Socioeconomic status and enrolment in science and maths: a machine-Learning approach

Secondary school subject choices play a crucial role in shaping enrolment in science and maths in higher education and are influenced by academic, social, and institutional factors. Using data from the Growing Up in Ireland study, we document substantial socioeconomic gaps in enrolment in science and maths and show that choices made by age 13 strongly predict later participation. Applying SHAP (Shapley Additive exPlanations) based decompositions, we find that these gaps are largely predicted by differences in ability already evident at age nine. Our findings highlight early numeracy and lower-secondary choices as key margins in science-and-maths inequality.