Need vs. merit. That's the debate among policy experts and presidents as they discuss institutional aid policies. And while merit aid may be best associated with some private colleges, it has become increasingly clear in recent years that the public universities are very much using merit aid to shape their classes.

In a study being released today, however, three scholars of higher education economics argue that what's going on in public higher education is less of a merit/need dichotomy, and that aid selections are reflecting the values colleges place on both kinds of aid. The scholars use a database of information from public flagship universities and focus not on different kinds of aid, but on different characteristics of students. In the process, they find correlations (but not necessarily causation) between characteristics of financial need and more aid and characteristics of merit and more aid. They also find a strong relationship between racial minority status and receiving more aid -- across other characteristics of students.

The study is part of College Success: What It Means and How to Make It Happen, a book being released today by the College Board. The authors of the study are Michael S. McPherson, president of the Spencer Foundation; Morton Owen Schapiro, president of Williams College; and Francie E. Streich, a research associate at the foundation. McPherson and Schapiro are also co-editors of the book.

The split between need and merit "is not so tidy" as some analysts make it seem, the study says. "Colleges and universities have been known to label aid allocations in one of these ways while distributing it for other reasons. In this analysis, we eschew the common categorical distinction between 'need-based' and 'merit-based' aid." Rather they study examines who gets aid at public flagships -- based on levels of family income (a proxy for the values associated with need-based aid) and SAT scores (a proxy for merit).

While part of the difficulty in measuring aid stems from the way colleges mix different kinds of assistance, another problem the authors note is "endogeneity" -- in that the factors that determine aid eligibility also may determine success in college. For instance, low income levels may qualify students for aid but also make it less likely that they come from high schools or socioeconomic backgrounds that make it likely for them to succeed. Aid based on merit may go to the students most likely to succeed anyway. So researchers and policy makers would do well to focus less on definitions of kinds of aid and look more simply at where the aid goes, the authors write.

"Rather than rely on the labels, let the data speak for themselves," said Schapiro in an interview.

The study found that, across selectivity level of public flagships, wealthier students received less in grant aid, suggesting a link to need in distribution. But 100-point increase in SAT scores, on average, yielded increases in students' grant sizes, especially at the least competitive flagships -- those frequently engaged in efforts to raise their competitiveness by attracting more top students. That suggests quite a bit of influence of merit in aid distribution. (The data come from 21 public flagships, from students who entered college during the 1999 academic year.)

Marginal Impact of Income and SAT Scores on Grant Size

McPherson said in an interview that the data suggest one flaw of the merit/need dichotomy is that it assumes one primary motivation for universities' aid policies, even if the motive may vary from institution to institution. "They have more than one goal and they are using more than one tool," he said. "They care about needy students and they consider it important to get more high scoring students."

While critics of merit aid suggest colleges use it merely to become more prestigious, the authors of the study say that before drawing conclusions, need to look at the motives and policies used. McPherson pointed out that research on "peer effects" shows that having some very top performers in a class has an impact -- positive academically -- on the rest of the student body. But Schapiro noted that such a positive impact would be the case only if these students are truly part of the student body. A public university that attracted such students with merit aid, and gave them an honors college for courses and a special dorm to live in, might not see the benefits on other students, he said.

The study also found that minority students receive more grant aid, on average, than white students across all aid and SAT levels at public flagships. Black students have larger grants than comparable students from other minority groups, the study says. The black-white gap in grant aid, across other factors, is about $2,000.

McPherson said that he believes this shows colleges are using "preferential packaging," giving minority students more grant aid and smaller loan requirements. And he noted that public universities use such packaging techniques for other groups too. Generally, he thinks the gaps are not due to race-specific scholarships.

Both Schapiro and McPherson said that the database they used -- from the Andrew W. Mellon Foundation -- created an unusually large sample for them to examine. At the same time, they noted that the data are from students who enroll -- and in some cases colleges may be offering aid in different proportions to students they want to attract, but who turn them down.