Student Loans and College Participation

Over the years, a large number of studies have been conducted on the relationship between student loans and students’ access to college, choice of institutions, persistence through college, and attainment of degrees.

Most of these studies, however, suffer from a number of analytic and methodological limitations. In the ideal world, in order to test the effectiveness of student loans on college participation one would conduct a true experiment.

Such an experiment could take a number of forms, but one would be to select a number of high school students who have financial need (i.e., their college costs exceed their families’ expected family contributions) and randomly assign them to groups.

Each group would be offered a different set of financial aid packages; the first might be grants alone, the second might be the same amount of grants, supplemented by loans, and the third group might be offered only loans (or perhaps different types of loans—subsidized, unsubsidized, private, different repayment schedules, etc.).

The students would then be tracked to determine the rates at which they enrolled in, persisted through, and graduated from college.

The experiment would allow you to determine whether student loans, as compared to grants, made a difference (either positively or negatively) in the students’ educational behavior, and whether different types of loans had varying impacts.

With sufficient information about the background and other characteristics of the students, you could also draw conclusions about the impact of different forms of aid on subgroups, such as students from different races or income categories.

Unfortunately, this type of true experiment rarely is carried out in education research. So what we are left with is a series of studies that attempt to model this real-world experience through statistical analyses of existing data, or what are often described as observational studies.

One of the limitations of many of these studies is that they do not follow the same cohort of students longitudinally from high school, into college, and through to degree attainment (and beyond). And those that have followed the same students generally suffer from limited information about student financing of postsecondary education in general, and even more so, about the use of loans specifically.

For example, the National Education Longitudinal Study of 1988 (NELS 88), conducted by the National Center for Education Statistics of the U.S. Department of Education, followed a nationally representative sample of students who were in the eighth grade in 1988 (National Center for Education Statistics, 2007).

It tracked the students periodically when they were in high school, and then beyond, up to the year 2000, or approximately age 26.

While this is a very rich resource for the educational experiences of students, it has very limited information about the use of student loans by those students who attended college.

Other studies have much more detailed information about the type and amount of loans students took out, but they only focus on students who were already enrolled in college and/or have insufficient information about other characteristics and behaviors of the students related to college participation that would allow the researcher to isolate the impact of the student loans themselves.

Another limitation is that many of the studies were conducted before the large increase in student borrowing that has occurred since the 1992 reauthorization of the Higher Education Act, so their studies may have limited applicability to current groups of students.

What is difficult to determine from observational studies is what the behavior of students would have been in the absence of student loans.

If they had not borrowed money to finance their education, would students have been more or less likely to enroll in college, enroll in a particular sector, persist, and ultimately attain a degree? A nuanced answer to this question would be, “It depends on how they replace the loans as a financing tool.”

For example, if they choose to take a job while in school (or work more hours if they already are employed), this decision may have a more negative impact on persistence than does borrowing.

This scenario is described by King (2002), who provides examples of student behavior that may seem rational in the short run, i.e., choosing to work more hours rather than incur debt, but in the longer term is more harmful to persistence and degree attainment.

Alternatively, consider the situation of the student contemplating going to college who has unmet need after taking into account all of the grants she is offered along with the amount of money her family is able and willing to contribute to her postsecondary education costs.

If she sees her situation as a dichotomous choice of either borrowing for college to help fill the gap between college costs and available resources, or not borrowing for college, then one could conclude that borrowing should have a positive impact on college access (or choice, if she is choosing between two or more alternative institutions with different levels of unmet need).

But the reality is that this is rarely the situation most students face. There are a variety of financing opportunities, particularly if one takes into account the options presented by attending differently priced institutions, each with its own financial aid offer.

This is where the observational studies begin to lose their predictive ability because of the lack of information available about the various alternatives students faced.

For example, a research database (such as the NPSAS surveys) may have detailed information about the type and amount of loans taken out by a student once enrolled in college, but it is likely to have much less information about the student financing alternatives offered to or considered by the student (either at the time of transitioning from high school to college or while persisting through college).

Without this detailed information, it is very difficult to assess just how effective student loans are in promoting or hindering college participation.

Having outlined these limitations of the existing research, in this section I summarize some of the findings of the work that has been conducted on the relationship between student loans and college participation.

The findings of these studies regarding the impact of student loans on college participation can at best be described as mixed. Cofer and Somers (2000) used the NPSAS of 1987 and 1993 to examine the relationship between the amount of borrowing and whether students persisted in college.

The authors found that higher levels of borrowing were more likely to negatively affect the persistence of students in private baccalaureate institutions (which, on average, have higher tuition levels) as compared to students in public colleges and universities, controlling for student and institutional characteristics, as well as grant receipt.

The authors noted that students in the private institutions had higher levels of borrowing, on average, as a possible explanation for this effect.

Dowd and Coury (2006) used the 1990 NPSAS data and its longitudinal component, the Beginning Postsecondary Students (BPS) Longitudinal Study, to examine the persistence (from first to second year) and associate degree attainment of community college students, to determine how loans affected these outcomes.

The longitudinal component of BPS interviewed students in 1994, and asked them about their college enrollment in each month between the initial survey in 1990 and the 1994 follow-up. The authors’ multivariate analyses found that while loans were negatively related to persistence from the first to the second year, there was no impact of loans on subsequent degree attainment.

The multivariate analyses controlled for a number of factors, including receipt of grant aid. This could indicate that the transition from the first to second year is particularly important among borrowers; if students were able to make it into the second year, their firstyear borrowing did not appear to affect their degree attainment.

Dowd and Coury (2006) emphasize that the decision to borrow for college, at least among the population of students they studied, is a complex process.

It includes aspects of eligibility, or qualification, for loans; self-assessment of the risk involved in borrowing; selection into (or out of) borrowing, as opposed to alternative financing opportunities; and calculation of cost subsidies (if any) provided by various forms of borrowing.

In a review of the literature on student loans and college participation, Campaigne and Hossler (1998) concluded that it was difficult to assess the impact of loans on overall access to college.

They noted that middle- and upper-income students are more likely to borrow than lower-income students, because of the loan aversion of this latter group.

However, they speculate that the difference in college enrollment rates between this group and their wealthier peers is likely due to inadequate grant aid for poorer students. St. John (2003) came to a very similar conclusion regarding the inadequacy of grant aid.

Campaigne and Hossler (1998) drew similar conclusions regarding the role of loans in promoting college choice.

Because middle- and upper-income students were more willing and likely to borrow, loans enabled them to attend more expensive, generally private, four-year institutions than they otherwise would have been able to afford.

The researchers found loans to have less of an impact on the decisions of poorer students, and they found more mixed results with respect to the effect of loans on persistence, concluding that “federal loans appear to have a very small, modest effect on persistence; however, much more research is needed on this topic” (p. 100).

In another review of the literature (Heller, 1997), I looked at a number of studies that examined the relationship between tuition prices, financial aid of various types, and enrollment in college. My review concluded that while some studies found a positive relationship between loans and college enrollments, overall the research indicates that college enrollments are less sensitive to loans than to grants.

While this may not be a surprising conclusion, it does help refute the arguments made by some that a dollar of loan aid may have an equivalent enrollment effect to a dollar of grants, because both serve to reduce the net price paid by the student at the time she enrolls in college.

Hu and St. John (2001) looked at the persistence of students in public colleges and universities in Indiana. Examining three cohorts of students from the 1990s, the authors sought to determine what types of financial aid packages were effective in promoting persistence.

They examined the mix of grants, loans, and other forms of aid in a package, but did not look at the specific amounts of each form of financial aid awarded to the students.

They concluded that student loans, particularly when packaged along with grants, had a positive impact on persistence for students from different racial groups (controlling for other student and institutional characteristics), with the largest impact on Hispanic and African American students, as compared to white students.

Another study that used the NPSAS data from 1987 also examined the impact of various forms of financial aid on persistence (Paulsen and St. John, 2002). The authors found that borrowing was negatively related to student persistence, particularly for lower-income students.

But rather than concluding that it was the act of borrowing to pay college expenses itself that was harming students’ persistence, they concluded that the problem was that financial aid of all forms was inadequate to help these students meet their college expenses.

Perna (2000) looked at differences in financial aid sensitivity among students from different racial groups, using NELS data to examine the relationship between a number of factors, including financial aid, and enrollment in fouryear institutions.

She found that student loans reduced the probability African American students would enroll in this sector concluding that:

[L]oans reduce the probability of enrolling for African Americans after controlling for sex, costs, benefits, ability, and social and cultural capital. The negative relationship between loans and college enrollment among African Americans may reflect inadequate knowledge about the availability of financial aid, an aversion or distaste for borrowing, and/or an expectation that future earnings will be insufficient to repay the loans. (p. 137)

Her study also found that grant aid had no impact on enrollment in four-year institutions for any of the three groups she examined (white, African American, and Hispanic students). She concluded that “financial aid alone is not sufficient to increase college access” (p. 137), as the other factors she examined played a more critical role.

As noted in the beginning of this section, the literature on the effectiveness of loans in promoting college participation is somewhat lacking, because of the lack of true experimental studies that would allow researchers to causally link the awarding of loans to students’ decisions to enroll in college and persist once there.

Most of the research addressing this question uses descriptive and or quasiexperimental methods that provide a much less definitive answer to the question of the effectiveness of loans.

A major problem in understanding the relationship between loans and college participation, particularly for studies conducted in recent years, is the fact that loans have become an increasingly important part of the college financing package for many students.

A generation ago, many students were able to finance their postsecondary educations through a combination of their own (and their family’s) resources, along with available grant aid provided by federal, state, institutional, and private sources.

Loans were often seen as a vehicle for the financing of postsecondary education at the more expensive private colleges.

But many of today’s students, even those attending relatively low-priced community colleges, must rely on loans to pay their tuition and other charges because grant aid is no longer sufficient to supplement their own resources.

The literature reviewed in this section can be summarized as finding that loans have little impact on college access and persistence for most students. But this finding needs to be understood in the context of a financial aid landscape in which many students do not receive sufficient grant aid to pay for college.

If grant aid were proportionally higher, then loans might provide more of a positive impact on college participation. But absent sufficient grant aid, simply piling on higher amounts of borrowing to students with large levels of unmet financial need may not be an effective vehicle for getting them to college.

Location: Indonesia