A Temporal Investigation of Factors Related to Timely Degree Completion

Introduction

Graduation, especially timely graduation, is an increasingly important policy issue, and for good reason. College graduates earn twice as much as high-school graduates and six times as much as high-school dropouts (Murphy & Welch, 1993); and their wealth is two and one-half times that of a high-school graduate and five times that of a high-school dropout (Diaz-Jiminez, Quadrini, & Rios-Rull, 1997). In addition to these financial rewards, the spouses of college graduates are more educated and their children do better in school and are less likely to get into trouble with the police (Jencks & Edlin, 1995; Murphy & Welch, 1993). Despite the obvious rewards, graduation from U.S. colleges and universities is far from assured (DeBrock, Hendricks, & Koenker, 1996). Lederman (1991) reported that of all students enrolled in Division-I schools in 1984, only 48% had graduated by August 1989. In the data used in this study, 41% of the entering class of 1991 had graduated within six years of enrollment. Of these graduates, 88% graduated without any interruption (henceforth "timely graduates").

The fact that only about one-half of college attendees graduate is widely perceived as a failure—a failure of either the student, the institution, or the entire educational system. This view is held particularly strongly by state legislatures. For example, Virginia intends to tie institutional funding to graduation rates and a number of other outcomes (Hebel, 1999). At the University of Minnesota, the institution examined in this article, graduation rates are one of the institutional effectiveness measures that have had funding explicitly tied to them for the past several years. Although concern about low graduation rates is understandable, the degree of this concern does not seem to square with the empirical evidence given that only 15% of students drop out because of academic troubles (Kalsner, 1991). Further, DeBrock et al. argue that those who fail to graduate do so as the consequence of a rational economic choice: "Each student must determine if the value of completing the degree makes persistence rational" (1996, p. 520). That is, are the net returns to persisting greater than the net returns of dropping out? As Altonji (1993) has argued, new information about preferences, abilities, and/or market opportunities may cause students to alter or interrupt their academic career. For them, the graduation decision is part of the overall labor market optimizing problem, an approach pioneered by Manski and Wise (1983). Two important conclusions follow from the optimization approach. First, it is incorrect to view those who fail to graduate as primarily a failure on the part of the university, and second, failure to graduate does not necessarily reflect a lack of ability on the part of the student. No matter one's perspective on why students fail to graduate, however, it is important to understand the determinants of successful and timely degree completion.

Most studies of student departure focus on the characteristics of students as determinants of success. The assumption is that poorly prepared students find college more difficult and are thus less likely to graduate (DeBrock et al., 1996). We too investigate the impact of student demographic characteristics, but because we have good controls for ability and academic performance, we interpret any effect as operating through relative rates of return to graduation rather than as measuring some failure on the part of the student. We add several attitudinal variables not usually considered and also include detailed data on financial aid. Financial aid plays an important role in higher education by lowering the costs of attendance (McPherson & Schapiro, 1998).

We also investigate whether students in different colleges within the university have different graduation rates. Such differences may reflect different standards among collegiate units. However, different rates of graduation may also reflect differential returns to employment in the labor market. For instance...