Evidence on credit constraints and university attendance.∗

The effects of credit constraints on university participation are investigated. Students most likely to face credit constraints have the same or higher probability of attending university as all other students, given their high school achievement. A novel approach to handle potential bias arising from unobservable heterogeneity is proposed and implemented, providing estimated unobservables based on post secondary plans reported during ninth grade. These estimated unobservables are important in explaining university attendance but results regarding the effects of credit constraints are robust to their inclusion in models estimated. Our new estimated unobservables approach is consistent with a related Copula estimation approach. JEL Classification: I21, I22, I28