The author exploits random assignment of student loan borrowers to student loan servicers of varying quality to study the direct effect of servicers on borrowers’ credit outcomes and to isolate variation in the likelihood of default that is not correlated with borrower characteristics. She finds that being assigned to a higher-default servicer increases a borrower’s likelihood of default by approximately 6 percent. However, there is a precisely estimated null effect of servicer assignment on measures of borrowers’ likelihood of financial distress, credit access, and zip code characteristics. These findings suggest that averting a servicer-induced default does not yield considerable benefits for marginal borrowers’ credit outcomes but that servicers are meaningful drivers of student loan repayment outcomes.

View the Full Working Paper