Mónica Costa-Dias - University of Bristol

Education is a main facilitator of social mobility, and higher education plays a major role in this. But while returns to higher education are large, they also vary widely with the characteristics of programmes, those of students and how the two align. This paper investigates the role of sorting between students and higher education programmes in explaining heterogeneity in the returns to higher education, and the extent to which students from different socio-economic backgrounds (SES) and with different skill sets benefit from higher education depending on the field of study and the selectivity of the programme. Using rich administrative data for England, we document that students from disadvantaged backgrounds are less likely to enroll in higher education programmes than their equally skilled higher SES peers, particularly in the most selective ones, and that low SES is associated with lower wages conditional on education and skills. To investigate the drivers of these gaps, we develop a life-cycle model of education, labour supply and earnings that allows for a rich characterisation of heterogeneity in the skills that students have and that different educational programmes provide. A key feature of our model is that students and programmes meet in a matching market in equilibrium, where the assignment of students to programmes is determined. Exploiting geographical and cohort size variation in combination with the structural model, we can separate the preferences of students and programmes and identify the returns to heterogeneous programmes. Our model reproduces empirical patterns accurately. We use the model to describe sorting in the higher education market in England, quantify the returns to these investments and the role of higher education for social mobility. We find that students from poorer backgrounds benefit more from investing in higher education than better off students, but on the margin benefit less from enrolling in better quality, more selective programmes. We also find that low SES students are less willing to pay for a match that promises higher future earnings. However, students’ preferences cannot fully explain differences in higher education sorting by SES. Rather, systematic differences in skills and admission rules that favour better off students also play a role. We use the model to run counterfactual analysis of policies targeting the demand and supply sides of the higher education market. We conclude that demand side policies subsidising tuition fees or incentivising investments in high-returns programmes are insufficient to move the needle on social mobility; supply side policies could potentially be highly effective.