As more undergraduates have taken out loans to attend college, the number of borrowers who fail to repay their student loans has increased. While previous research has focused on students' likelihood to default, this study employed institutional cohort default rates (CDRs) as an outcome variable. Using Integrated Postsecondary Education Data System, this study investigated the association between institutional effectiveness and CDRs. Coupled with multilevel modeling, the study also observed the effects of state-level factors, such as state appropriation and unemployment, on CDRs. The results showed that institutional characteristics--e.g., proportion of minority students, admission test scores, retention rates, and instructional expenses--are strongly associated with institutional CDRs. This suggests that institutional default rates are mainly a function of the students that institutions enroll, and future studies should include institutional as well as student factors to provide policy makers and researchers with a more comprehensive understanding of institutional CDRs.