An official website of the United States government
If higher ability individuals are more likely to attend college and if there is significant ability bias in the college education premium, then a significant portion of the observed complementarity between the college and non-college educated is due to changes in the ability composition of education groups. If college attainment rose to over half the population, this composition effect would reverse, as is illustrated with high school attainment. If there is little ability bias, the ability distribution is nearly degenerate, with the awkward implication that the most productive in- dividuals would earn barely more without a college education than the least.