ABSTRACT
We investigate the effect of an educational cash transfer on schooling and working of the recipients and their non-recipient siblings in Indonesia, using a matched difference-in-differences strategy. We find that the cash transfer increases the probability of schooling for all recipients. Specifically, the likelihood of schooling for the senior secondary school children increases by 19 percentage points. However, there is no effect on the recipient's probability to work. Furthermore, there is no significant spill-over effect on non-recipient siblings' schooling. While the transfer reduces the incidence of child labour for non-recipient girls, it increases the probability of non-recipient boys to work.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Since the individuals are aged between 13 and 18 years in 2014, we use age dummies as controls in all regression estimations.
2 Synthetic control method is not appropriate here as it requires many pre-treatment periods in order for the method to be credible (Abadie Citation2021).
3 We also consider bootstrapped standard errors with 5000 replications and the results are qualitatively similar (Table S6).
4 There results are with standard errors clustered at the province level.
5 Since fixed effects logit model does not converge, we use a linear probability model with fixed effects in this analysis (King and Zeng Citation2001).