Schooling and Labor Market Consequences of School Construction in Indonesia: Comment

Abstract: I revisit Duflo (2001), which exploits a 1970s schooling expansion in Indonesia to estimate the returns to schooling. I correct data errors, check sensitivity to two specification choices not discussed in the original text, adjust for potential sources of bias (error correlation within the geographic units of treatment, endogenous weighting, non-parallel trends, weak instruments), and follow up on affected cohorts later in life. The reduced-form impact estimate for wages is robust, while that for years spent in school is not, making the former hard to attribute to the latter.