HOME SHORT TAKES OUR ESSAYS PODCASTS LINKS ABOUT US CAU  Subscribe MTC on Facebook  Find us on Twitter

OUR ESSAYS


September 2, 2010

The Underperformance Problem

By Russell K. Nieli

On average black students do much worse on the SAT and many other standardized tests than whites. While encouraging progress was made in the 1970s and early 1980s in improving black SAT scores and reducing the black/white test score gap, progress in this direction came to a halt by the early 1990s, and today the gap stands pretty much where it was twenty years ago. Whereas whites and Asians today average a little over 500 on the math and reading portions of the SAT, blacks score only a little over 400 -- in statistical metric a gap of a full standard deviation. Only about one in six blacks does as well on the SAT as the average white or Asian.

This state of affairs is well known uncomfortable though it may be to bring up in public. Less well known is what in the scholarly literature is called "the underperformance problem." Once in college blacks with the same entering SAT scores as whites and Asians earn substantially lower grades over their college careers and wind up with substantially lower class rankings. This gap in grade performance, moreover, is not reduced by adding high school grades or socio-economic status to the criteria for matching students. Blacks equally matched with whites or Asians in terms of their entering scholastic credentials and socio-economic backgrounds simply do not perform as well as their Asian and white counterparts in college. And the degree of underperformance is often very substantial.

This is contrary to what many people have been led to believe. Standardized tests are "culturally biased," it is said, and do not fairly indicate the abilities or promise of racial minorities growing up outside the dominant white, middle-class, Anglo-Saxon culture. Often this claim is bolstered by reciting items on long outdated verbal tests asking for the meaning of words like "regatta" or "cotillion" that only upper-class whites are likely to know. The implication is usually that those from minority cultures will do better in college in terms of grades than their test scores would predict. The "cultural bias" argument, however, is not only questionable on its face -- since the clearly non-Anglo Saxon Asians do better than whites on most standardized tests of mathematical abilities including the SAT, while the equally non-Anglo Saxon Ashkenazic Jews outperform everyone else on tests of English verbal ability -- but fails to account for the fact that in terms of grade performance blacks in college consistently do worse, not better, than their standardized test scores would predict. Standardized tests such as the SAT and ACT overpredict, not underpredict, how well blacks will do in college, and in this sense the tests are predictively biased in favor of blacks, not against them.

This "underperformance problem" has been well documented for over forty years. It first received widespread attention with the publication in 1985 of Robert Klitgaard's outstanding monograph, Choosing Elites, which dealt with admissions to America's most prestigious colleges and professional schools. It was also an important topic in William Bowen and Derek Bok's The Shape of the River (1998) -- the highly influential defense of racial preference policies at elite universities by former presidents of Princeton and Harvard. It was an important topic, too, in the two subsequent River Books sponsored by the Andrew W. Mellon Foundation -- The Source of the River (2003), by Douglas Massey and his colleagues, and Taming the River (2009), by Camille Charles and her colleagues.

The scope of the underperformance problem is suggested in The Shape of the River. Using the College and Beyond (C&B) database consisting of detailed statistics on students at 28 highly selective liberal arts colleges and research universities, the authors found that blacks and whites were separated by a rank-in-class gap of 30 percentile points -- white students averaged at the 53d percentile, blacks at the 23d. However, less than half of this rank-in-class gap (14 points) could be explained in a complex regression analysis by factors such as the blacks' lower entering SAT scores, their lower high school grades, and numerous other relevant background characteristics. The remaining gap (16 points) had to be chalked up to "underperformance" -- i.e., to the fact that blacks simply didn't do as well as whites even when their many differences in entering characteristics were closely controlled for statistically. This same pattern, Bowen and Bok went on to explain, "holds within each group of C&B schools considered separately, whether the schools are classified solely by their degree of selectivity, by their status as a college or university, or as a public or private institution. At the various sets of schools, the adjusted black-white gap in class rank that remains after controlling for other variables ranges from 15 to 21 points." "This unmistakable pattern," Bowen and Bok add, "is found not only in the C&B colleges but in professional schools as well."

Stereotype Threat

Explanations for the underperformance problem have been elusive. The most frequently discussed idea is psychologist Claude Steele's theory of "stereotype threat." Much of the problem that black college students encounter in achieving grades commensurate with their proven abilities, Steele believes, is due to their tendency to internalize the image of black intellectual inferiority projected by the wider culture. As a result of this internalization, blacks often experience heightened test anxiety, Steele says, and do more poorly on college exams than would be the case in the absence of such a race-linked handicap. Steele and other researchers have shown within laboratory settings that black students do considerably worse on tests if they are told that the test is designed to determine their intellectual abilities than if the test is presented in a less threatening manner as a simple exercise in showing the mechanics of problem solving. (While Steele will only say so when pushed, our current system of racial preferences that places blacks in institutions where they must compete with whites and Asians who have usually gained entrance under more exacting standards almost surely contributes to this stereotype vulnerability.)

The stereotype threat theory, however, is at best only a partial explanation of the underperformance conundrum. Two salient and undisputed facts limit its explanatory power. First, the problem of underperformance has been shown to extend far beyond the kinds of majors where test-taking is a central part of a student's overall evaluation. In many humanities, social science, and history courses written assignments and term papers -- not timed, in-class exams -- are the basis of a student's academic assessment but the underperformance of blacks has been shown to be an across-the-board phenomenon that is found even among black students majoring in these areas. Though there is some evidence that the underperformance problem may be more serious in test-taking courses, especially those in the natural sciences, there surely is more at work here in depressing black grades than test-taking anxiety.

An even more serious problem with the stereotype threat explanation is the fact that the greatest gap in college grades between blacks and whites with similar SAT scores and high school grades is found among those students with the highest measures of these background characteristics. Blacks with the highest SAT scores and high school GPAs, many of whom have scores higher than the typical white or Asian student at the colleges they attend -- and thus have good reason to feel intellectually superior, not inferior, to most of their classmates -- are those underperforming the most. Those least likely to be troubled by feelings of intellectual inferiority or lack of academic self-confidence show the greatest performance gap when matched with whites and Asians of equivalently high SAT scores and high school grades.

Stereotype threat theory cannot explain any of this. The problem is laid out most succinctly by William Bowen and Frederick Vars in a widely read study of blacks and whites at 11 highly competitive colleges and universities (Bowen and Vars use a subset of the same College and Beyond database used by Bowen and Bok in The Shape of the River). Summing up their study's findings, they write: "At every level of SAT score blacks earn lower grades than their white counterparts, and this remains true after controlling for other variables, including high school grades and socioeconomic status. ... African-American students have lower GPAs than one would predict on the basis of SAT scores and high school grades. In our analysis this overprediction ... is present among both males and females, and in both colleges and universities. While it is most pronounced in the sciences, it exists in all major fields of study. ... Most sobering of all, the performance gap is greatest for the black students with the highest SATs. ... In selective colleges and universities, black students at the highest levels of SAT score are especially likely to underperform relative to white classmates with similar scores and characteristics." Bowen and Vars highlight their analysis by saying that "the reasons for this gap are not well understood."

The Disincentives of the Affirmative Action System

While acknowledging the probable effect of "stereotype threat," authors like Bowen and Vars realize that such an explanation goes only so far and cannot explain why blacks in all college subfields do worse than comparable whites and Asians, nor why it is the smartest blacks who underperform the most. The underperformance problem, however, is not nearly as enigmatic as these and other writers contend. Its most fundamental cause has been well explained in the past by critics of affirmative action in a manner that anyone with common sense can easily grasp. The real culprit here in discouraging black students to do their best and perform up to their potential, these critics explain, is the all-pervasive system of racial preferences that makes it much easier for black high school students to get into good colleges, and for black college students to gain access to good graduate schools, professional schools, and corporate sector jobs. The black students know that they don't have to do nearly as well in school as their white and Asian classmates to reach the same level of acceptance and achievement

The Manhattan Institute scholar John McWhorter, a linguist by training, explains the situation in its most cogent terms drawing from both his own experience as a black high school student in a mixed-race Philadelphia private school and from his many years of teaching black college students at Berkeley. "I can attest," McWhorter writes, "that in secondary school I quite deliberately refrained from working to my highest potential because I knew that I would be accepted to even top universities without doing so. Almost any black child knows from an early age that there is something called affirmative action which means that black students are admitted to schools under lower standards than white; I was aware of this [from] at least the age of ten. And so I was quite satisfied to make B+'s and A-'s rather than the A's and A+'s I could have made with a little extra time and effort. Granted, having the knack for school that I did, I was lucky that my less-than-optimum efforts still put me within reach of fine schools. However, there is no reason that the same sentiment would not operate even in black students who happen to be less nerdy than I was. ... In general, one could think of few better ways to depress a race's propensity for pushing itself to do its best in school than a policy ensuring that less-than-best efforts will have disproportionately high yield."

McWhorter sees a similar dynamic at work with black college students as with black high school students since they know they will be given a huge boost over their white and Asian classmates when the time comes to apply for jobs and places in graduate and professional schools. When the huge boosts that blacks receive under the reigning affirmative action system are combined with what McWhorter calls the "cultural disconnect" that he believes leads many black youth to hold schoolwork "at half an arm's length," the inevitable result is a huge racial gap in achievement.

Shelby Steele, another prominent black critic of affirmative action, makes the same claim even more pithily. Racial preferences, says Steele, send out the message to black students that "mediocrity will win for them what only excellence wins for others." In view of this message, for many black students the most attractive strategy becomes one of sitting back in class, taking it easy in school in terms of studying and hard work, and letting the whites and Asians toil away to get their good grades. The black students, says Steele, know that they don't have to perform at nearly the same level as their white and Asian schoolmates to advance along a similar career path. They can make up through race what they lack in performance.

"The top quartile of black American students," Steele writes, who often come "from two-parent families with six-figure incomes and private school educations," is precisely the group "that has been most aggressively pursued for the last thirty years with affirmative action preferences." "Infusing the atmosphere of their education from early childhood," he continues, "is not the idea that they will have to steel themselves to face stiff competition but that they will receive a racial preference. ... Out of deference, elite universities have offered the license not to compete to the most privileged segment of black youth."

Other prominent black critics of affirmative action, including economist Walter Williams, Vanderbilt law professor Carol Swain, and Supreme Court Justice Clarence Thomas have offered similar observations. Like McWhorter and Steele they all present what might be called a "preference disincentive theory" to explain the seemingly inexplicable phenomenon of black underperformance. And this is the only theory out there that can explain why underperformance is observed across the entire spectrum of black talent including among those least likely to experience "stereotype threat." It is also the only theory that can explain the fact that it is those blacks with the highest SAT scores who underperform the most.

The high-SAT blacks, who are often the most intellectually gifted and have the highest IQs, know that professional schools and graduate schools look high and low for black students of their intellectual caliber and will admit them with much lower grades than their white and Asian counterparts. Since it is the most competitive graduate and professional schools that extend the greatest affirmative action boost to black applicants, it is not surprising that it is the best and brightest blacks who underperform the most. They know they need not get grades in college that come anywhere close to those of their white and Asian classmates to gain places in the most prestigious graduate institutions and in the most sought after corporations and law firms. While some students are internally driven to achieve at their maximum regardless of external rewards, and others receive relentless grade pressure from home, this is not how typical American students behave, and certainly not typical black students. (Participant observer studies have shown just how half-heartedly many black teen subcultures support academic striving even when there is no overt subcultural hostility of the kind that would stigmatize getting good grades as "acting white.").

A Threatening Truth

Despite its common sense appeal and the testimony of many prominent black commentators, the "preference disincentive theory" is rarely taken up by proponents of affirmative action. It appears to be too threatening to their policy preferences even to be considered very seriously lest it turn out to be true. In The Shape of the River the theory is discussed on just one page (of a 400 page book), though the discussion is superficial and concludes with Bowen and Bok saying "we know of no way to test this hypothesis." In the Bowen and Vars study the theory is taken up -- and dismissed -- in a single, two-sentence footnote. "While some commentators have argued that affirmative action in the workplace weakens incentives for black students to perform academically," they write, "even if affirmative action were to shift upward career prospects for black [college] graduates, the marginal payoffs to academic achievement should remain constant."

This last statement is true, of course, in the sense that it is almost always to the advantage of black college students in terms of their careers to get better grades even if they know they will be given various post-college racial boosts. Marginal grade improvements will almost always pay off careerwise. And if black college students were always single-mindedly focused on getting into the best professional schools, or landing the most prestigious positions in the corporate world, affirmative action would have no effect on how hard they work in college. But black students must not to be confused with profit-maximizing business firms, nor are they analogous to Max Weber's inner-worldly Puritan ascetics restlessly striving to prove themselves before God through their abstemiousness and tireless labor.

Academic work in college is often arduous and demanding and surely less attractive for most young people than the alternatives of socializing with friends, listening to music, playing sports, and the like. Trade-offs will be made, and if black students are told that they can advance just as fast and go just as far in their career trajectory as similarly talented white or Asian students while doing much less demanding academic work than these other students, it is a deal many black students will readily accept. In social science terminology, they will "satisfice" rather than "maximize" their academic and career potential. Like John McWhorter in his Philadelphia high school, they will accept decent or mediocre grades even though by means of more diligent work they could have achieved -- and been additionally rewarded for achieving -- substantially better grades. This is what causes black underperformance.

None of what I have said here is rocket science or string theory. The failure of the academic establishment to acknowledge the obvious here in understanding the underperformance phenomenon can only be chalked up to a fear of discrediting the affirmative action policies they have long championed.

-----------------------------------------------

Russell K. Nieli is a Senior Preceptor in the Executive Precept Program of the James Madison Program in American Ideals and Institutions at Princeton University, and a Lecturer in Princeton's Politics Department. His new book, Wounds that Will Not Heal: Affirmative Action and Our Continuing Racial Divide, will be published next year by Encounter Books.



Comments (3)

Deckin:

Simply put the possibility of the statistical fallacy known as absent data does not appear to have been taken into account. When Nieli note that blacks underperform their SAT scores, he notes that these results come even after controlling for the selectivity of the institutions. I daresay that kind of control is not good enough. One effect of affirmative action is that blacks with SAT scores of X will get into an institution much more selective and much more rigorous than an asian or a white or student student with the same SAT score. Grouping institutions together as 'selective' is not good enough. I would want information for the same institution and ideally, the same major at that institution. What we're likely seeing here is nothing more than an artifact of the fact these students are getting into places and majors they really aren't qualified for and their grades reflect that. Even institutions grouped together as 'selective' can vary widely as to their grading policies and course rigor. Until we see more and better data, the rational person ought to remain unconvinced.

Another excellent essay by Professor Nieli -- many thanks to him for writing it and to Minding the Campus for posting it!

bob sykes:

The average IQ of American blacks is 85, one full standard deviation below the average white IQ of 100. Asians and Jews have yet higher IQ.

This means that only 1/6 of blacks will score as well or better than the white mean. That's all there is too it. IQs are largely hereditary (50 to 80%), so these ratios will not change. The so-called gap closure in the NYC schools has been shown to be due to fraud.

The rest of the discussion is utter nonsense.

Post a comment

Leave comments here. Unless they are vicious or obscene, they will be printed.


MONTHLY ARCHIVES:

 

RECENT ESSAYS

Texas Leads the Way on Higher-Ed Accountability
Thomas K. Lindsay

More Decline in the U. of Chicago Core
Adam Kissel

Do We Over-Invest in Non-Traditional Students?
Richard Vedder

The Liberal Arts Are in Trouble--Should We Celebrate?: Part 2
Symposium

The Liberal Arts Are in Trouble--Should We Celebrate?
Symposium

All Essays >>>

Diversity

Published by the Manhattan Institute
The Manhattan Insitute's Center for the American University.