Longer Exposure, Better Outcomes: Evidence from Indonesia's BOS Program

This thesis studies whether Indonesia’s Bantuan Operasional Sekolah (BOS) program improved educational attainment by reducing the recurring cost burden of schooling. BOS was introduced in 2005 as a national school operational grant intended to support school operating costs and reduce routine charges faced by households in basic education. Using data from the Indonesia Family Life Survey (IFLS), I compare siblings within the same origin household who differed in poten- tial BOS exposure because of their birth cohort. I combine this within-household cohort variation with cross-province variation in a BOS-related implementation- intensity proxy constructed from the 2007 IFLS school-facility data.

One additional year of potential BOS exposure is associated with about a 1 percentage point increase in the probability of reaching senior high school in higher- intensity provinces relative to lower-intensity provinces. Among higher-exposure children in the analytic sample, the median child had six potential BOS-exposed school years; at that level of exposure, the preferred estimate implies roughly a 6 percentage point higher probability of senior-high attainment. The estimates for completed years of schooling are weaker and less stable. Taken together with three sensitivity checks, the results provide suggestive causal evidence that BOS supported progression to senior high school where implementation intensity was stronger, while also highlighting the need for richer administrative data on BOS disbursement and execution.