- and middle income countries. Most of them are, however, based on simple correlations and do not establish causal effects. Hence, this study aims to investigate the causal effect of education on obesity in Indonesia.
We estimate the effect of years of schooling on being overweight/obese using high-quality individual-level data from the Indonesia Family Life Survey, which provides detailed information on education histories and health status. We start by estimating simple probit models. To account for potential endogeneity, we then use exposure to a substantial primary school construction program (SD INPRES) as an instrument for years of schooling. For this, the individual-level data is linked to district-level data capturing the intensity of the program across regions from 1973-1978. All models control for relevant individual characteristics and a regional specific confounders