We employ a new data set from Suriname to estimate private and social returns to technical, vocational and two tracks of general education (mathematics and language). Return estimates are based on gender-specific wage equations, corrected for sample selection bias and adjusted for unemployment....