Specialists or All-rounders: How best to select university students?

This paper studies whether universities should select their students only using specialised subject-specific tests, or on the basis of a broader set of skills and knowledge. The empirical analysis is guided by a theoretical framework. The theoretical model shows that even if broader skills are not improving graduates’ outcomes in the labour market, the university chooses to use them as a criterion for selection alongside the mastery of more subject-specific tools. This is so because broader skills allow the university to select candidates who are on average abler. I test the model on a large administrative dataset of Portuguese students. Within programmes, I exploit the variation between specific and non-specific entrance exam sets. My central finding is that, on average, universities with less specialised admission policies admit a pool of students who obtain a higher final GPA.