Post-secondary education in Canada: can ability bias explain the earnings gap between college and university graduates?

Using the Canadian General Social Survey we compute returns to post-secondary education relative to high-school. Unlike previous research using Canadian data, our dataset allows us to control for ability selection into higher education. We find strong evidence of positive ability selection into all levels of post-secondary education for men and weaker positive selection for women. Since the ability selection is stronger for higher levels of education, particularly for university, the difference in returns between university and college or trades education decreases slightly after accounting for ability bias. However, a puzzling large gap persists, with university-educated men still earning over 20% more than men with college or trades education. Moreover, contrary to previous Canadian literature that reports higher returns for women, we document that the OLS hourly wage returns to university education are the same for men and women. OLS returns are higher for women only if weekly or yearly wages are considered instead, because university-educated women work more hours than the average. Nevertheless, once we account for ability selection into post-secondary education, we generally find higher returns for women than for men for all wage measures as a result of the stronger ability selection for men.