Three essays on schooling and health in Indonesia. Assessing the effects of family planning on fertility and of supply-side education programmes on BMI, schooling attainment, and wages

In 1969, Indonesia established a national family planning programme and total fertility has declined rapidly since but there is little consensus over the relative contribution of family planning to the observed decline. The first chapter constructs a new measure of family planning exposure to examine the role of family planning in reducing fertility. The causal effects of infant mortality is also examined based on a new instrumental variable, water supply and sanitation programme exposure, and that of schooling using father’s schooling as an instrument. The findings strongly indicate that family planning contributes to lower fertility together with reductions in infant deaths and improvements in women’s schooling, and that the effects of family planning and decreases in infant mortality are larger than that of schooling. In 2002, nearly one-in-ten men and more than one-in-five women in Indonesia were overweight and noncommunicable diseases had become the main cause of death but there exists no evidence on the causal effect of schooling on BMI for developing countries. The second chapter assesses whether more schooling causes healthier BMI in Indonesia by using two instrumental variables to capture exogenous variation in schooling. The first instrument takes advantage of the primary school construction programme (SD INPRES) in the 1970s; the second instrument is father’s schooling. Two results stand out: more schooling causes higher BMI for men and there is no causal effect of schooling on BMI for women. This chapter also provides some very preliminary evidence that the shift from blue collar to white collar and service sector occupations is one contributing factor to why more schooling increases BMI for men. The third chapter also uses the SD INPRES programme but to examine the effect of increased school supply on schooling attainment: overall, by gender, and by socioeconomic background. It also constructs a new SD INPRES programme exposure variable as an instrument for schooling to assess the causal effect of schooling on wages. The results strongly suggest that the SD INPRES programme increased schooling for men and women but that women benefited more as did individuals from less advantageous socioeconomic backgrounds. More schooling also causes higher wages and there appears to be an added positive effect for women through the additional schooling induced by the SD INPRES programme.