When colleges defend the use of financial aid based on academic merit, they almost always make the case that it’s not an either/or question with regard to students from low-income families. An institution can benefit from recruiting top students with merit aid and still maintain its commitment to those with lots of academic talent but not much money. And by attracting students with merit aid, the argument goes, an institution may improve in quality, attract more funds, and even be able to do more for low income students.

A new study -- while not challenging the intentions of those making such claims -- suggests that, in fact, the adoption of merit aid by private colleges may achieve something for colleges that care about SAT averages. At the same time, the impact may be negative when it comes to economic or racial diversity. In fact, the study finds that, 10 years after private colleges begin offering merit aid, they are likely to be enrolling smaller shares of Pell Grant recipients and black students than they were prior to using merit aid.

The study -- “Keeping Up With the Joneses: Institutional Changes Following the Adoption of a Merit Aid Policy” -- has just been published by the Cornell Higher Education Research Institute. The author is Amanda L. Griffith, an assistant professor of economics at Wake Forest University.

For the study, Griffith used a database created from an annual survey of colleges conducted by the College Board on selected financial aid policies. The survey data run from 1987 through 2005, and Griffith isolated for analysis those 93 private colleges that were not offering aid at the beginning of the time period, but started doing so during it. Most of them started fairly early in the time period, so Griffith has data 10 years out for most of the colleges that started to offer merit awards. She then used various other databases to examine trends in the demographics and other characteristics of the colleges.

Among the key findings:

• Three to five years after colleges start offering merit aid, the percentage of Pell Grant recipients starts to drop at middle and top tier colleges (as measured by selectivity, using SAT scores as a proxy.) Six to 10 years after starting to offer merit aid, these colleges have seen their percentage of Pell Grant recipients drop by an average of five percentage points. The change is much smaller for less competitive institutions.
• In the immediate few years after merit aid starts, there is not a notable impact on the enrollment of black students. But after that, top and middle tier institutions start to see a decrease of 1.5 percentage points in black enrollment, growing to 2 percentage points 10 years out. (Much smaller shifts are seen for Latino students.)
• Merit aid may have a positive impact on diversity with regard to international students. Many American colleges provide relatively little if any need-based aid to students from outside the United States, so relatively modest merit awards may have a significant impact on enrollments. Within five years after starting merit aid, middle and top tier colleges see a 2 percentage point increase in international enrollments, and bottom tier colleges see an increase of 3.5 percentage points.

What about the impact that many colleges want when they add merit aid -- the enrollment of more top students as measured by factors like SAT scores, which are part of the rankings game? Over all, SAT medians go up after the introduction of merit aid. But for the most competitive institutions that start merit aid, there tends to be a drop in the first year or so -- most likely because institutions of this caliber that are starting merit aid are likely to be doing so because they are losing students to other institutions already. There is a rebound a few years out. Middle tier institutions see a modest gain of 22 SAT points a few years out, which increases to 35 SAT points 10 years out.

The paper notes that “it takes a few cycles of offering merit aid before word gets out and the program begins to attract many higher test score students.”

The SAT gains for bottom tier institutions are small.

The paper ends by expressing concerns about the trends it documents. “It is worrisome, given the already low levels of representation of low-income and minority students at four-year colleges, to find that the introduction of a merit aid policy is associated with a decrease in the percentage of low-income and black students, particularly at the more selective institutions in the sample. This crowding-out is likely due to an increase in merit aid spending at the expense of need-based financial aid. In conjunction with the rising costs to students following the switch to merit, this relationship is something that needs more research.”