Left Out of the Elites

Low income, high performing students are plentiful enough that selective colleges should be able to admit more of them, study finds.
December 8, 2005

Elite institutions like Harvard, Yale and Princeton should increase by 30 percent the number of low income students they admit – and they can do so without lowering the quality of their student bodies, a new report by researchers at Williams College asserts. The report’s authors and some other scholars are confident that such shifts in policy will happen -- probably not fast enough for advocates for underprivileged students, but perhaps too quickly for some in the upper class.

A recent discussion paper by the Williams Project on the Economics of Higher Education indicates that if “high ability” is defined as an SAT score of 1300 or above, then highly selective private colleges -- the 31 leading institutions that conduct joint financial aid surveys for comparative purposes -- should move collectively from having 10 percent to nearly 16 percent of their students come from the bottom two family income quintiles (those who made less than $41,001 as of 2002).Therefore, instead of about 2,750 low income students matriculating at those colleges per year, there should be about 4,300.

The paper also notes evidence -- based on analyses of SAT scores -- that there are enough low-income students of “high ability” in the United States for the institutions to be able to admit enough such students to largely mirror the distribution of such students nationally. College officials sometimes explain the relative underrepresentation of students from low-income families on a dearth of high performing students -- but Gordon C. Winston,a professor of economics at Williams who co-wrote the paper, says the research suggests that's not the real problem.

“We do have an issue of disparity,” says Winston. “We have to ask why this is happening and how we are excluding low-income folks in ways we don’t know.”

Winston says that an “overreliance on the SAT” in the application and admissions processes at private institutions may be one way that certain income populations are being excluded.  “We do most of our searches for students based on SATs, but this neglects a whole part of the country where the ACT is paramount,” he says. “Is this intention or diabolical?” 

Disparities also exist in terms of cost, according to a new study led by Winston, which appears latest issue of The Journal of Human Resources.  It found that although many highly selective colleges tailor their tuition and other prices to students' financial means, students in the lowest family income quintile spent nearly half of their annual family income on one year at an elite college or university. Students in the other four income groups spent an average of between 21 percent and 26 percent of their annual family incomes on one year's college costs.

The study also shows that the lower the students' family income, the smaller percentage of the "sticker price" -- which, on average, amounted to almost $34,000 at elite private institutions in  the 2001-2 academic year -- the students paid. Students in the lowest income quintile, which had a median annual family income of $15,347, paid an average of $7,552 (22 percent of the sticker price) in 2001-2, reports Winston. 

"The kids in the bottom quintile are getting a significant price reduction, but relative to their family income, it's not nearly the price reduction that other kids are getting," says Winston. "It's really encouraging, though, that among these expensive, elite schools, the proportion of family income paid is falling precipitously for low income students."

When it comes to the smartest of the smart poor students (at least in terms of performance on the SAT), Winston notes that selective private institutions, collectively, do very well in terms of admission. That is, the poorest students who scored a 1600 on their SATs in 2003 were disproportionately found in selective private colleges as opposed to other institutions.

However, moving from the poorest to second poorest income quintile of students who score that high, selective private colleges’ relative share of low-income students drops.  The paper notes that middle and upper-middle income students who score a 1600 are markedly underrepresented at such institutions, while those from the highest income families are overrepresented. A forthcoming paper will examine this "middle class issue," says Winston.

Chris Avery, a Harvard professor of public policy and management, says that as more low- and middle-income students are recruited to selective privates -- which he says is “inevitable” in light of such research -- upper-income students will face increased competition for a limited number of seats.  But, he notes, such students have long held advantages in higher education over lower-income groups, in areas such as counseling, academic preparation and familial legacy connections.  “These advantages of class are slowly being eroded,” he says. 

“I think it’s a policy issue,” Winston says of this matter. “If we argue that lower-income kids should have a larger share, then we have to ask how we’re going to go about it and at what cost?”

“That’s a tough one” for college officials, he says.    

The paper’s authors note limitations in their analyses.  For instance, anywhere from 35 to 50 percent of “high ability” students who take the SAT or ACT do not report family income. Still, the authors note that studies have found low-income students to be less likely to report family income than high-income students.  Therefore, conclude the authors, non-reporting likely leads to an understatement of the population of low income, “high ability” students. 

Jay Rosner, executive director of the Princeton Review Foundation, has concerns about the paper’s definition of “high ability” and how it relates to college performance. “1300 might sound reasonable in the abstract; however, there are many thousands of inner-city kids and rural kids scoring hundreds of points lower who could be successful at highly selective colleges,” he said via e-mail. “One example is Cedric Jennings, profiled in Hope in the Unseen by Ron Suskind. Cedric scored under 1000, and was successful at Brown as an undergrad.  He has a master’s from Harvard.  Of course, he was/is unusually strong and motivated, but hardly unique.”
“In my opinion and experience, a difference of as much as four or five hundred points on the SAT [math and verbal] may or may not be significant in indicating potential for success in college,” he said, noting that SAT scores only correlate somewhat with first year college GPA’s.

Winston, however, says such critiques are unwise if the goal is to have more low-income students enrolled at private selective universities.  “There are plenty of low-income well-qualified students out there,” he says.  “We don’t have to lower the bar -- we have to change the policies.”


Back to Top