• DOI:10.2139/ssrn.4333118
• Corpus ID: 250644022

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

@article{Roodman2022SchoolingAL,
  title={Schooling and Labor Market Consequences of School Construction in Indonesia: Comment},
  author={G. David Roodman},
  journal={SSRN Electronic Journal},
  year={2022}
}

• G. David Roodman
• Published 19 July 2022
• Economics
• SSRN Electronic Journal

Duflo (2001) exploits a 1970s primary schooling expansion to estimate the returns to schooling in Indonesia. Under t he study’s difference-in-differences (DID) design, two patterns in the data — negative selection in treatment and shallower wage-schooling gradients for younger workers — can together bias results upward. In response, I follow-up later; test for trend breaks timed to the intervention; and perform changes-in-changes, which relaxes the parallel trends assumption. I also correct data errors, cluster variance estimates, incorporate endogenous survey weights, and test for (and detect) instrument weakness. Weak identification – robust inference yields positive but low-precision estimates. CIC estimates are mostly indistinguishable from zero. and Jonathan Morduch, and excellently Most estimates and inferences performed with ivreg2 (Baum, Schaffer, and Stillman 2007), reghdfe (Correia 2014), cic and Santangelo 2015), cmp (Roodman 2011), or boottest (Roodman et 2019). The packages esttab 2007), coefplot 2014), palettes 2022), and blindschemes (Bischof 2017) make tables and graphs.