Alert: Shameless self-promotion. The paper was in the top 10 downloaded for the week at SSRN. Enrollment managers reject the work, ironically – for they do not believe the data driven analysis.

Tuesday, May 15, 2007

Colleges’ Financial-Aid Allocations May Often Be Inefficient, Study Suggests, By DAVID GLENN

Colleges may frequently be “overspending” by offering students larger financial-aid offers than are actually necessary to entice them to enroll, according to a working paper released last week by three economists.

In a detailed examination of the admissions practices of two selective private colleges, the economists found that the colleges generally made larger financial-aid offers to students who were already highly likely to attend. Those practices contradict scholars’ typical idealized models of financial aid. The standard model assumes that financial-aid offers are incentives to enroll, and that colleges will direct those incentives at “marginal” students — that is, students who might easily decide to attend a different institution.

“The incentives of the school and the incentives of an enrollment manager aren’t always perfectly aligned,” said Michael J. Rizzo, a senior economist at the American Institute for Economic Research, a nonprofit organization in Massachusetts, in an interview on Monday. In other words, admissions officials might be so anxious about meeting enrollment goals or student-quality targets that they spend their budgets inefficiently.

Mr. Rizzo wrote the paper, “The Cost of Crafting a Class: (In)Efficient Financial Aid Allocation at Two Private Colleges,” with Robert E. Martin, a professor of economics at Centre College, and Randy Campbell, an assistant professor of economics at Mississippi State University.

The authors’ suggestion that colleges should trim some of their financial-aid offers might sound vaguely heartless — and might seem reminiscent of “enrollment management” strategies that have been heavily criticized for apparently allowing colleges to chase prestige at the expense of low-income students.

Mr. Rizzo and his colleagues argue, however, that if the two colleges in their study spent their financial-aid budgets more efficiently, they could free up resources for any number of possible goals, including offering more need-based aid, diversifying their student bodies, or hiring more faculty members.

The two colleges in the study insisted on anonymity. One of them has a large endowment and a total enrollment of roughly 1,000 students. The other has a small endowment, relatively low tuition, and a total enrollment of roughly 5,500 students.

Mr. Rizzo and his colleagues were drawn to this project because they wanted to demonstrate that enrollment managers’ typical formulas are oversimplified. The standard formulas, they argue, do not take account of the fact that admissions officials already have opinions about how likely a student is to enroll before they calculate that student’s financial-aid offer. (In general, colleges calculate an applicant’s ‘enrollment propensity’ by looking at the past behavior of applicants with similar academic and demographic profiles. Admissions officers also sometimes make individualized judgments of an applicant’s leanings based on interviews and correspondence.)

“Financial-aid offers don’t just drop out of the sky,” Mr. Rizzo said. “They’re jointly determined with whether the school thinks the kid may or may not come.”

Professional enrollment managers “will probably be very apprehensive about our results,” Mr. Rizzo said. “Their models often sort of live in a static world. They’ll say, ‘All the kids from New York should get a scholarship of $5,000.’ But that’s ignoring the fact that all the colleges in New Jersey may decide to offer that same scholarship. And you can’t easily model that.” (Mr. Rizzo himself sometimes works as a consultant for Scannell & Kurz Inc., an enrollment-management firm based in New York, and he said that he has had friendly arguments about this paper with his employers there.)

Mr. Rizzo cautioned that the study involves only two institutions. “I would never say that this is automatically generalizable to the whole universe of colleges,” he said. He is now working on a follow-up study that includes data from hundreds of colleges.

Pressure to Produce Enrollments

The new paper “really touches a raw nerve about how an institution’s competing interests can sometimes work against themselves,” said David Hawkins, director of public policy at the National Association for College Admission Counseling.

“Our members are typically under the gun to snag those enrollments early, to get that yield rate nailed down quickly,” Mr. Hawkins said. “So it doesn’t surprise me that they’ve found that money is perhaps going to students who are already likely to attend.” With yield numbers declining even at highly selective institutions, Mr. Hawkins said, such pressures are only increasing. “Yield” refers to the percentage of students accepted at a college who choose to attend.

Mr. Rizzo offered a similar account. “As you get toward the deadline of admitting your class, admissions offices panic,” he said. “They just start throwing money at kids.”

Larry D. Singell Jr., a professor of economics at the University of Oregon, said in an interview on Monday that he finds the paper intriguing but not entirely persuasive.

“I think there are a lot of institutions that give out far too much aid, in the sense that they don’t need it to get people in the door,” said Mr. Singell, who is an associate editor of the journal Economics of Education Review. But he added that he was not certain that that is necessarily happening with the two colleges in Mr. Rizzo’s paper.

Mr. Rizzo and his colleagues “have made a reasonable start at looking at these issues,” Mr. Singell said. But he said that as far as he can tell, it is possible that the two colleges are in fact spending their financial-aid budgets efficiently, perhaps by bumping certain students’ likelihood of enrolling from, say, 80 percent to 90 percent.

“This is a question that we have thought about a lot at Oregon,” Mr. Singell said. “Who exactly is the marginal student? When is it that we should be giving them aid, in order to get them over the threshold? My guess is that a move from 50 to 60 percent [likelihood of enrolling] isn’t the same as a move from 80 to 90.”