Class Notes: A Story of Bias in 2003 and Today

I told you I’ve been covering class bias in higher education for a long time. I was going through some old files when I found this unpublished piece on from 2003. I had written the original version in 1999 and kept trying to place it, updating it as new events and research warranted.

I used different elements from this piece in shorter articles for the Progressive Populist in 2004 and Razorcake in 2009. Stylistically, I prefer those later pieces, but the 2003 version includes details I still haven’t seen anywhere else. In this article, I argued that class-based affirmative action should be added to other forms of affirmative action. Everyone else at the time argued that we could either keep race as a criterion or replace it with class. (Gender-based affirmative action typically did not appear in those debates. Most colleges asserted that they did not practice gender preferences at the undergraduate level, which was the main locus of debate on affirmative action.)

Today, most writers on affirmative action still play the class-versus-race card. They also ignore all the evidence showing that college administrators have extreme prejudices against working-class applicants. In this piece, I discussed that problem in the context of social profiling. I believe the article was ahead of its time. It still is, in fact.

Anyway, here it is. Many of the links in the footnotes are out of date, but I left them in order to show the original sources.

Class Notes: Social Status and Affirmative Action (2003 version)

The case for including socioeconomic status within affirmative action criteria has suffered more from its advocates than from its critics. Class-based affirmative action is rarely discussed, except when conservatives offer it as an alternative to race-based affirmative action, as in the Bush administration’s brief in the Michigan Law School case.1 Of course, such maneuvers are purely tactical: conservative interest in the college-admission prospects of working-class and poor applicants never goes beyond the level of gestures made during disputes over race-sensitive admissions.

Despite the neglect of liberals and the shameless ploys of conservatives, we urgently need a remedy for social discrimination. Also, broadening the scope of affirmative action to include class as well as race would shore up support for the policy at a time when it is in serious danger.

The increasing gap between rich and poor in America is nowhere more evident than in the area of college education. A 1998 Presidential Commission on the Cost of Higher Education reported sharply rising costs at undergraduate institutions, especially the private ones that increasingly dominate the US News & World Report rankings and other measures of prestige.2 In 1980, tuition at the average private university was $2,971 higher than tuition at the average public institution. By 1996, that gap had increased to $12,430.

Due to upward redistribution of wealth since 1980, working-class and poor families have borne the brunt of tuition increases at both public and private institutions, while rich families have taken tuition hikes in stride. Between 1980 and 1996, the ratio of average tuition price to household income approximately doubled for the poorest 20% of families, while staying almost exactly the same for the richest 20%.3

Likewise, education scholars Michael McPherson and Morton Shapiro discovered that cuts in the federal Pell Grant program and the failure of university-based scholarships to keep up with tuition have left working-class applicants especially vulnerable to rising tuition costs. Based on enrollment figures from the mid-1970s to the mid-1990s, they found that upper- and middle-class representation at colleges was unaffected by tuition increases, while percentages of lower-income students (defined as those from households with incomes below $20,000 in 1990 dollars) declined sharply as tuition rose.4

Perhaps the best source of information on the admissions practices and demographics of top universities is The Shape of the River (1998) written by former Princeton President William G. Bowen and former Harvard President Derek Bok, along with four other authors. In the book, the authors effectively countered objections to race-based affirmative action, but an unintended product of their work was to reveal the extent of social elitism at top universities.

Bowen and Bok drew their statistics from the Mellon Foundation’s “College & Beyond” database, a collection of information obtained from admissions statistics and student questionnaires That information came from 28 colleges and universities, heavily weighted toward elite private institutions, in three separate years: 1951, 1976 and 1989. Four Ivy League universities are included in the list, as are Stanford, a number of prestigious liberal arts colleges, and two of the highest rated public universities: Michigan and North Carolina.5 Subsequent discussion pertains to the most recent College & Beyond study (1989), unless otherwise indicated.

One aspect of admissions policy examined by Bowen and Bok was the practice of giving preference to children of alumni or “legacies.” Legacy admissions made the news during the debate over the Michigan affirmative-action case, and administrators at elite universities have been eager to downplay their significance. The Princeton University Admissions Office acknowledges that its committee takes legacy status into consideration, “when such an applicant is in the top part of our applicant group,” but adds that, “no student is admitted simply because he or she is the offspring of a Princeton undergraduate.”6

The Brown University Admissions Office states:

All other things being equal, qualified applicants from families that have a relationship with Brown may be at a slight advantage. Please keep in mind that happenstance of birth alone will not get one admitted to Brown; academic and personal achievement and promise will.7

Using admissions data obtained confidentially from three unspecified private institutions in the College & Beyond database, Bowen and Bok compared 1989 admission rates for African-Americans to those for white legacies and white non-legacies. (The authors did not compare data for African-American legacies and non-legacies, probably because the former category was too small to constitute a significant sample.)

White legacies were admitted at a greater rate than African-American applicants (44% to 39%) and at twice the rate of white, non-legacy applicants (44% to 22%). Among applicants who scored 1300 on the SATs, legacies were two-and-a-half times as likely to be admitted as non-legacies (60% to 24%).8 In fact, non-legacies with SAT scores above 1300 had about the same chance of admission as legacies who only scored in the 1100s (24% versus 22%).9 There is nothing “slight” about the advantages of legacy applicants.

In a counterpoint to their discussion of legacy admissions, the two former university presidents maintained that “the C[ollege] & B[eyond] schools seek to enroll individuals from disadvantaged backgrounds.”10 Likewise, in their friend of the court brief in the Michigan admissions case, lawyers for eight institutions (Brown, Chicago, Dartmouth, Duke, Harvard, Penn, Princeton and Yale) argued that their schools already applied a sort of affirmative action for working-class and poor applicants. “Admissions officials give special attention to, among others, applicants from economically and/or culturally disadvantaged backgrounds.”11

If that is so, Bowen and Bok’s statistics suggest that there is a wide gap between “giving special attention to” economically disadvantaged applicants and actually admitting them. The two authors compared enrollment figures for blacks and whites from different economic backgrounds in the 1989 College & Beyond cohort. They divided students into three economic categories, according to annual household income and whether or not the students’ parents held bachelor’s degrees. The categories were:

  1. High status, consisting of students from households with annual incomes above $70,000 and where at least one parent was a college graduate.
  2. Low status, consisting of those from households with incomes below $22,000 and where neither parent was a college graduate.
  3. Middle status, encompassing all others.12

Using census data, Bowen and Bok found that approximately 9% of Americans fit the high status category, while roughly 64% were middle status, and 28% low status (the figures do not add up to 100% because they are rounded off). At institutions in the Mellon database, the enrollment breakdown was as follows. The percentages for black students appear in the first column, percentages for whites in the second.

High status          15%               44%

Middle status      71%               54%

Low status           14%                  2%13

The authors called the small percentages of low-status students “striking,” and the fact that 50% of all African-Americans fall into the low-status category makes it even more so.14 A chart in the book’s appendices revealed that the percentage of low-status students at the Mellon schools was cut approximately in half between 1976 and 1989. In 1976, 4% of white students at College & Beyond database schools fit the low status category, compared to 2% in 1989. The proportion of black students from low-status backgrounds fell from 26% in 1976 to 14% in 1989. At the same time, the percentage of students from the top social category increased from 40% to 44% among whites, and from 11% to 15% among blacks.15

Turning to the most selective institutions in the database, Bowen and Bok declined to provide percentages for all income groups, but did specify that 12% of blacks at those schools, and only 1% of the whites, came from the low status category.16 It is unclear just how small the latter percentage is, because the authors always rounded off to the nearest whole number, which leaves the possibility that the 1% figure is actually rounded up. For comparison, Andover, President Bush’s prep school, accounted for 1.2% of full-time undergraduates enrolled by Yale between 1998 and 2001, even though Andover’s enrollment is under 1,100.17

Bok and Bowen’s social statistics appear yet more disturbing when we consider that officials at top institutions claim to place a high premium on the educational advantages of diversity. Such advantages are justifiably among the chief reasons cited for maintaining affirmative action. The friend-of-the-court brief submitted by Harvard and seven other institutions states that admissions officers at those schools, “have always sought to enroll a broad cross-section of students who can bring a critical mix of experiences and perspectives into the university community.” The lawyers also added this statement: “Diversity helps students confront perspectives other than their own and thus to think more rigorously and imaginatively; it helps students learn to relate better to people from different backgrounds.”18

That is true with regard to race, and equally true with regard to class. How are classroom discussions of poverty or economic policy at the most selective institutions influenced by the fact that 99% of white students and 88% of black students are upper- or middle-status, as defined by Bowen and Bok? Does the near total absence of low-status students at the most prestigious colleges encourage undergraduates at those schools to accept or reject the stereotype that poor people are lazy? Would not the addition of an economic category of affirmative action improve social diversity among black and Hispanic students at top colleges, and also concentrate the benefits of the policy more directly on the minority applicants who have suffered the worst discrimination?

Despite the grim evidence of social exclusion they uncovered, and its negative implications for universities’ quest for diversity, Bowen and Bok came out against class-based affirmative admission. They offered statistics that proved that minority enrollments would fall if class-based affirmative action replaced affirmative action by race, but they never considered the possibility that the two types of affirmative action could operate in conjunction.

Likewise, once the subject shifted from race to class, the two authors stopped basing their judgments on hard data and rigorous analysis. At one point they wrote: “Although some critics may believe that universities do not try hard enough to find qualified low-income applicants, this charge is probably unjustified.”19 Probably?

Nowhere in the main body of the book, nor in the 150+ pages of statistical appendices, did the authors test that claim, but evidence against it can be found in a source they cite extensively: an article by economist Thomas Kane, studying the results of achievement tests taken by graduating high school seniors in 1992. Kane analyzed the backgrounds of students who scored in the top 10% of all test-takers in both math and reading, separating the students by race (blacks and Hispanics in one group, whites in the other) and by two broad income categories (those from households with annual incomes above $20,000 and those with incomes below that figure).20

One finding of the study was that among whites who took the test, 6.1% of the top-scorers came from households in the below-$20,000 category.21 Those students overcame a great deal to outperform many test-takers from more privileged backgrounds, and that is precisely the sort of achievement admissions officers claim to be looking for in their applicants.

Recall, however, that only 1% of white students at the most selective College & Beyond institutions came from Bowen and Bok’s low-status category ($22,000 income and neither parent with a BA–comparable to the bottom social category as defined by Kane). That disparity alone indicates that there are many more highly qualified, low-income candidates for admission than top universities are currently enrolling. Another of Bowen and Bok’s objections to class-based affirmative action was that the policy was “probably prohibitively expensive.”22

That plea of poverty is especially odd, considering the enormous growth in top universities’ endowments during the 1990s. A 1997 list of university endowments showed that every Ivy League institution except Brown possessed an endowment of over $1 billion (Brown’s endowment measured $949 million). Harvard was first in the nation at just under $11 billion, and a year later that figure had risen to $13 billion. A study of endowments at all American universities in 1995-96 showed an average annual growth rate of over 17%.23

When building their endowments, colleges and universities benefit immensely from the tax exemptions that come with their non-profit status. In fact, the legal requirements for receiving a tax exemption are more lax for institutions of higher learning than for registered charities and other private, non-profit institutions. Unlike other non-profits, colleges and universities are not required to spend 5% of their endowments annually in order to maintain their tax-exempt status, and many colleges do not spend that amount.24

Private universities got a scare in 1995, when a Pennsylvania court revoked the state tax exemption for Washington and Jefferson College, an expensive private institution. The court found that the college failed to meet a number of criteria normally required of non-profit organizations. Particularly, it did not offer enough of its services free of charge nor did it benefit enough persons who were worthy of charity.25 An appeals court overturned the ruling on a 4-3 vote, but the question posed by the lower court is apposite: why should taxpayers face higher taxes in order to provide exemptions for institutions that are not only quite wealthy, but are truly open only to well-to-do applicants?

In a final attack on the idea of class-based affirmative action, Bowen and Bok offered this warning to admissions officers: “students with low socioeconomic backgrounds are less likely than students of equivalent ability from high socioeconomic backgrounds to complete their studies, attain professional or doctoral degrees, and earn high incomes.”26

Of course, that piece of social profiling ignores many basic social factors that affect graduation rates and lifetime income. Most obviously, students from low-income backgrounds are more likely to have to drop out of school and find jobs in order to help support their families. Wealthy students with strong family connections in business or the professions hold obvious advantages over those without them, and elite fraternities, sororities and clubs allow rich students to make valuable contacts that improve career prospects and increase earning power. A particularly egregious case of the second phenomenon is Bowen’s own Princeton University, with its long tradition of socially exclusive dining clubs.

Bowen and Bok would not treat the obstacles faced by many African-American students as a reason not to admit them in the first place. However, they did just that in the case of low-income applicants from all racial backgrounds. Given their standing, they must have known that their book was going to be essential reading in college admission offices, and it was utterly irresponsible for them to sanction prejudices against economically disadvantaged applicants. Despite its effective defense of race-based affirmative action, The Shape of the River reveals the problems that stem from leaving affirmative action policy, and the defense of it, to university administrators.

It is time for progressives in the union movement, the civil rights movement and the press to take up the matter. The tools for implementing class-based affirmative action are already in place on American campuses. Colleges already collect financial information from applicants, and Bowen and Bok’s criteria of parental income and degree status make a good starting point for identifying economically disadvantaged candidates, though it would also be useful to distinguish between parents who are manual laborers and those whose occupations are non-manual. Of course, the new policy would spell the end of “need blind” admissions at the institutions where that practice exists. However, given the many indicators of class contained in the information available to admissions offices (home town, high school, legacy status), the idea of need blind admissions is chimerical anyway. And attempts to follow that policy have failed to make colleges more socially inclusive.

The best long-term solution would be to add socioeconomic status to the categories of discrimination prohibited by civil rights laws. Under the 1987 Civil Rights Restoration Act, institutions that receive federal funds, including those that participate in guaranteed student loan programs, are subject to US civil rights legislation. The federal government could therefore ban legacies and set goals and timetables for increased enrollment of economically disadvantaged students any time the country has a president and Congress willing to start treating socially exclusive practices at universities as a form of invidious discrimination. So far, politicians have shown little interest, though Senator John Edwards’s proposal to end legacy preferences is a welcome start. Another means of promoting national discussion on the topic would be to focus attention on the tax exemptions granted to rich colleges and universities, and on the small percentage of well-to-do taxpayers whose children get to attend those institutions.

All signs indicate that the Republicans intend to use affirmative action as a major issue in the 2004 campaign. If supporters of affirmative action do not find a way to counter the administration’s playing of the race card, they can look forward to increasingly effective employment of such tactics by conservatives in years to come. Campaigning for the expansion of affirmative action would allow progressives to blunt the edge of Republican racial politics, and also highlight the GOP’s love of social privilege. More fundamentally, such a plan would make a start on the first order of business for progressive politics today: finding ways to unite working-class people currently divided along racial, urban-rural or other lines.

  1. No. 02-516 In the Supreme Court of the United States. Jennifer Gratz and Patrick Hammacher, Petitioners v. Lee Bollinger et al. Brief for the United States as Amicus Curiae supporting petitioners (On-line press release of the Office of the Solicitor General:, 7.
  2. In US News & World Report “America’s Best Colleges” (2004), the top 20 universities are all private.
  3. Straight talk about college costs and prices: Report of the National Commission on the Cost of Higher Education (Phoenix, AZ: National Commission on the Cost of Higher Education and Oryx Press, 1998), 160.
  4. Michael S. McPherson and Morton Owen Shapiro, The Student Aid Game: Meeting Need and Rewarding Talent in American Higher Education (Princeton, NJ: Princeton University Press, 1998), 38-49, 136-144; Michael S. McPherson and Morton Owen Shapiro, “Does Student Aid Affect College Enrollment? New Evidence on a Persistent Controversy,” American Economic Review, Volume 81, No. 1 (March, 1991), 309-317.
  5. William G. Bowen and Derek Bok, in collaboration with James L. Shulman, Thomas L. Nygren, Stacy Berg Dale, and Lauren A. Meserve, The Shape of the River: Long-term Consequences of Considering Race in College and University Admissions (Princeton, NJ: Princeton University Press, 1998), xxviii-xxix The twenty-eight institutions in the database are seventeen research universities (Columbia, Duke, Emory, Miami (Ohio), Michigan, North Carolina-Chapel Hill, Northwestern, Pennsylvania, Penn State, Princeton, Rice, Stanford, Tufts, Tulane, Vanderbilt, Washington University in St. Louis and Yale) and eleven liberal arts colleges (Barnard, Bryn Mawr, Denison, Hamilton, Kenyon, Oberlin, Smith, Swarthmore, Wellesley, Wesleyan, Williams). All the institutions are private except the following four: Miami, Michigan, North Carolina and Penn State.
  8. Bowen and Bok, 28, 28n.
  9. Ibid.
  10. Ibid., 48.
  11. Nos. 02-241 and 02-516 In the Supreme Court of the United States. Barbara Grutter, Petitioner  v. Lee Bollinger, et al; Jennifer Gratz and Patrick Hamacher, Petitioners v. Lee Bollinger, et al. Brief of Harvard University, Brown University, the University of Chicago, Dartmouth College, Duke University, the University of Pennsylvania, Princeton University, and Yale University as Amici Curiae supporting respondents (On-line press release of Harvard University:, 20.
  12. Bowen and Bok, 48n.
  13. Ibid., 48-49.
  14. Ibid.
  15. Ibid., 341.
  16. Bowen and Bok, 40, 49-50. The selectivity rating was based on the average SAT scores of enrollees. The colleges and universities in the “most selective” category were Bryn Mawr, Duke, Princeton, Rice, Stanford, Swarthmore, Williams and Yale.
  17. That percentage is based on college placement figures made public by the Andover College Counseling Office ( and on Yale’s enrollment in 2001 as recorded in US News & World Report’s America’s Best Colleges (2001).
  18. Brief of Harvard University, etc., 19, 8.
  19. Bowen and Bok, 50.
  20. Thomas J. Kane, “Racial and Ethnic Preferences in College Admissions,” in Christopher Jencks and Meredith Phillips (eds.), The Black-White Test Score Gap (Washington, DC: Brookings Institution Press, 1998), 449-451.
  21. Kane, 449.
  22. Bowen and Bok, 50.
  23. Ronald G. Ehrenberg, Tuition Rising: Why College Costs So Much (Cambridge, Mass.: Harvard University Press, 2000), 36-38.
  24. Ibid., 269.
  25. Ibid., 268-69.
  26. Bowen and Bok, 271.