In the early 1990s, when merit aid first began to flourish, the rule of thumb was that $1,000 could tip the decision for a student on the fence between two institutions; $4,000 would get him to accept his second choice, and $6,000 his third choice. (Colleges can determine students' preferences from the questionnaires that all students fill out before taking a College Board or ACT test.) Armed with this knowledge, private schools went on a spending spree, buying up "meritorious" students (defined as gifted or rich, but ideally both). And public schools jumped at the technique as well: from 1992 to 2000 the proportion of state aid based on merit rose from just under 10 percent, where it had hovered throughout the 1980s, to 25 percent. Since 1990 the average discount on "sticker-price" tuition has risen from 26.5 percent to just under 40 percent, and the proportion of students not paying full price has grown from 62.5 percent to 80 percent. The resulting confusion about the actual cost of college gives the enrollment manager a free hand to manipulate prices for individuals.
To decide how to parcel out financial aid, the enrollment manager puts admitted students onto a grid with need on one axis and academic ability on the other. This is called "segmenting the class" or "table analysis." The school then adjusts financial aid for students by group, with the goal of increasing the "yield rate" for the most desirable prospects—typically academic stars and those willing to pay most or all of the tuition ("full-pays"). A school with a revenue problem puts its money toward rich students; a school that's going after prestige pushes it toward students with high SAT scores. Where the school might be paying more than is necessary to attract a candidate (for a wealthy student with low grades, for instance, or an in-state student with few other options), aid is cut accordingly. Some schools are content to fine-tune their financial-aid packages for different groups by trial and error from year to year. But more-advanced enrollment managers, and all the major consulting companies, use a statistical method called logistical regression to determine how each group will respond to a different award, based on how students have behaved in the past.
One of the basic texts of enrollment management recommends a book about pricing techniques developed by the airlines: Revenue Management: Hard-Core Tactics for Market Domination. Using the logic of the Saturday-night stay and the fourteen-day advance purchase, advanced financial-aid leveraging goes beyond general categories to forecast how much each student is willing to pay, and guarantee the best class at the lowest price. Schools and consultants combine test scores, grades, and class rankings from the testing services and students' high schools with demographic and financial data purchased from a credit-reporting agency such as Equifax. All this information is eventually reduced to the seven or eight variables that best predict a student's responsiveness to price.
In the least desirable categories (usually poor students with lower test scores) accepted students are often "gapped"—given a fraction of what they would need to attend, even after the maximum possible contribution from their families. (A school interested mainly in revenue might even give more money to a wealthy student with lousy scores than to a better-qualified poor student.) Some schools leave gaps as high as $34,000 a year. From 1995 to 1999 the average unmet need for families earning over $60,000 either stayed constant or narrowed slightly; for families earning $40,000 to $60,000 it grew by three percent; and for families earning under $40,000 it grew by 27 percent. Some schools have no choice but to gap students once they've exhausted their aid budgets. Others will intentionally gap poor students so severely that they decide not to attend in the first place—or, if they enroll, the long hours of work-study and mounting debts eventually force them to drop out. Called "admit-deny," this practice allows a college to keep poor students out while publicly claiming that it doesn't consider a student's finances when making admissions decisions.
"Admit-deny is when you give someone a financial-aid package that is so rotten that you hope they get the message: 'Don't come,'" says Mark Heffron, a senior vice-president at Noel-Levitz, one of the largest enrollment-management consulting companies. (His financial-aid division currently has 140 clients.) Unfortunately, "they don't always get the message." When consulting for a school that gaps students to a point where they are likely to drop out, Heffron encourages schools to call students and tell them that unless they can find an additional source of money—such as a generous relative—they should decline the offer of admission.
However nasty, admit-deny allows schools to avoid the controversy associated with publicly abandoning need-blind admissions. That students are rejected on the basis of income is one of the most closely held secrets in admissions; enrollment managers say the practice is far more prevalent than most schools let on.
"Good luck getting any institution to tell you exactly how they handle ability to pay as a driver in their admit decision," said one enrollment manager who requested anonymity. "What they will say is 'We're need-blind.' That's bullshit. They would never tell you exactly how they do it, but they do it all the time."
Schools also use detailed data to systematically cut aid to students whose behavior shows they are likely to accept admission anyway. A student who lists a school first when asked where to send test scores, files for financial aid with that school, and then visits campus has tipped his hand, and some schools will figure, why waste money to attract a sure thing? Understandably, schools don't publicize this practice: in 1996 Johns Hopkins found itself on the front page of The Wall Street Journal for even considering it. The incident remains notorious in enrollment-management circles.
"I think that's self-defeating, and first of all I don't like the morality of it," Heffron told me. "It's one of those things where if anyone finds out you're doing it, you're dead."
The demand for revenue and prestige also increasingly controls the earlier, mass-marketing and recruitment stages of the admissions "funnel," by which a student goes, in the industry lingo, from "suspect" to "prospect" to "admit" to "matric." The ACT and the College Board don't just sell hundreds of thousands of student profiles to schools; they also offer software and consulting services that can be used to set crude wealth and test-score cutoffs, to target or eliminate students before they apply.