New York City’s vast public school system enrolls 1.1 million students, some 18,000 of whom attend nine “specialized” high schools, where the curriculum is particularly rigorous and admission is both widely sought and highly competitive. Stuyvesant, Bronx Science, and Brooklyn Tech are the oldest, largest, and most famous such institutions. Eight of these schools base admission decisions solely on applicants’ scores on the Specialized High School Admissions Test (SHSAT), developed by an education assessment company under contract to the New York school system, which began using it in 1971. (The ninth concentrates on art, music, and the performing arts. It admits students on the basis of portfolios or auditions, since no standardized test can reliably identify those 13-year-olds who will, over the ensuing four years, turn out to be the most annoying.)

In June, New York Mayor Bill de Blasio called for his city’s schools to replace SHSAT with an admissions process relying on two measures: middle-school class rank, and scores on a test taken by every student in New York state. Using SHSAT is a “monumental injustice,” he contended, because blacks and Hispanics account for two thirds of all New York City public school students but only one tenth of those enrolled in the specialized high schools. For de Blasio, this gap shows that using SHSAT denies students “an equal chance to get into one of their city’s best high schools.” Under the admissions procedure the mayor has proposed, which cannot be implemented without the New York state legislature’s approval, the specialized high schools will “start looking like New York City.” Black and Hispanic students, that is, would account for about 45% of enrollment, much higher than the current figure, though still only two thirds of their numbers in New York’s entire school system.

Tellingly, de Blasio treats SHSAT’s unrepresentative outcome as proof that its use constitutes an unfair process. His reasoning applies to a specific situation the general principle recently expressed by Ibram X. Kendi: “As an anti-racist, when I see racial disparities, I see racism.” Kendi, a historian who directs the Anti-Racist Research and Policy Center at American University, is the author of Stamped from the Beginning: The Definitive History of Racist Ideas in America, which won a National Book Award in 2016. He recently wrote in the New York Times that “many racist Americans” resort to what he considers the only alternative explanation for racial disparities: “black inferiority.” In the same spirit, de Blasio writes that objections that his admissions proposals will “lower the standard” at the specialized schools are based on a “narrative” that not only “traps students in a grossly unfair environment,” but “actually blames them for it.”

Disparate Impact

The idea that disparities are guilty until proven innocent is central to anti-discrimination law and politics. The Supreme Court’s 1971 decision in Griggs v. Duke Power Company held that employment criteria, such as completing high school or not having a criminal record, could be “neutral on their face, and even neutral in terms of intent” but still constitute a discriminatory business practice. If a criterion has a “disparate impact” that causes, say, more blacks than whites to be removed from the pool of job applicants, an employer can avoid legal jeopardy only by demonstrating that it has “a manifest relationship to the employment in question” or derives from a “business necessity.” Failure to meet this burden of proof exposes an employer to charges of illegal discrimination on the basis of race, sex, national origin, etc.

The standard developed by court rulings and regulatory policies since Griggs is the “four-fifths rule.” If a minority group’s representation at a workplace is less than 80% of its proportion in the local population, then the employer can be charged with illegal discrimination. Exoneration in those circumstances requires demonstrating that the personnel practices that yielded the unrepresentative workforce meet the “manifest relationship” or “business necessity” test. Many employers choose, instead, to avoid the costs, risks, and bad publicity of litigation by settling out of court, or preemptively altering their business practices to keep from triggering the four-fifths rule.

Clear thinking about demographic disparities begins with the recognition that when allocating a finite number of benefits, such as admission to a prestigious educational institution, the under-representation of some groups necessarily means the over-representation of others. And white students do indeed account for 24% of those enrolled in the schools employing SHSAT, despite being only 15% of New York City public school students. (The least youthful of New York’s major demographic groups, non-Hispanic whites accounted for 33.3% of the city’s population in the 2010 census.)

The real over-representation, however, is of Asians, 16% of all students citywide (and 12.7% of the five boroughs’ total population) but 62% of those attending the specialized high schools. There is no evident way to account for this fact within de Blasio and Kendi’s framework. One logical possibility—that New York City is run by its Asian minority to the detriment of all other groups, the way South Africa was run by its white minority—can be dismissed for lack of evidence. The alternative, however, is equally implausible: white New Yorkers are in charge of the city, where they implement policies that are injurious to blacks and Hispanics, modestly beneficial to themselves, but hugely advantageous to Asians. If whites did possess such power, surely they would be the main, or even sole, beneficiaries of the opportunities denied to other groups.

Asians’ over-representation in New York’s specialized high schools doesn’t prove that the current admissions system is the best one possible. But when a particular minority group, the object of animus and discrimination throughout much of American history, flourishes in an admissions process relying entirely on SHSAT, we must acknowledge the strong possibility that this test is given for meritocratic rather than racist purposes. If the over-representation of Asians in the specialized schools cannot be ascribed to favoritism, either by them or by whites toward them, then to explain blacks and Hispanics’ under-representation we must consider factors other than pervasive bias against those groups.

Reality Check

When economist Thomas Sowell sees disparities, he, like Ibram Kendi, sees bias driven by hatred or ignorance. Unlike Kendi, however, Sowell sees other causes also at work. Sowell is the Rose and Milton Friedman Senior Fellow on Public Policy at Stanford University’s Hoover Institution, which is fitting, since he received his doctorate from the University of Chicago after studying with Milton Friedman. His 45 books cover many topics, but a recurring subject is the nature and social effects of race, culture, and discrimination. Some of his most original and courageous work has examined this issue.

If every disparity is explained completely by prejudice, then a world without prejudice would be one without disparities, where each identifiable group is proportionally represented in every socioeconomic category. Stuyvesant High School’s student body would be a demographic miniature of New York’s adolescent population—or, at least, every group’s representation at Stuyvesant would surpass 80% of its total in the city—just as the inmates detained at Rikers Island or swells listed in the Social Register would be demographic miniatures of the city’s adult population. This proportionality isn’t the rule, Sowell argues, and it’s not even an exception: there’s no clear evidence that the standard we are supposed to judge by, and strive for, has ever existed anywhere. In Intellectuals and Race (2013) Sowell cites political scientist Myron Weiner’s assessment of “the universality of ethnic inequality,” which holds that, “All multi-ethnic societies exhibit a tendency for ethnic groups to engage in different occupations, have different levels (and, often, types) of education, receive different incomes, and occupy a different place in the social hierarchy.”

Kendi and de Blasio might reply that these differences simply demonstrate the ubiquity and power of discrimination, but Sowell believes the world offers too many counterexamples disputing this conclusion. He notes that Malaysia’s ethnic Chinese minority, presently about a quarter of the country’s population, is politically weak but economically powerful. In the 1960s, for example, students from this minority group earned more than 400 degrees in engineering, while students from the Malay majority earned just four. Sowell’s new book, Discrimination and Disparities, points out that from 1950 to 2000, Jews—less than 1% of global population, and the objects of hatred and abuse in many parts of the world for centuries—received 22% of the Nobel Prizes in chemistry and 32% in both medicine and physics.

Other than discrimination, what could account for such disparities among groups? A fair summary of Sowell’s work since his third book, Race and Economics, was published 43 years ago, would be reality—economic, historical, and anthropological. Sowell is profoundly skeptical of the pat idea that every disparity constitutes prima facie evidence of injustice and oppression. Any such assumption, he argues, disregards the myriad ways that human variety and freedom generate widely divergent outcomes.

Cosmic Injustice

Many differences begin, literally, at home. Discrimination and Disparities cites one study showing that those born first in their families were more likely to be National Merit Scholarship finalists than all their younger siblings combined, even in families with as many as five children. Similarly, 22 of the 29 original astronauts in the Apollo program were their family’s oldest (or only) child. “If there is not equality of outcomes among people born to the same parents and raised under the same roof,” Sowell asks, “why should equality of outcomes be expected—or assumed—when conditions are not nearly so comparable?”

At the other end of the scale from intimate family relationships, tectonic geopolitical forces acting over centuries also generate relentless pressures that increase inequality, often in ways so elemental that it is hard to step back far enough to see them. Two millennia ago, according to Intellectuals and Race, the people of western Europe suffered “ruthless slaughters” and “brutal oppressions” not visited upon those in eastern Europe, who were never forcibly incorporated into the Roman Empire. One result, however, was that the western Europeans’ descendants acquired an enormous advantage vis-à-vis eastern Europeans: written Latin provided an alphabet the peoples ruled by Rome adapted, which brought literacy to western Europe long before its advent farther east. This head start in using written languages conferred a massive advantage in the retention and accumulation of knowledge, which helped the West develop such civilizational multipliers as coins, paved streets, castles, and printing presses centuries before they appeared in eastern Europe.

By the same token, Sowell argues, the poverty and strife endured by the people of sub-Saharan Africa, and their descendants in the New World, have causes beyond slavery and colonialism. (And, he has pointed out in many writings, slavery and subjugation were ubiquitous around the world and throughout recorded history, rather than a unique depredation visited by Europeans on the indigenous people of what is now called the Third World.) Africa, separated from the rest of the world by enormous oceans and a desert larger than the continental United States, suffered acutely from geographical isolation, thereby limiting its peoples’ exposure to, and adaptations of, useful innovations. Other isolating factors included “a dearth of navigable waterways within sub-Saharan Africa, as well as a dearth of natural harbors, [and] the difficulties of maintaining draft animals because of the disease-carrying tsetse fly.” Geography isolated Africans not only from the outside world but from one another. One result, Sowell notes, is that Africa accounts for one tenth of the world’s people but one third of its languages.

One of Sowell’s most clarifying distinctions is between social justice and cosmic justice. Slavery and Jim Crow were, clearly, massive social injustices. Some people subjected others to injurious, humiliating treatment, practices that those who inflicted the wounds would never voluntarily submit to themselves. The least censorious but also least contestable thing to be said about such cruel depredations is that they were unfair.

It makes no sense, however, to say that it is either fair or unfair to be a sibling other than the oldest, deprived of undivided parental attention in one’s earliest years. Or to come from an economic and cultural background shaped by the reality of acquiring literacy centuries after other peoples had done so. Or to be descended from people who were geographically isolated and suffered the resulting consequences. These are cosmic injustices, not social ones. To call such burdens unfair requires a principle that all people, in all times and places, should start life with a full and equal allotment of every kind of advantage that families, communities, and civilizations can bestow. We can hardly hope to comprehend this standard of fairness, much less devise any practical plan for realizing it.

It’s hard to apportion causality between social and cosmic injustices for a particular disparity, such as that between the median household incomes of black and white Americans, or their respective high school completion rates. Sowell’s complaint is that those, like de Blasio and Kendi, who effectively insist that social injustice explains 100% of any disparity and cosmic injustice accounts for 0%, assume we could look around the world and across time to see people in general living the same way, wanting the same things, and pursuing them with identical degrees of skill and determination. The mountain of evidence to the contrary, however, argues for the profound, ongoing impact of causes other than social injustice.

The past, according to Intellectuals and Race, “is as much beyond our control as the geographic settings and historic happenstances that have left not only different individuals or races, but whole nations and civilizations, with very different heritages.” These heritages, mutable but also powerful, cannot simply be exchanged for other sets of habits and dispositions. Sowell quotes historian Ulrich Bonnell Phillips: “We do not live in the past, but the past in us.” There is, moreover, no plausible reason to believe the comforting axiom that all civilizations are created equal, in the sense that each one leaves the people it has shaped identically equipped to flourish in any and every historical setting.

Culture Club

In Civil Rights: Rhetoric or Reality? (1984), Sowell anticipated Kendi’s either/or proposition: “The civil rights vision tends to dichotomize the spectrum of possible reasons for group differences into (1) discrimination and (2) innate inferiority.” Sowell has argued in many writings for a third possibility: culture, derived from adaptations to the circumstances the world presented to a particular people in a particular time and place. Intellectuals and Race defines cultures as “whole ways of life,” which sounds too encompassing to serve as an analytical concept. In Race And Culture: A World View (1994), Sowell focuses on something more limited and useable: “those aspects of culture which provide the material requirements of life itself—the specific skills, general work habits, saving propensities, and attitudes towards education and entrepreneurship—in short, what economists call ‘human capital.’” Taking discrimination, innate differences, and culture as independent variables to explain the disparities in New York’s specialized high schools, we would seek and weigh evidence that: 1) the admissions process is biased against blacks and Hispanics but in favor of Asians and, to a lesser extent, whites; 2) Asian teenagers possess more innate scholastic ability than whites, who in turn possess more than blacks and Hispanics; and 3) Asians’ skills, habits, and attitudes toward education give them a significant cultural advantage vis-à-vis whites, and an even bigger one compared to blacks and Hispanics.

Discrimination and Disparities emphasizes one factor that can derive from each of these three causes: particular groups have different attainments because they often have different aspirations. There are, Sowell writes, “inescapable differences” in “what individuals and groups want to do, and are prepared to prioritize.” In 1950, 6% of America’s doctors were women. In 2018 the figure is 34%, and certain to continue increasing. According to the Association of American Medical Colleges, in 2016-17 women accounted for 45.8% of the students in “active residencies,” the final stage in medical training in which a doctor selects a specialty before beginning a career.

The advent of gender parity among doctors does not, however, translate into parity within medical subfields. Men accounted for 85.1% of the residents in orthopedic surgery, 73.9% in diagnostic radiology, 68.1% in nuclear medicine, and 64.7% in anesthesiology. Women were 82.7% of all residents specializing in obstetrics and gynecology, 73% of those in pediatrics, 64.4% in dermatology, and 55% in family medicine.

Advocates for each of Sowell’s explanatory categories can interpret these disparities as vindicating their own assessment. Perhaps, despite overall parity, the medical education process still discriminates on the basis of sex, inducing men to pursue some specialties and women others. Or, possibly, men and women have natural differences that shape the trajectories of their medical careers. In 2017, Google fired one of its engineers, James Damore, for writing an internal memo that some of his colleagues and superiors considered sexist. One of its assertions was, “Women on average show a higher interest in people and men in things.” This contention, however troglodytic one might find it, really is a useable description of the skills and interests that set obstetricians and pediatricians apart from surgeons, radiologists, and anesthesiologists.

Finally, it may be that men and women’s different preferences and aptitudes are culturally conditioned rather than innate. “A just future would be one without gender,” feminist philosopher Susan Moller Okin wrote in 1989. “In social structures and practices, one’s sex would have no more relevance than one’s eye color or the length of one’s toes.” She goes on to describe that future as one where men and women participate “in more or less equal numbers in every sphere of life, from infant care to different kinds of paid work to high-level politics.” We may safely conclude that to realize this vision will require changed laws and practices, but also new attitudes and habits. (Whether we want to realize it, and whether there is any anthropological basis for thinking we can, are separate questions.) If men care about things and women about people, for example, making gender obsolescent would require that we set ourselves against any assumption that they do so because men and women are that way, and proceed to identify and eliminate every socialization process that causes them to become that way.

Just Google It

One reason Sowell rejects the equation of disparities with discrimination is that such analyses ignore, as he says in his most recent book (published just before he turned 88), “the cost of discrimination to the discriminator.” Ignoring this cost means disregarding the ways discrimination is self-curtailing, a quality that increases the likelihood disparities result from other causes. It is true, for example, that blacks applying for home mortgage loans are rejected more frequently than white applicants. Ascribing this disparity to racism, however, ignores the inconvenient fact that black-owned banks reject black mortgage applicants at a higher rate than white-owned banks do. (Nor can this finding be ascribed to an apples-to-oranges comparison, skewed because riskier black borrowers apply to black-owned banks. A 1997 study published in the Journal of Financial Services Research, and cited in Discrimination and Disparities, found that “black-owned banks are more likely than white-owned banks to reject similarly situated black applicants.”) And, like New York’s specialized high schools, banks in general reject white mortgage applicants at a higher rate than Asian ones. These facts argue that objective realities rather than subjective prejudices explain the mortgage application rate disparity, at least in significant part.

In the same way, the de Blasio-Kendi framework disregards the benefit of non-discrimination to the non-discriminator. Google’s workforce, for example, is highly unrepresentative of the American population, especially in its technological departments, where 41.1% of the employees are Asian, 21.4% are women, 2.8% are Hispanic, and 1.5% are black. Judging by Google’s stated intentions, these disparities cannot be ascribed to prejudice or callous indifference. One corporate vice president is the company’s Chief Diversity and Inclusion Officer. Google boasts that it has made parental benefits gender-neutral and that 84% of its “people managers have taken Unconscious Bias training.” In firing James Damore after his memo about diversity at Google became public, CEO Sundar Pichai said, “To suggest a group of our colleagues have traits that make them less biologically suited to [their] work is offensive and not OK. It is contrary to our basic values and our Code of Conduct.”

And yet all this earnestness appears to be having results that are just north of negligible, consistent with Sowell’s deadpan rule that there is “no necessary correlation between what people say and what they do.” The company reports that between 2014 and 2017 the proportion of its tech hires who were women increased from 20.8% to 24.5%. At that rate, women won’t account for half of Google’s tech hires until after 2030, which means it will take many years beyond 2030 for women to constitute half of its tech staff. The proportion of tech hires who are black has soared from 1.9% to 2.0%.

Perhaps Google goes on and on about diversity and inclusion as part of a deliberate, cynical attempt to distract from, and perpetuate, its discriminatory practices. Or, perhaps, the company is sincere but feckless, genuinely baffled about how to pursue its diverse, inclusive aspirations.

The least plausible explanation is that Google can’t look after itself, and needs Unconscious Bias trainers to keep it from missing the big profits to be had by hiring and retaining a highly skilled, highly diverse workforce. In the decade after the firm went public in 2004, its adjusted share price increased by over 1,000%, compared to roughly 100% for the S&P 500, and its gross revenue grew 20-fold, from $800 million to $16 billion. A defining feature of the technology industry is that personnel practices are central to any competitive advantage. All a firm’s productive assets, that is, are its people. Advantages such as mineral deposits or access to shipping lanes don’t matter. Despite personnel practices that, by its own standards, have abjectly failed to achieve diversity and inclusion, Google has gone from being two Stanford doctoral students’ research project in 1996 to having its recently created parent company, Alphabet, ranked 22nd on the Fortune 500.

Of course, had Google’s personnel practices been more diverse and inclusive all along—giving it the benefit of larger numbers of superb but undervalued female, black, and Hispanic tech employees, while saddling it with fewer B-plus but overvalued male, white, and Asian ones—the company might be ranked even higher on the Fortune 500, and its initial investors would own more and bigger yachts and mansions. The best reason to doubt this counterfactual scenario is that if it were correct there would be tangible evidence for it, not just hypothetical speculation about it. There would be, that is, at least one other technology company, somewhere, that realized it could flourish by hiring a tech workforce that looks like America, not like Google. That company, perceiving the opportunity others missed or disdained, would have made a fortune by availing itself of the reserve army of able, ambitious, and underemployed female, black, and Hispanic software engineers.

The technology industry is not famously diffident about seizing opportunities to make fortunes, yet this diverse, inclusive, and obscenely profitable company has yet to be created. According to a recent diversity ranking of 100 tech companies, exactly one, Everlane, has gender parity in its tech workforce—seven men and seven women. (In 2017 total employment at Google’s parent company, Alphabet, was 88,110.) Only nine other firms had a tech workforce that was at least 30% female. Sixty-eight of the 100 companies, by contrast, had a smaller proportion of female tech employees than Google’s 21.4%. And in only 12 of the 100 companies did the proportion of tech employees who were black exceed 5%.

There’s a further reason to doubt that the tech industry looks so unlike America due to sexism and racism, rather than meritocracy. Technology is not one of those fields where the importance of interpersonal relationships gives more diverse organizations an advantage over less diverse ones. A police department or public school district in an ethnically diverse city might reasonably and justifiably seek a workforce demographically similar to the population it serves, hoping to strengthen relationships with all members of the community. Similarly, a realtor or car dealership with black or Hispanic customers might well conclude that it would be profitable to hire more black and Hispanic salespeople.

A computer, by contrast, knows nothing about the demographic characteristics of the person who wrote the software it runs. The program either works or it doesn’t. The fact that no technology firm has succeeded in hiring a workforce that looks like America strongly suggests that the large numbers of exceptionally talented black, Hispanic, and female programmers whose existence is posited by the industry’s fulsome rhetoric about diversity and inclusion…are just not out there. Until that changes, if it ever does, the campaign for a more diverse, inclusive tech industry necessarily entails a trade-off: accepting a less skilled workforce in order to employ a more demographically representative one. The notorious Damore memo may have been contrary to Google’s basic values, but it reflected very accurately the company’s actual practices in response to the basic talent-pool realities it confronts, ones that no amount of Unconscious Bias training will alter.

The Liberal Project

Sowell argues that one reason social injustice is over-diagnosed, and social justice remedies over-prescribed, is that the whole concept of discrimination harbors a basic ambiguity. The word “discriminate” derives from the Latin one “discriminare,” which the Oxford English Dictionary translates as “to divide, separate, to distinguish, differentiate.” In English, the term has come to encompass two opposite meanings: to make relevant, justifiable distinctions and, alternatively, to make irrelevant, even invidious, ones. We use the former sense when saying, with approval, that a wine expert has a discriminating palate, or an orchestra conductor a discriminating sense of tempo and dynamics. The second, disapproving sense of the term means, according to Sowell, “treating people negatively, based on arbitrary assumptions or aversions concerning individuals of a particular race or sex, for example.”

This central ambiguity about discrimination helps explain why those who oppose it routinely equate making distinctions with making unfair distinctions. As Sowell writes, the “implicit assumption that there would be no disparate outcomes unless there were disparate treatment” is “almost impervious to evidence.” To see disparities and immediately see racism (or sexism, ethnocentrism, homophobia, etc.) comes naturally to those who work from the assumption that no disparity ever has an innocent explanation.

This disposition, in turn, results from sincere zealotry rather than cynical or obtuse misinterpretations of the facts. In a debate with author Sam Harris about whether it is intellectually and morally licit to investigate genetic causes for racial differences, Ezra Klein of Vox.com said that to entertain this hypothesis would reassure those who have given up on furthering black progress through social policies. If people come to think that problems like high black poverty rates result from causes that are, in Klein’s phrase, “immutable,” then their reaction will be:

This is not on us, white America, or America broadly, and we don’t have to kind of feel so bad. We can embrace the politics of difference. We can begin removing some of these social supports. Don’t need to have as much affirmative action. Don’t need this employment nondiscrimination stuff. We can cut the size of the social welfare state.

This assertion is different from saying that decent people cannot consider the questions raised by books like Richard Herrnstein and Charles Murray’s The Bell Curve (1994), or the recently published Who We Are and How We Got Here by geneticist David Reich, lest doing so legitimize the beliefs of scientific racists and eugenicists. That is, in itself, a dubious epistemological rule: evidence in support of proposition x might make bad people feel vindicated; therefore, x is false. A 1973 academic conference asserted, for example, that the possibility that heredity accounts for some portion of the differences between races is, in the New York Times’s summary, “unfit to discuss.” One speaker, historian Garland Allen, “said he believed in academic freedom, but only in conjunction with the ideas of the local community. ‘It should be like what goes on in China,’ he said”—at the height of the Cultural Revolution.

Klein’s point, rather, is that the liberal project carried forward since Lyndon Johnson’s Great Society must not be undermined. That project was to use activist government’s entire toolkit to help all Americans, but especially those who had been excluded or deprived prior to the civil rights movement and the other great egalitarian crusades of the 1960s. This quest for what historian Samuel Moyn calls “a fully inclusive social compact” will be undermined if people come to believe that the inequalities we see today result from cosmic injustices, which would mean that the story culminating in our present circumstances has no villains. And if no one’s actions caused today’s disparities, then it’s unlikely that anyone’s actions can cure them.

Klein fears that if people come to view the disparities the liberal project exists to reduce as the consequences of anything other than past and present discrimination (of the invidious kind), people’s feelings of guilt and moral responsibility will diminish. As they do, the liberal project’s political support will come to rest heavily, and precariously, on motives of altruism and enlightened self-interest. It’s imperative, then, that whites not be “let off the hook” for disparities disadvantageous to blacks, which is the effect of considering causes other than discrimination as the source of those disparities. By extension, straight white males must not be let off the hook for disparities disadvantageous to any American outside their diminishing ranks. This, the entirety of America’s inequalities, is on them, all of it, and they must be made to feel very, very bad.

All or Nothing

As an African-American, Sowell has a basis to resent theories about blacks’ innate cognitive shortcomings, virtually all of whose proponents are European-Americans. But his review, published in the American Spectator in February 1995, called The Bell Curve “very sober, very thorough, and very honest.” Sowell accepts Herrnstein and Murray’s presentation of data supporting and opposing the idea of a genetic basis for black-white I.Q. differences, and subsequent declaration that they are “agnostic” on the question.

Sowell is more skeptical than agnostic. He treats the question of innate differences between groups as far from answered, and possibly not answerable, given the analytical tools presently available. It is, nonetheless, askable. His review emphasized that I.Q. scores have increased significantly over the past several decades, too brief a period for genetic changes to account for the phenomenon. As The Bell Curve points out, the 15-point I.Q. gap between whites and blacks is comparable to the gap between whites today and those tested two generations ago. For Sowell, the fact that “the average level of mental test performance has changed very significantly for whole populations over time,” along with evidence that “particular ethnic groups within the population have changed their relative position” on the ranking of test score results, “undermines the case for a genetic explanation of interracial I.Q. differences.”

Sowell clearly believes that culture, as it affects the development of human capital, is a better candidate for explaining inter-group differences than either discrimination or genetics. The “high correlation between the amount of work that different groups put into their education and the quality of their outcomes does not bode well for theories of genetic determinism,” he writes in Discrimination and Disparities. “When we find some race whose lazy students get educational results superior to the results of hard-working students in other races, this would be evidence supporting that hypothesis, but such evidence does not seem to be available.” Asians’ over-representation in New York’s specialized high schools would come as no surprise to Sowell, especially given the proliferation of test preparation centers in the city’s Asian neighborhoods. The Asian American Achievement Paradox (2015), by Jennifer Lee and Min Zhou, points out that the culture of academic achievement among Asian-Americans is so vigorous that a student who brings home a report card with an A-minus is said to have received “the Asian F.”

One might suppose that cultural determinism would be accepted as a more respectful explanation of disparities than genetic determinism. To the contrary. If there is some hereditary basis for disparities disadvantageous to blacks, not only is it futile to devise remedies but also to assign blame…to anyone, white or black. Cultural determinism, however, ascribes agency to disadvantaged American groups, which necessarily recognizes responsibility and the possibility of blame as well. For sociologist Michael Eric Dyson, President Obama’s 2013 address at historically black Morehouse College warned its students “against using racism as an excuse for failure,” and condemned “black pathology, such as absentee fathers.” Such admonitions, Dyson lamented, were seen by white America as “heroic battles against black deficiency.” Another black public intellectual, Ta-Nehisi Coates, dismisses as “lazy” any cultural explanation for disparate outcomes that leave blacks worse off than whites.

The great problem with the Morehouse speech and the Obama presidency in general, wrote Dyson, was the failure to “speak of race in a way that holds whites even partially responsible for black suffering.” As it is for Ezra Klein, the imperative becomes not to let whites off the hook. By contrast, ascribing every disparity adverse to blacks entirely in terms of discrimination holds whites wholly responsible for black suffering. Necessarily, then, it absolves blacks of any responsibility for such suffering, or any need to pursue changes other than demanding that whites renounce their advantages, as in the jihad against “white privilege” or Coates’s advocacy of huge monetary reparations for slavery and segregation.

If all disparities are explained by discrimination, and all Americans fit into one of just two categories—victims of discrimination, or perpetrators and beneficiaries of it—this all-or-nothing frame for determining moral responsibility becomes inevitable. Unfortunately, so does a war between republicanism and pluralism. Ibram Kendi urges us to supplant the flawed principle that all men are created equal with the “perfectly egalitarian” declaration that “all human groups are equal.” This, he says, “is the creed of anti-racism. All human groups are biologically and behaviorally equal; they are all on the same level despite their physical and cultural differences.”

This dictate about group equality, however, nullifies rather than enhances the principle of human equality. A nation dedicated to the proposition that all behaviors and cultures are to be regarded as equal forecloses the prospect of justified pride in order to banish the possibility of self-reproach. If their behaviors and cultures lead some groups to different levels than others, there must be endless, fevered efforts to keep everyone at the same level, as in Mayor de Blasio’s idea that proportional results are the sole criterion defining a fair process. This conception of equality sets it against, and demands that it prevail against, freedom. Constant state interventions will be needed to minimize the consequences, good and bad, of individuals’ choices, habits, and dispositions. For the sake of group equality, the disciplined, responsible, and ambitious will be penalized so that those who can’t or won’t manifest these qualities are rewarded. The result, concludes Thomas Sowell in his series of luminous books, written over the course of a long and very American life, will be an ever less free and democratic nation that will be hard to sustain and impossible to admire.