Estimate of the private return on education in Indonesia: Evidence from sibling data

• The aim reestimates ROE to prove the high ROE from former studies and the low enrollment rate for college.
• This study found the OLS, the 2 Step Heckman, and the 2SLS estimation results are biased.
• This study just can rely on the Household Fixed Effect (HFE) estimates using siblings’ data.
• The HFE estimates show ROE reduced by more than a half from 10.8% (OLS estimates) to 5%.
• It means ROE is not as high as earlier estimated and may explain why the enrollment rate for college is low.

This study reexamines the return on education (ROE) in Indonesia to solve the puzzle between the high ROE from previous studies and the low enrollment rate for higher education by taking into account endogeneity and sample selectivity issues. This study finds the OLS estimates of ROE in between 10% and 12%. After controlling for unobserved family background using sibling data, the Household Fixed Effect estimate indicates ROE reduces from 10.8% to 5%. This study argues ROE is not as high as previously estimated and may explain why the enrollment rate for college has been stagnating in Indonesia.