In an article in the Summer 1969 Journal of Negro Education, Harvard sociologist Nathan Glazer raised a cautionary flag about the prospects of attaining ethnic group parity in the area of educational achievement. “History and social research convince me,” he wrote, “that there are deep and enduring differences between various ethnic groups in their educational achievement and in the broader cultural characteristics in which these differences are, I believe, rooted.” Such differences, Glazer continued, “cannot be simply associated with the immediate conditions under which these groups live, whether we define these conditions as being those of levels of poverty and exploitation, or prejudice and discrimination.” “If we are to have a decent society,” he concluded,
men must learn to live with some measure of group differences in educational achievement, to tolerate them, and to accept in some degree the disproportion in the distribution of rewards that may flow from differences in educational achievement….We need to press not only our research on these differences, their origins, their extent, their causes, the measures that reduce them, but also develop and strengthen a political and social philosophy that permits a society to accept them, to live with them, and be stronger because of them.1
In the first half of the twentieth century, Glazer pointed out in his article, groups like the Japanese, Chinese, and Jews did outstandingly well in the public educational system in America and reaped the commensurate rewards in terms of financial and occupational advancement. Other groups, including the very low-achieving Italians (my own ethnic group), did very poorly and advanced much more slowly, though the more successful groups surely experienced no less discrimination or socio-economic disadvantage than the less successful ones (they often experienced more). Such performance differences, Glazer explained, are found not only in America, but around the world, and it is neither accurate nor wise, he implied, to attribute such disparities either to the malevolence of oppressors or the lasting effects of such oppression; nor is it wise to expect that they can be overcome entirely by ameliorative public policies, however desirable such policies may be.
I was remembering these thoughts of Glazer (from whom I have learned much over the years about ethnic groups in America), when reading Thomas Espenshade and Alexandria Walston Radford’s new study, No Longer Separate, Not Yet Equal: Race and Class in Elite College Admission and Campus Life, which deals with student characteristics and performance at eight highly selective American colleges and universities.2 Espenshade and Radford write from what might be loosely described as a “social democratic” perspective that equates economic and educational inequality with social injustice, and their major concern seems to be with eliminating differential outcomes and disproportionate group success rates among those seeking admission to elite colleges across socio-economic and ethno-racial categories. But as they indicate in the final chapter, they are also concerned with maintaining at least some kind of merit-based selection system, where high school grades and standardized test scores count for the bulk of a college applicant’s entrance requirements—something they recognize has deep resonance with the American public. It is the academic performance gaps, they realize, that upset most attempts to achieve socio-economic and ethno-racial group parity in elite college admissions, and these, they believe, should be what public policy tries to eliminate.
In my judgment, the authors could use a good dose of Glazer’s cautionary realism on the subject of differential performance by ethnic groups, just as they might think seriously about the cautionary realism of those social scientists who argue that there are difficult-to-change class-cultural barriers that negatively impact the academic performance of those from certain lower-class cultures,3 and genetic characteristics among many poor and working-class white groups that can have a similar education-retarding effect.4
The authors at times appear to believe that it is an unjust feature of American society that it permits the economically and educationally more successful (whether individuals or ethnic groups) to pass on some of the fruits of their success to their children in the form of improved schooling, enhanced educational opportunities, and superior cognitive development in the home. But most Americans do not think this way. Many years ago, the political scientist Jennifer Hochschild, a left-leaning social democrat, conducted extensive interviews with ordinary Americans on the question of economic fairness, and found—much to her dismay, it appears—that even many poor and working-class Americans think it perfectly fair and reasonable for the economically and educationally more successful to improve the educational and other life opportunities of their offspring. They would do the same, they told Hochschild, if given the opportunity.5 This should not have been too surprising, since one reason so many people from so many diverse places around the world have come to America is precisely to get ahead and pass on the advantages they have acquired to their children and grandchildren.
The American ideal has been that each new generation is permitted —and indeed encouraged—to build on the success of those who have come before. Many call it “the American Dream.” Hochschild’s working-class respondents seemed generally unconcerned with “distributive justice,” “economic equality,” “economic democracy,” “Chancengleichheit,” and similar ideals that preoccupy European-style socialists and social democrats and their many counterparts in America. Most Americans, in short, are not social democrats and do not look to redistributive welfare states such as Denmark and Sweden as their national role models. For politicians in most areas of the country to embrace socialism or social-democratic ideals is politically the kiss of death.
Many academic intellectuals, of course, would like to change this situation, but for the most part they have not succeeded. So when Espenshade and Radford remark that “high SES [socioeconomic status] parents are more likely than low SES parents to send their children to private schools, move to neighborhoods where the public schools are purportedly better, secure a private tutor to help with homework and basic organizational skills, and encourage their children to take a large number of AP exams and SAT II subject tests” (19), this is not something that most ordinary Americans consider unjust, even though such behavior is likely to lead to substantial class-based outcome inequalities of the kind that social democrats and those on the left find so objectionable. Most Americans want to be able to pass on any socio-economic and educational gains they have made in their lives to their children, the possibility of which provides a very considerable part of many ordinary people’s incentive to work hard and develop a more focused orientation toward the future. Were this possibility eliminated, one wonders whether America would remain the magnet it has been for the upwardly mobile, energetic, and creative people of the world, and how economically and technologically progressive it would continue to be.6
Some Intriguing Findings
Despite its overarching social-democratic perspective—which clashes sharply with my own more classical liberal values—I found much in No Longer Separate, Not Yet Equal to admire. The book is based on a detailed analysis of students who enrolled as freshmen in the fall of 1983, 1993, and 1997 at eight highly selective public and private four-year colleges and universities (average SAT score 1360),7 and contains over 140 charts and tables analyzing a huge array of factors regarding student performance and student life. There is so much interesting information here that at times it overwhelms. Much of the material analyzed comes from an extensive sixteen-page questionnaire filled out and returned by almost 9,100 students. The study was sponsored by the Andrew W. Mellon Foundation—the same organization that sponsored the three pro-affirmative action “River Books” previously reviewed in these pages.8 Here are some of Espenshade and Radford’s findings that AQ readers might find intriguing:
- Black students enjoy an enormous advantage in admission to competitive institutions in comparison with whites from similar socio-economic backgrounds, Hispanics experience a lesser but still very substantial advantage, while Asians suffer a substantial admissions penalty. Holding background characteristics constant, being black confers a 310-point SAT advantage (on a 1600-point scale) over whites, and being Hispanic a 130-point advantage. Being Asian, however, is equivalent to a 140-point disadvantage compared to equally qualified whites. With socio-economic and other background factors held constant, an Asian with a 1450 SAT score has about the same chances of admission as a white student with a 1310, a Hispanic student with an 1180, and a black student with a 1000. The black-over-Asian advantage is equivalent to 450 SAT points.
- Because of the enormous achievement gap nationwide in terms of SAT scores and high school grades among the four ethno-racial groupings, replacing current preference policies with policies that give greater weight to socio-economic deprivation (“class-based affirmative action”) would do little to achieve the kind of minority group representation that academic administrators desire. Statistical simulations indicate that such enhanced class-based policies at the competitive private institutions surveyed, where substantially more bonus-point credits are given than are given now to students from lower-SES backgrounds, would result in the reduction of the current representation of blacks from 8.3 to 4.0 percent, and of Hispanics from 7.9 to 6.2 percent. Asian representation, however, would rise, as would that of whites. If racial preferences at these schools were eliminated and no additional weight given than what is presently accorded to students from lower socio-economic backgrounds, black representation would decline to 3.6 percent and Hispanic representation to 5.8 percent. If admission to these schools were based just on SAT scores with no special consideration for race or class, current black representation would dwindle to 0.9 percent and Hispanic representation to 2.5 percent, while Asian representation would jump to 38.3 percent (more than ten times the black and Hispanic populations combined).
- >Across the eight institutions, public and private, more than half the black students graduate in the lowest fifth of their class, while only one in twenty graduate in the top fifth. The average white student graduates at the 57th percentile, the average black at the 20th percentile—a gap of 37 percentile points. Statistical models indicate that only 20 points of this 37-point percentile gap can be attributed to the lower high school grades, SAT scores, and class rank of the entering black students. The remaining 17 percentile points is due to blacks simply doing less well in terms of college grades than whites with similar entering scholastic credentials. Numerous other studies have documented this “underperformance.” Although Espenshade and Radford say that this underperformance has not been adequately explained, at one point they offer a suggestion: “In his dissent in the Grutter case, Justice Clarence Thomas argued that affirmative action stigmatizes blacks and other beneficiaries of racial preferences and lowers a necessary incentive for hard work and academic preparation because underperforming students believe that affirmative action will give them the necessary edge” (302n).
- Black, Hispanic, and Asian applicants from lower socio-economic backgrounds enjoy a considerable admissions advantage over applicants from their own ethnic group coming from higher SES backgrounds when comparisons are made of those who have similar high school scholastic records and other relevant background characteristics. On an other-things-equal basis, lower-class blacks applying to private colleges have five times the odds of getting admitted as blacks from the upper class (.87 odds ratio versus .17), lower-class Hispanics three times the odds of upper-class Hispanics (.65 versus .22), and lower-class Asians more than four times the odds of upper-class Asians (.58 versus .13).
- >While applicants from non-white ethno-racial groups enjoy a huge admissions advantage if they are from poor families, whites from lower-class backgrounds suffer huge admissions disadvantages not only in comparison to their counterparts among non-whites, but among whites from wealthier backgrounds. At the private institutions surveyed, otherwise equally qualified whites are only one-third as likely to get accepted if they come from lower-class families as those from upper-class backgrounds (predicted probabilities .08 odds ratio versus .23). Lower-class whites clearly experience a large class penalty. When the comparison is across ethno-racial groups, the lower-class white penalty becomes enormous: lower-class Asian applicants are seven times as likely to be accepted to competitive private institutions as similarly qualified lower-class whites (.58 versus .08), lower-class Hispanic applicants eight times as likely (.65 versus .08), and lower-class blacks ten times as likely (.87 versus .08).
- The enormous disadvantages for lower-class whites in comparison to non-whites and wealthier whites is partially explained by the fact that, except for the very best endowed, private institutions are reluctant to admit students who cannot afford their high tuitions, and since they have a limited amount of money to give out for scholarship aid, they reserve this money for those who can be counted in their enrollment statistics as diversity-enhancing racial minorities. Poor whites are given little weight as enhancers of campus diversity, whereas blacks and Hispanics of any socio-economic category, even the wealthiest, as well as Asians of the lower socio-economic category, are. Private institutions, say Espenshade and Radford, “intentionally save their scarce financial aid dollars for students who will help them look good on their numbers of minority students” (99n). Quoting a study by New York University researcher Mitchell Stevens, they write: “ultimate evaluative preference for members of disadvantaged groups was reserved for applicants who could be counted in the college’s multicultural statistics. This caused some admissions officers no small amount of ethical dismay” (99n).
- While participation in high school extracurricular activities usually offered some advantage in the college admissions process, participation in high school ROTC, 4-H clubs, and Future Farmers of America actually somewhat diminished a student’s admissions chances on an other-things-equal basis. Although Espenshade and Radford do not say so, there seems to be at least a modest admissions bias against “red-state culture” and “red-state occupations,” and surely no efforts are being made to get more students with farming or military backgrounds or aspirations admitted to prestige colleges under their stated diversity-enhancement programs.
- Immigrants and their offspring are doing extraordinarily well in getting into the more competitive institutions. Of the eight institutions surveyed, almost four in ten students are foreign-born immigrants or the children or grandchildren of immigrants, a much higher proportion than such people in the American population at large. At the private colleges surveyed, six in ten of the black students in the 1997 applicant pool were either immigrants or their children or those of mixed racial backgrounds. Estimates at Ivy League institutions indicate similar immigrant black overrepresentation, with one study suggesting that four in ten blacks at Ivy League institutions have foreign-born parents. Blacks whose families have been in America for many generations and are likely to be able to trace their American ancestry to slavery times are doing very poorly in comparison with the offspring of more recent black arrivals.
A Call to Refocus on the K–12 Performance Gap
Many AQ readers will find the major conclusion of No Longer Separate, Not Yet Equal pleasantly surprising given its Mellon Foundation sponsorship. While the authors do not call for eliminating racial preference policies, they are not their die-hard defenders, and they recognize not only how unpopular racial preference policies are with the general public, but how precarious their future existence may be. The combination of state anti-affirmative action referenda and a Supreme Court increasingly hostile to race-based decision-making, they suggest, may render impossible the continuation of current policies. Their final recommendation is one that many of affirmative action’s severest critics could readily endorse:
Improving the academic performance of underrepresented minority students constitutes the only viable, long-run strategy for preserving meaningful minority representation on elite college campuses if race-based affirmative action is eliminated. No amount of tinkering with admission practices can match the power or attractiveness of this solution….We need to make closing the racial achievement gap a high social priority and to move aggressively and with the greatest determination to make it happen….The racial gap in grades, test scores, and other measures of skills, abilities, and knowledge that children acquire is arguably the most pressing domestic issue facing the United States at the beginning of the twenty-first century. (377, 398)
For Espenshade and Radford the day of reckoning has arrived and their main policy proposal is a massive government-financed study, conducted on an unprecedented scale (“a Manhattan Project for the behavioral and social sciences” ), that would monitor the progress of perhaps tens of thousands of school children in a single birth cohort as they progress through the K–12 school system. The purpose of such a study would be to get a better grasp of why black and Hispanic students do so poorly in school and what can be done to remedy the situation.
Espenshade and Radford, in my view, are clearly on the right track here when they refocus attention away from the acrimonious and divisive affirmative action debate to the racial achievement gap in learning—what some call the “pipeline problem.” I have three reservations, however, about their Manhattan Project idea. First, while better social science research might be in order, it is simply untrue, as Espenshade and Radford seem to believe, that we currently lack knowledge of the kinds of programs and schools that work well with black and Hispanic students. The pioneering work on Catholic schools by James Coleman and Thomas Hoffer, and the subsequent work on the same topic by Anthony Bryk and his colleagues, as well as decades of “effective school” research beginning in the 1970s, have given us a good indication of what kinds of schooling works well with black and Hispanic youngsters, especially those from troubled neighborhoods and family circumstances.9 The small, usually church-related private schools have a much better track record than the neighborhood public schools, and if it weren’t for the opposition to vouchers by the teachers’ unions, many more such institutions would flourish.
My second objection is with the grandiosity of Espenshade and Radford’s proposal, which almost certainly would raise expectations beyond the capacity of any government action or set of public policies to satisfy. If we have learned anything from the failure of Lyndon Johnson’s Great Society extravaganza, it is that we should be wary of raising hopes and expectations beyond our capacity to fulfill them, since the inevitable failures that follow can lead to a generalized despair and reluctance to consider more modest proposals that might actually achieve their goals.
My final objection is a recapitulation of Glazer’s warning about the inevitable persistence of ethno-racial group differences and the need to adopt a system of expectations and values that accepts unequal group outcomes among a free people, even as we try to improve the performance of those lower-achieving individuals and groups who may have special educational needs. A grandiose social science research project of the kind that Espenshade and Radford propose might well discover that an important source of group differences are persistent cultural impediments such as those that Glazer stresses, or the presumably even more persistent genetic differences that Arthur Jensen and his followers claim. If this were the case and we were to adopt the implied Espenshade/Radford value system that holds such group differences to be forms of social injustice, we would find ourselves in a difficult situation indeed.
To save us from this fate, I propose a much more modest goal of opportunity enhancement for the least advantaged individuals of whatever race or ethnicity, combined with a value system closer to the classical liberal ideals of such thinkers as Locke, Smith, Madison, Hayek, and Kristol. These are ideals with strong resonance in America and do justice to human decency and pragmatic common sense. Socialism and social democracy have never been very popular in America—and we remain a much stronger nation and a better people because of it.
1Nathan Glazer, “Ethnic Groups and Education: Towards the Tolerance of Difference,” Journal of Negro Education 38, no. 3 (Summer 1969): 187, 195. In the same year this piece appeared, Berkeley psychologist Arthur Jensen published his famous—and for many incendiary—“How Much Can We Boost IQ and Scholastic Achievement” in the Harvard Education Review, which argued that much of the black/white achievement gap in the classroom was genetic in origin. Glazer has always adhered to a cultural interpretation of differential performance among ethnic groups, and does not speculate about genes, though if there is any truth to the claim of Jensen and his followers about genetic differences, Glazer’s cautionary tale about the prospects of achieving group parity in academic attainment would obviously gain added strength.
2These eight schools remained anonymous in the study as a condition of giving Espenshade and Radford data on their students.
3The Unheavenly City (Little Brown and Company, 1968), Harvard urban specialist Edward Banfield showed that those from lower-class cultures are much more present-time oriented and less concerned with improving the life chances of their offspring than those from working-class or middle-class cultures. As one ascends the class hierarchy, Banfield claimed, people tend to focus more on long-range plans for themselves and their progeny.
4The Bell Curve (The Free Press, 1994), Charles Murray and Richard Herrnstein make the argument (first articulated by Herrnstein in a 1971 Atlantic Monthly article), that (a) academic talent, which is correlated closely with things like IQ test scores, is partially heritable, (b) American society rewards handsomely those with superior academic talent, and (c) as a result of this the offspring of the academically and economically most successful often have a genetic advantage in school. Behavioral geneticists have for years shown the huge component of heredity in IQ differences (adopted children, for instance, have IQs much closer to those of their biological parents than to those of their adoptive parents), though Espenshade and Radford seem unacquainted with the extensive behavioral genetics literature on this topic.
5Jennifer Hochschild, What’s Fair? American Beliefs about Distributive Justice (Cambridge, MA: Harvard University Press, 1981).
6Although socialists and social democrats rarely discuss the matter, the “equal life chances” ethic they champion would presumably envision an ideal world as one in which the government saw to it that children of parents who had achieved the highest SES status have the same chance of winding up in the lowest SES bracket as the children of parents who are in this bracket—so that a doctor’s son or daughter would have the same chance of becoming an unskilled laborer as the unskilled laborer’s offspring. This would eliminate “the social reproduction of inequality across generations” (380) that Espenshade and Radford apparently deplore. It would be wrong according to the social democratic ideal for parents who have attained above-average socio-economic standing to try to maintain (or improve) that above-average standing in the lives of their offspring.
7All pre-1995 exam scores were converted to the upwardly re-centered, post-1995 scoring system.
8Those interested can see my “Selling Merit Down the River,” which appeared in the Fall 2009 Academic Questions (vol. 22, no. 4), a review essay of William Bowen and Derek Bok, The Shape of the River: Long-Term Consequences of Considering Race in College and University Admissions (Princeton University Press, 1998); Douglas Massey, Camille Charles, Garvey Lundy, and Mary Fischer, The Source of the River: The Social Origins of Freshmen at America’s Selective Colleges and Universities (Princeton University Press, 2003); and Camille Charles, Mary Fischer, Margarita Mooney, and Douglas Massey, Taming the River: Negotiating the Academic, Financial, and Social Currents in Selective Colleges and Universities (Princeton University Press, 2009).
9See James Coleman and Thomas Hoffer, Public, Catholic, and Private Schools: The Importance of Communities (New York: Basic Books, 1987); Anthony Bryk, Valerie Lee, and Peter Holland, Catholic Schools and the Common Good (Cambridge, MA: Harvard University Press, 1993); Sara Lightfoot, The Good High School: Portraits of Character and Culture (New York: Basic Books,1983); Andrew M. Greeley, Catholic High Schools and Minority Students (New Brunswick, NJ: Transaction Publishers, 1982); Abigail and Stephan Thernstrom, No Excuses: Closing the Racial Gap in Learning (New York: Simon and Schuster, 2003); and Terry Moe, Schools, Vouchers, and the American Public (Washington, DC: Brookings Institution Press, 2001)