Why Kevin Carey is (mostly) wrong about merit aid

Kevin Carey is one of the smartest and most eloquent education analysts in the United States. He is also mostly wrong about the uses of merit aid.

Carey's recent Chronicle of Higher Education piece, "Too Much "Merit" Aid Requires No Merit," argues that a significant amount of merit aid (institutional scholarships based on academic performance) is given to students whose academic record does not merit it.  In giving such aid to the "stupid sons of the rich" (here Carey is quoting Harvard's turn-of-the-20th century president Charles W. Eliot) higher education both forces taxpayers to subsidize other students more heavily than it otherwise would and debases the meaning of the word "merit."

His article is wrapped around an anecdote about the son of friend.  The parents were intelligent and rich; the son shared his parents wealth but not their academic prowess.  Nonetheless, he was admitted to several decent private colleges, two of whom offered him a merit scholarship--in Carey's telling in order to entice his wealthy family to pay the rest of the $50,000 annual bill. The parents were incredulous, the father telling Carey, "He's never gotten a 'merit' anything before...He's not a very good student."

It is true that there are students at many institutions like the student in this story, who were indifferent academically but still qualified for merit aid. And it is true that giving such scholarships to the children of the wealthy somehow seems, well, wrong.  It is, in my experience, rarely the case that such a scholarship is given to a student only because s/he is from a rich family, though. And it is also true that the alternatives to awarding merit aid in this way are not much better, either for students or for the institution.

Think about the three pricing models that private colleges use to attract students.  In the first, the college prices tuition at what it costs to educate the student and offers almost no scholarships. In the second, the college prices tuition at roughly what the market will bear, but then provides institutional aid (almost entirely in the form of a discount) based on family need.  In the third, the college again chooses a market price, but then provides institutional aid (again, as discount) based on merit.  What happens for students and for the school?

In model one, schooling is largely available to students of middle and upper-middle class families who can afford the $15K+  that it costs to cover the costs of education.  The student body is thus from roughly the same economic strata, but of varying levels of academic preparation. Will the school be able to enroll enough  students to keep the doors open?  Only if its reputation is strong enough, its quality high enough, or its costs low enough to generate demand.  Otherwise, students have no reason to choose the school.

In model two, schooling is largely available  either to wealthy students who aren't academically strong enough to go somewhere else, or to students with significant enough need, met by need-based aid, that the cost of attending becomes affordable. Again, academic preparation varies widely, but so does the economic well-being of families.  And again, the school will struggle to enroll enough students, unless it has very generous donors whose gifts offset discount, or its reputation is so distinctive that it can attract both of its potential main audiences.

In model three, schooling is available to students with a relatively narrow range of academic preparation (those whose grades qualify them for merit under the school's criteria), and with a relatively narrow economic range as well.  The school will attract students who can afford it, who are attracted to its message, and who may be enticed by a reward for their prior academic performance.

None of these models is inherently better than another.  Schools choose them based on a mixture of their position in the marketplace, their mission, and their view of the social ends of education.

In practice, no school uses one of these pure models.  Most use a mix of models 2 and 3, supplemented by federal or state aid.  In the case of Westminster College, for example, applicants with an ACT score above 21 and a high school GPA above 3.0 earn merit scholarships. Scholarship amounts and ranges are posted publicly. Very few students below that range come to Westminster.  But at each merit scholarship level, students are also evaluated for need-based aid, some from the college, and some in the form of federal loans and Pell grants.  It is the case, therefore, that attending Westminster is actually less expensive for a student with financial need than for a student with a comparable academic background from a more prosperous family.  It is also less expensive for a student with a strong academic background, regardless of his/her economic situation, than it is for a student with a weaker academic record but the same economic profile.

These results--that the cost of school is less for needier families than richer ones, less for strong students than weak ones, and  least for needy, bright students--are defensible on social and educational grounds. But for a school that accepts these results, there are several implications. For enrollment managers, the biggest challenge is balancing the number of prosperous and less-prosperous students, and the number of academically strong and less strong students, so that the institution earns enough revenue to stay open and meet its goals of maintaining access to a quality education.

I have written critically of this practice, known as financial aid leveraging, because it makes price opaque to students and because it may not be well-founded in human psychology. It can cause academic problems as well, since the range of academic preparation can (though needn't necessarily) vary widely at leveraged schools. (That can also be the case at schools that hardly leverage at all). Further, students who are among the weakest academically at any institution are less likely to be retained.  But so are students who are needier, so leveraging is no guarantee of college success.  And for someone like Carey who is looking at higher ed as a whole, the practice raises questions about the quality of the system, since because different schools attract different pools of potential applicants, a student with a 1000 on the SAT may get merit aid at one college, but nothing at another.

Kevin Carey decries financial aid leveraging also when he criticizes institutions for giving merit aid to "the stupid sons of the rich." But Carey is wrong about schools' motivation for giving merit aid to wealthy underachievers. Schools don't do it because they want to enroll more stupid sons of the rich.  We do it to enroll more bright daughters of the poor.  That goal may be worth a few thousand dollars of merit aid to a "stupid son", even for a kid whose parents don't think he deserves it.