How Much Should Public Universities Spend on Merit Aid?

Critics claim that public universities spend far too much on merit aid at a time when the focus should be on providing more need-based assistance, but the uses of merit aid are many, as are the reasons that drive the aid decisions at individual institutions.

The main problem for leading institutions, especially, is how to balance quality, access, state interests (including revenue), and public perception. In general, the most vocal critics of merit aid believe that access should trump all the other factors.

Two recent examples of that criticism come from the New America Foundation: Colleges’ Pursuit of Prestige and Revenue Is Hurting Low-Income Students and The Out of State Student Arms Race, both by analyst Stephen Burd.

First of all, some of the arguments in these and other reports are valid. For one thing, there is no doubt that the U.S. News rankings drive many colleges to spend money on generating better metrics, especially those related to test scores, selectivity, and student/faculty ratios. Some schools have become proficient in gaming the system.

The U.S. News methodology currently gives a combined weight of 9.25% to test scores and selection ratios. The use of the latter should be scrapped, given the increased use of the Common App and marketing geared to ramping up applications just for the sake of lowering acceptance ratios. (As for test scores, there are ways that colleges can game that metric as well.) The methodology also assigns a weight of 22.5% to multiple financial metrics that also pressure colleges to raise and spend more money.

State budget cuts and rising costs for instruction, research, and administration have also led to the need for more revenue. Just how much of the additional revenue is actually necessary for improved instruction is a matter of contention. (See for example Baumol’s Cost Disease and The Bowen Effect.) The combined effects of state disinvestment and the obsession with prestige and rankings have undoubtedly led to the intense focus on increasing revenues.

Yet after granting the critics a fair measure of credit, we come back to the four main factors that affect the allocation of merit aid, discussed below. And here’s a proposed standard for balancing the factors: If merit aid is denied to highly qualified, low-income students who are residents of the state, and goes instead to out-of-state students whose qualifications are about the same or less, then the merit aid is being used excessively for revenue purposes.

Quality–As noted elsewhere on this blog, the elite colleges and universities in this country, almost all of them private, simply do not have enough slots for the top 8-10% of students, based on test scores. Most of the highly talented students who are not accepted by elite private colleges will end up at public universities. Those public universities that allocate funds to support smaller classes and undergraduate research for talented students through honors programs, along with merit aid, are not only spending money to recruit students with higher test scores in order to enhance their prestige; they are also filling a real need by providing more slots for talented students. In addition, many are trying to keep talented students in state rather than seeing them leave, never to return. All too often, critics of public university spending ignore these needs.

Access–The relationship of merit aid to greater access for lower-income students is complex. Rankings and prestige have an impact on merit aid allocations, but that impact is not always what the critics see.

Some elite public universities (UC Berkeley, Michigan) offer higher percentages of merit aid than other public universities that are excellent but not so elite (although the average amounts of merit aid from Berkeley and Michigan are not especially large.) Why? The competition for UC Berkeley and Michigan includes many private elite schools, and sometimes even modest merit aid can be the deciding factor. Private universities such as Chicago, Northwestern, and Rice also offer significant merit aid, and do so to compete with the Ivies, Stanford, etc., who are so much in demand that they don’t have to offer non-need-based aid.

The University of Virginia and the University of North Carolina also compete effectively against private elites, but they have chosen to provide very limited merit aid.

For public universities at the next level, Washington, Illinois, UT Austin, Wisconsin, the competition is often with other publics, and they more than hold their own. Partly as a result of being in high demand regardless of aid, UT Austin has one of the highest enrollments of Pell Grant students and offers little in the way of merit aid.

But when it comes to public universities with lesser reputations than those listed above, the balance between aid for quality or aid for access may tip too far toward quality, sometimes with an eye on improving rankings and revenue. New America singles out the University of Alabama and the University of South Carolina for criticism. How much of this criticism is valid?

State Interests/Revenue–The state of South Carolina now funds only 10 percent of the cost of education at the flagship university. Moreover, the number of college-age students in the state is declining. New America criticizes the University of South Carolina for awarding too much merit aid to out-of-state students, who still end up providing more revenue out of pocket than in-state students, and also help to sustain enrollment levels.

If the university allocated most or almost all of its aid to need-based students within the state, the revenue would drop dramatically and the expense per student would rise. The university would probably be unable to support its excellent honors college; for that matter, the university would eventually be unable to serve as many students period. So even if the state legislature undervalues higher education, the university and many citizens believe it is in the interest of the state to increase the number of college graduates (and their families) over the long haul, and not diminish the university in the process.

Is the percentage of non-resident freshmen (45%) too high, and the merit aid they receive too much? To answer those questions, one would need to know (1) whether many highly qualified (but low-income) in-state students are not receiving aid because the aid is going to out-of-state students with equal or lesser qualifications; and (2) how many of talented out of state students will remain in South Carolina after graduation.  To the extent that highly qualified, low-income, in-state students are losing out, then the out  of state aid should be reduced.

Public Perception–Funding honors programs and offering merit aid to talented students can certainly increase the selectivity profile of a university and eventually enhance rankings and public perception. But we would draw a distinction between the aggressive gaming of the rankings and the more justifiable funding that is related to legitimate state interests. New America suggests that the extremely generous merit aid that the University of Alabama offers to talented out-of-state students is mainly to enhance rankings. But, contrary to what New America claims, the Alabama U.S. News ranking has actually fallen 13 places since 2012.

The University of Mississippi is another flagship that offers generous merit aid. What is also true is that the state of Mississippi has the second lowest percentage of college grads in the nation, and Alabama the 7th lowest. Surely these states should find ways to sustain their flagship institutions, and merit aid, for now, is one of those ways. Who knows but that some day they might join UVA, UNC Chapel Hill, UT Austin, Wisconsin and others that can go forward without so much emphasis on merit aid.

Again, we agree with New America that many universities, including some major public institutions, do use merit aid, at least in part, for purposes of moving up in college rankings and sometimes excessively for revenue purposes. But the total picture is much more complicated, resulting in one of the most difficult issues to emerge from state disinvestment in higher education.

Is It True that 80% of Elite Students Are Accepted by Elite Colleges?


Editor’s Note: The following post examines the claim by the New America Foundation that acceptance to elite colleges is not any more difficult than in the past, with 80% of highly qualified applicants still being accepted by elite schools. What we found is that even with the additional elite slots offered by honors colleges and programs, there are not enough slots to accommodate 80% of elite applicants. Although the acceptance rate as indicated by probability statistics (given multiple applications by each student) might show a high acceptance rate, the enrollment rate must be significantly lower than 80% for elite students. Updated with important changes, March 21, 2016; updated September 18, 2018, with references to new SAT scores.

Kevin Carey is director of education policy at the New American Foundation, and recently he has argued (1) that the end of college as we know it fast approaching, thanks to technology, and that (2) reaching a good college isn’t as hard as it seems.

Carey has little use for research universities, a “hybrid” form of higher education whereby students are offered the “bait” of being able to study with well-known research professors only to experience a “switch” once they are enrolled, finding themselves under the inept tutelage of callow TA’s and lower level faculty. The advent of MOOCs will give rise to the university of, and for, everyone, leading  also to the “brutal unmasking” of hybrid universities as the pretenders Carey believes them to be.

At the end of the post, there is a list of colleges, universities, and public honors programs with mean SAT scores of 1380 or above.

Setting aside Carey’s curious approach of proclaiming the end times for traditional colleges on the one hand while announcing the cheerful prospects for enrolling in these same colleges on the other, we will focus only on his assertion that approximately 80% of highly qualified college applicants are now able to gain acceptance to an “elite” college, never mind the dire comments on how much tougher it is now to get into such institutions.

According to Carey, “the slots themselves [at elite colleges] aren’t becoming more scarce and the number of students competing with one another isn’t growing.”As we suggest below, the problem with this statement is that the slots Carey is talking about seem to be plentiful because the colleges he includes are not as elite as the students he is tracking.

Therefore, while it may be true that 80% of applicants gain acceptance to the 113 colleges he is tracking, it is highly unlikely that there are in fact 113 schools whose mean test scores match the threshold scores of his applicant cohort, or that have sufficient slots to actually enroll those students. Indeed, the only way all of these students can attend colleges that even approximate their level of credentials is for many of the students to attend the very “hybrid” universities Carey criticizes.

(The New America Foundation also takes public universities to task for offering too much of their financial aid resources to students who are meritorious but not so needy as others. We will comment on this assertion in a later post.)

Carey defines elite applicants as those with old SAT scores of 1300 or higher (now ~1380], or with a comparable ACT score (29). Elite colleges are those that are among “the 113 schools identified by Barron’s Profiles of American Colleges as the most selective.”

We followed up on Carey’s assertion by doing some research of our own. Using the 25th percentile scores and the 75th percentile scores in the 2015 U.S. News rankings and dividing the total by two in order to arrive at an approximate average, we found that, rather than 113 colleges with average (old) SAT scores of 1300 or higher, there are only about 86. In addition, there are 51 public honors colleges/programs with mean (old) SAT scores of 1300 or higher. For purposes of illustration, we assumed that there was a like number of private honors colleges and programs, though we believe this is a generous estimate.

Then we calculated the number of students who took the ACT and SAT in 2014 and who score in the 91st or 92nd percentile or higher on one or both tests. An SAT score of 1300 is the 91st percentile; an ACT score of 29 or higher is also at the 92nd percentile level. We estimated that 25 percent of the 345,250 students took both tests; again, this is probably a generous estimate. This estimate should also take into account the relatively small number of high scorers who do not have high school GPA’s commensurate with their test scores.

The result: about 259,000 students met the test score threshold in 2014.

Next, we calculated that of the 86 colleges and universities with mean test scores at the threshold or higher, there were 131,077 places for freshmen in 2014. This leaves a deficit of 127,923 places for highly talented students, or more than 49%.

Using exact figures for public honors colleges and programs with similar test score profiles, we added another 20,917 places. Then we added another 20,917 places for private university honors colleges even though we doubt that there are that many honors places in private schools.

Therefore, the total number of “elite” places in 2014, including honors program places, was approximately 172,911. Subtracting this number from 259,000 still yields a deficit of elite places in the amount of 86,089.

New America claims 80 percent of elite students were accepted, but it is extremely unlikely that they could have actually found a place in the elite group of schools we have identified. So, if we take .8 x 259,000 students, the result is 207,000 students who should have been accepted by elite schools. Subtracting 172,911 places from 207,000 still leaves a deficit of elite places in the amount of 34,289 places. Again, the actual deficit is probably higher.

From these calculations, it appears that the acceptance rate for elite students, if they were to apply to schools with mean test scores at the threshold level or higher and not to less selective schools, would not be 80%, and it is very unlikely that there could be enough truly elite places for 80%. This is based on the calculation 172,911/259,000. That there is “room” for about 67% of elite students in elite colleges may not be a panic situation, but some elite students should be prepared to attend colleges whose mean scores are somewhat lower than their scores.

For the time being, there are solid alternatives at relatively low cost, at least for in-state students: those discredited “hybrid” institutions. Washington, Wisconsin, UT Austin, Florida, Penn State, UC Davis, UC Irvine, UC Santa Barbara, Ohio State, Georgia, Rutgers, Connecticut, Purdue, and Clemson may not have had average test scores in 2014 that meet the New America definition; but surely many of the elite students choose to attend these universities, in addition to those who are accepted by the more selective honors programs at most of these schools. Indeed, New America probably included these and other prominent public and private universities in its list of 113, despite their average test scores being somewhat lower than those of the elite students they tracked.

The estimated total number of freshmen slots at the 14 state schools above is about 104,771. At these schools, there are perhaps 40,000 to 52,500 freshmen who meet the elite definition.

This should leave no doubt about the need for these universities to sustain or enhance their current level of excellence: they are necessary, as are public and private honors programs, if top students are to actually “reach” schools that approximate their abilities.

Below are the universities, colleges, and public honors programs we have identified as having mean test scores of SAT 1300/ACT 29 or higher. Not included is a list of private honors colleges that meet the threshold; as noted above, we have estimated the number of students accepted by these honors colleges. The public honors colleges and programs are in listed in bold type at the end.

Carnegie Mellon
Boston College
Johns Hopkins
Notre Dame
Wash U
UC Berkeley
North Carolina
UC San Diego
Georgia Tech
Wake Forest
Case Western
William & Mary
Univ of Miami
Stevens Inst of Tech
Claremont McKenna
Washington Lee
Harvey Mudd
Bryn Mawr
Colorado College
Holy Cross
Franklin and Marshall
Mt. Holyoke
Connecticut College
South Carolina
UT Austin Plan II
Virginia Tech
Stony Brook
Ohio State
Penn State
Miami Oh
Texas A&M
UC Davis
Florida St
Ohio Univ
UC Santa Barbara
Michigan St
North Carolina St
UC Irvine
Univ at Buffalo
Colorado St
Oregon St
Oklahoma St
Arizona St