background characteristics. Background characteristics can be demographic descriptors (like ethnicity, gender, and family income), factors that describe events or experiences prior to attending college (like SAT score), or variables that otherwise do not reflect student performance in college (like the amount of financial need a borrower has and the adjusted gross income of the borrower’s parents). In most cases, borrowers can do little to change background variables.
Prior research has emphasized the importance of both performance and background characteristics. Graduation status (or earned degree), a performance variable, has frequently been one of the most important variables in the outcomes of past default studies (Dynarski, 1994; Knapp & Seaks, 1990; Meyer, 1998; Volkwein & Szelest, 1995; Volkwein et. al., 1995; Wilms, Moore & Bolus, 1987; Podgursky et. al., 2000; Woo, 2002). However, these same studies have also found ethnicity, a background characteristic, to have a relationship to default that is almost as strong or as strong as the association between graduation and default. In addition, the studies have determined that other background variables are also important predictors of default.
The current study, while different in design from the others, nevertheless finds the background variables to be much less important to default behavior at Texas A&M University than the performance variables. This study finds that background variables in the College Preparedness section, which describes borrowers’ experiences in high school, the Demographic section and the Financial Aid section have weak relationships to default. In contrast, the performance variables in the College Success section and Attendance Pattern section represent most of the borrower characteristics that have strong associations to default. In particular, grade point average has the strongest association to default of any variable in the study, and ethnicity has a very weak relationship to default.
The structure of the present study might partially account for its conclusions being at odds with the findings of many previous studies, even though both sets of conclusions are valid. Unlike the studies cited above, this study analyzes only one institution – Texas A&M University. Other studies looked at many postsecondary institutions simultaneously. Examining multiple institutions might increase the chances of detecting large differences in default behavior between groups of borrowers with different backgrounds. If, for example, a group of borrowers with a particular ethnic background has a higher propensity to default on average and also tends to enroll at particular postsecondary institutions, while borrowers of other ethnic backgrounds tend to study at other schools, then only an analysis that incorporates a cross-section of these institutions can more fully account for the differences in default behavior that are associated with ethnicity. For similar reasons, other background variables might also have less explanatory or predictive power in a single-institution study than in a multiple-institution study.
At any rate, this outcome of the study – that background variables are less important than performance variables at TAMU – is somewhat heartening. Borrowers are succeeding in the student loan programs largely without respect to their ethnicity, their parents’ educations or their family income. Moreover, the study suggests that default behavior hinges more strongly on factors that are at least partially under the borrower’s control: whether the borrower graduates, how long the borrower spends in college and how well the borrower does in his or her college coursework. Additionally, the importance of the college success variables is consistent with one of the core principles of the student loan programs: student loans are worthwhile investments that borrowers will be able to repay, particularly when borrowers succeed in their programs of study.
At this point, one can only speculate about the precise relationship between college success and default. Graduating and achieving a high GPA do not directly cause a borrower to repay. Those factors might make borrowers better able to repay, by garnering them better jobs in the labor