Higher education dropouts, access to credit, and labor market outcomes: Evidence from Chile

In this paper we estimate a structural model of sequential decision of higher education and dropout to evaluate the impact of short-term credit constraint on dropouts. In particular, we analyze the impact on dropouts of the State Guaranteed Credit program (CAE) the most important funding program for higher education in Chile. The model allows for heterogeneity in observable and unobservable individual characteristics (Heckman et al. 2006) and controls for selectivity. Our data combine different sources of information, including individual data from standardized tests scores (PSU), higher education enrollment and unemployment insurance system (UI) data. The results show the important role of the abilities of individuals on educational choice. Individuals with higher ability tend to enroll in universities and not to drop out (sorting on ability). We show that household income and access to credit influence the probability of dropping out. Specifically, the results suggest that CAE has a positive impact on reducing the dropouts from higher education, where the program reduces the first year dropout rate in 15.5% for those enrolled in a university and a 24% for enrolled in a Center or Technical Formation (CFT) and Professional Institute (IP). We also found that the CAE is more effective in reducing the probability of dropping out for low-skill individuals from low-income families. However, our results show that CAE beneficiaries have lower wages than those who are not beneficiaries (even after controlling for characteristics, ability and selectivity bias). We attribute this to a incentive problems in the design of CAE which may lead to higher education institutions to reduce the quality of education. The evidence then calls to revise the design of CAE.