Student loans have become an increasingly important source of financial aid for college students. Between 1992–93 and 2000–04, the proportion of all undergraduates borrowing in a given year to help pay for their education increased from 20 to 35 percent at the undergraduate level (Tuma and Geis 1995; Berkner 2005), and from 19 to 42 percent at the graduate level (Choy and Premo 1995; Choy and Cataldi 2006). As borrowing has increased, long-standing concerns about students’ ability to repay their loans and the effect of the debt on their lives after college have intensified.
The first part of this report describes the undergraduate borrowing patterns of 1992–93 bachelor’s degree recipients and their graduate enrollment and additional borrowing through 2003. These graduates would have completed their undergraduate borrowing prior to the changes introduced by the 1992 reauthorization of the Higher Education Act. At that time, only students with financial need could have participated in federal loan programs as undergraduates.
The second part examines the repayment of undergraduate loans for bachelor’s degree recipients who had no additional degree enrollment, providing details on how many had finished repaying their loans by 2003, who were still repaying and how much, what their debt burden was, and how they had managed their Stafford loan repayment over the 10-year period.
The report uses data from the 1992–93 Baccalaureate and Beyond Longitudinal Study (B&B:93/03), a longitudinal study of students who earned a bachelor’s degree during the 1992–93 academic year. Base-year information on this cohort was collected as part of the 1992–93 National Postsecondary Student Aid Study (NPSAS:93). Graduates were interviewed again in 1994, 1997, and 2003. These data were supplemented with data from the National Student Loan Data System (NSLDS), which contains detailed records on the repayment history and 2003 status of Stafford loans taken out by the 1992–93 graduates.1 All comparisons made in the text were tested using Student’s t statistic. All differences cited were statistically significant at the .05 level.