Merit scholarships have a real impact on the yield of top admitted students, but unless those scholarships are exceptionally large, the yield is likely to remain small. That is the conclusion of a new study of an unusual experiment in which a private college created a random sample and control group to test the effectiveness of merit awards.
The rapid growth in merit scholarships has been controversial: Many institutions (public and private) say that the awards allow them to better shape their classes and to attract talented applicants who might otherwise go elsewhere. Yields -- the percentage of accepted applicants who enroll -- go up. Critics have said that merit scholarships may help institutions, but don’t truly help students (most recipients of merit scholarships have many options) or broad social needs in higher education (the merit awards divert attention and funds from the needs of low-income students). Amid this debate, many institutions are standing by their merit scholarships, while others are scaling them back or abandoning them.
In a study released by the Cornell Higher Education Research Institute, James Monks notes that there is relatively little research on the true impact of merit scholarships. That's because the institutions that offer them tend to do so as part of very specific strategies that they want to execute, so they focus available funds on those applicants that they most want, and comparisons to other groups of applicants aren’t valid. Monks is associate professor of economics at the University of Richmond and one of his research areas is the economics of higher education.
In his study, Monks took advantage of the decision of a private university to undertake a true experiment with merit scholarships. This university decided to offer $7,000 renewable merit scholarships (about 17 percent of tuition, room and board and fees) to 230 of its top rated applicants. But rather than offer the grants to those at the very top, the institution offered them completely at random to a subset of a larger group of applicants who had been identified as being at the top of the pool of those admitted. The pool excluded those who were receiving any need-based financial aid or other merit scholarships. There were 319 admitted applicants who were thus judged equally desirable as the first group, but who by random choice received nothing extra from the university.
Some previous research on student behavior with regard to scholarships has suggested that their impact goes beyond financial to the perceived “scholarship” or “price illusion” effect, in which students feel good about being identified as being special or saving money, and so may be responding to more than the money. And this of course extends to parents, Monks notes, since a merit scholarship for a child can lead to “cocktail party bragging rights.” To include these not-strictly-financial factors, the students in the group receiving merit awards were told that they had been identified as being “exceptionally talented even among a deep pool of outstanding and accomplished students.” The merit scholarship was termed an Academic Recognition Award.
An additional wrinkle: The $7,000 merit award matched the size of a one-time tuition increase the university had made in 2005, as part of a strategy to reposition itself. So the university could also compare the results to students who were identified as the top of the applicant pool the previous year and who had not been offered merit awards. They had the same sticker price as those who did receive merit awards in the year studied.
The results: The yield for 2005 for the top applicants in the pool who received no aid was 3.2 percent. The yield for those who received the $7,000 annual scholarships was 7.1 percent – a statistically significant difference from the no scholarship group. But the previous year, the yield for similar applicants, receiving no aid but paying the same sticker price as the 2005 merit recipients, was 5.2 percent. So the numbers suggest that the scholarships not only have an impact financially, but also possibly psychologically, since the yield was better than for those facing identical bills a year before.
At the same time, the study suggests that the impact of this kind of merit aid may not be large -- the yield rate was low even for those receiving it. (Many colleges that offer merit awards offer much larger grants, sometimes worth half of total sticker price, or even more.)
In terms of judging impact, Monks said in an e-mail interview that impact may be “in the eye of the beholder.” He explained that “this institution does not matriculate a very high percentage of its most highly rated admitted students without the incentive of some form of aid, either merit and/or need based. The merit awards do have a significant (statistically speaking) impact on enrollment probabilities, but without generous aid awards few of the most talented students matriculate to this institution.”
The study does not name the institution. However, all of the characteristics noted in the study (enrollment, tuition rates, the timing of an unusually large tuition increase) match the University of Richmond (and not other institutions). Monks said that he was not able to disclose the institution studied.