Why DNA Is No Key to Social Equality: On Kathryn Paige Harden’s “The Genetic Lottery”

DNA PLAYS A major role, indeed a starring role, in generating socioeconomic inequality in the United States, according to Kathryn Paige Harden, a behavioral geneticist at the University of Texas. In her provocative new book, The Genetic Lottery: Why DNA Matters for Social Equality, she contends that our genes predispose us to getting more or less education, which then largely determines our place in the social order. This argument isn’t new. It has appeared perhaps most notoriously in Arthur Jensen’s infamous 1969 article in the Harvard Educational Review (“How Much Can We Boost IQ and Scholastic Achievement?”) and in The Bell Curve by Richard Herrnstein and Charles Murray, published in 1994. Harden updates the argument in three ways. First, she grounds her claims in cutting-edge genomic research utilizing a technique called genome-wide association. Second, she explicitly rejects the racist claims made by Jensen, Herrnstein, and Murray, arguing that genetic differences account only for socioeconomic inequality between individuals within racially defined groups, not between racially defined groups. Third, she argues that attributing socioeconomic inequality to “nature” rather than “nurture” does not absolve society from ameliorating it.

In constructing this argument, Harden tries to thread a narrow political needle and thereby bring liberals and conservatives into agreement. On the one hand, she wants to convince political liberals — who may already support redistribution but are loath to view inequality as genetic in origin — that genetic differences between individuals cause socioeconomic inequality, and policies aimed at social justice must therefore take genetics into account. On the other hand, she aims to convince political conservatives — who may be more open to genetic explanations for socioeconomic inequality but opposed to redistribution — that the genes predisposing individuals to success or failure are allocated randomly at conception, and it is therefore unfair to distribute resources on the basis of this genetic luck.

While we admire Harden’s social justice aims, we remain unconvinced by her biological explanation for socioeconomic inequality. In making her case to liberals, we believe Harden extrapolates beyond what current scientific results allow. She has expanded an interesting but narrow finding — that DNA influences educational attainment and other social outcomes for individuals of European ancestry in high-income countries — into a unified theory of society and a basis for sweeping social reform. Her extrapolations are unwarranted by the research she cites, particularly as the research does not translate to people from other ancestry groups. We also think she is unlikely to convince many conservative readers that the randomness of Mendelian inheritance should lead to generous income redistribution policies. Instead, her efforts serve to naturalize the current distribution of power in society, even as she implores the genetically privileged to be more compassionate to the genetically disadvantaged.

Genome-Wide Association Studies

To make her case that genes matter, Harden relies on the findings of genome-wide association studies (GWAS), a cutting-edge technique in molecular genetics. In GWAS, hundreds of thousands of individuals — sometimes over a million — have their DNA analyzed at hundreds of thousands of loci, sometimes over a million. Researchers genotype single nucleotide polymorphisms, or SNPs (pronounced “snips”), which are spots on the genome known to vary among individuals. They then examine whether variation at each SNP correlates statistically with a disease, such as diabetes or schizophrenia, or with a trait, such as height or educational attainment. These complex traits and diseases are considered “polygenic,” meaning that many SNPs are involved, each having a tiny influence on the outcome. Therefore, rather than focusing on a single gene, geneticists sum up the effect of each relevant SNP to produce an individual’s “polygenic index,” or genomic propensity toward the outcome in question.

Harden’s focus is on the most recent GWAS of educational attainment. She takes a folksy approach to explaining how GWAS are conducted and polygenic indices calculated, resulting in one of the most engaging overviews of the topic we have encountered. Unlike some proponents of these methods, Harden is honest about what GWASs can and can’t tell us and about what polygenic indices do and don’t mean. She demonstrates that, on average, people with very high polygenic indices for educational attainment are more likely to graduate from college than are people with very low polygenic indices for educational attainment. These people will also be more likable to others, have more prestigious jobs, make more money, and accumulate more wealth. Harden terms them — us, really, as she assumes her readers fall into this category — the “naturally advantaged.” After a long discussion of the meaning of causality in the social sciences, Harden concludes that genetic differences cause differences in educational attainment in what she terms a “thin” sense: we don’t know the mechanisms at play, but different genomic variants produce different outcomes.

Yet in her effort to convince readers that genes matter, Harden overstates the degree to which they matter. She tells readers that, “in samples of White people living in high-income countries, a polygenic index created from the educational attainment GWAS typically captures about 10–15 percent of the variance in outcomes like years of schooling, performance on standardized academic tests, or intelligence test scores.” She compares this figure to that for household income, which accounts for 11 percent of the variance. What Harden doesn’t tell readers is that much more of the variance is explained by parental education: about 17 percent when only one parent is considered and over 20 percent when both are. The polygenic index for educational attainment therefore captures an underwhelming amount of variance in educational attainment and other socioeconomic outcomes — certainly not enough to justify putting it at the center of policy solutions.

Harden expects that, as GWAS samples grow, the polygenic index will become more predictive, but exactly how it predicts educational attainment is not at all straightforward. Consider how Harden chooses to present the 10–15 percent figure, making it account for educational attainment through biological mechanisms. She tells her readers that the genes involved are expressed preferentially in our brains, where they increase the bearer’s intelligence, executive function, grit, and perseverance — the cognitive and non-cognitive skills rewarded in our educational system and labor market. What Harden doesn’t tell us is that these genes are also “expressed” in our environments. People with higher polygenic indices for educational attainment are both more likely to be raised by parents with higher socioeconomic status and to go to well-funded schools. A study of adoptees suggests that about half of the effect of the polygenic index operates through these indirect mechanisms. Harden acknowledges this complex causality, demonstrating that small differences early in life lead to children being placed into environments that magnify those differences. For her, these are all genetic causes because, with different genes, we also would experience different environments. By identifying social mechanisms as “genetic,” Harden is naturalizing them, attributing the inequality they produce to the individuals who benefit from or are harmed by them rather than to the policies and practices that privilege some genotypes over others.

Race in Genetic Research

Although Harden contends that genetics play an important role in the attainment of education, wealth, and other social goods among White Americans, she emphasizes that “there is zero evidence that genetics explains racial differences in outcomes like education.” Genes can’t explain why White Americans are more likely to graduate from college than are Black Americans. This is an important lesson as research in behavior genetics is routinely called upon — sometimes by behavior geneticists themselves — to authorize White supremacy. The fallacy, Harden tells readers, is in moving from the claim that genes play a causal role in explaining differences in social outcomes for individuals within a racially defined group to the claim that genes explain average differences in social outcomes between groups.

To date, GWAS of educational attainment have only been done on White people with European genetic ancestry, using samples from North America, Europe, and Australia. Most GWAS use ancestrally homogeneous samples in order to reduce the risk of mistaking random variation between ancestry groups for SNPs that influence the outcome in question. It is an empirical question whether polygenic indices based on these studies are “portable” to members of other groups. Current research suggests that, for many traits, GWAS results for individuals of European descent are in fact not portable: they are inaccurate predictors for individuals of non–European descent. For example, the existing polygenic index for educational attainment accounts for much less of the variance in years of formal education among older African Americans (1.6 percent) than it does among older Americans with only European genetic ancestry (10.6 percent).

Harden recognizes that GWAS have been nondiverse, with results not transferring adequately between ancestry groups, so the benefits of genetic findings will only accrue to people of European ancestry. She advocates for investing “preferentially” to address these inequities and expects that scientists will soon develop “a polygenic score that is as strongly related, statistically, to academic achievement in Black students as it is in White students.” But there is no evidence that such research is underway, and geneticists with whom we have consulted don’t even know what this would look like (for example, is it a single polygenic index for everyone or separate indices for White and Black students?), much less how it would be carried out. Harden’s expectation doesn’t account for the fact, which she acknowledges, that genes play a smaller role in educational outcomes under adverse conditions. This phenomenon has been demonstrated for students with low socioeconomic status, and for female students, for whom the polygenic score for educational attainment was less predictive earlier in the 20th century and more predictive later in the century as gender discrimination abated. Given the persistence of racial discrimination, polygenic scores for African Americans may simply be less predictive in general than they are for White people, which means that the returns on genetic research will continue to be distributed inequitably. The dilemma Harden faces is this: advocating for policy based on current GWAS findings exacerbates racial inequities, but expanding genetic research to include non-European ancestry populations is not urgently needed for addressing the existing inequities either. Institutional antiracist interventions are the best antidote to the social inequalities faced by African Americans.

Harden works to distinguish her view as explicitly antiracist, in contrast with the racist projects of other behavior geneticists, such as the aforementioned Jensen and Herrnstein. Her commitment to antiracism is commendable, and she provides a clear explanation for why race is not a meaningful scientific category. And yet her book is plagued by the race concept. As she explains, race is not a biological or genetic category, and thus not a valid way to make group comparisons at the genetic level. Race concepts are meant to indicate discrete subgroups of the human population. But discrete, genetic, racial essences are in fact nowhere to be found. No genetic variants exist in all Black people and absent in all White people. Rather, human genetic variation is “clinal,” gradually changing with geography. Scientists therefore group people on the basis of genetic ancestry: gene variants that cluster in people whose ancestors came from specific geographical regions. Most people have complex, far-reaching ancestries that cross the discrete social category of race. Thus, race and ancestry are not interchangeable concepts. After taking the time to distinguish between race and ancestry, however, Harden proceeds to blur these categories, often using White as a synonym for people with only European ancestry and Black as a synonym for people with any African ancestry.

But this is too simplistic. Race relies on politically motivated rules for assignment, not biological ones, which means we can’t make inferences from someone’s socially identified race about their ancestry. The one-drop rule, which had the racist purpose of keeping the White race “pure,” says that to be White in America, you must have no proximate ancestry in Africa; if you have any appreciable African ancestry, you are Black. This explains why 90 percent of Black people in the United States have some European ancestry, while only 0.3 percent of self-identified Whites have African ancestry. It also shows how membership in a race does not reliably or consistently indicate what is “under the hood.”

Even genetic ancestry categories — for example European and African — are not the firm biological groupings Harden would have readers believe. Since genetic ancestry is nothing more than the variants held in common by people with a geographically bounded family tree, analysts can choose how tightly or loosely to group people based on similarities in their genomes: Are the groups at the level of a city? A country? A continent? In GWAS, which require enormous sample sizes, the unit of analysis is typically the continent, no doubt influenced by the convention in the US of grouping people into continental races. This is not a rigorous way to categorize people because those continental-level populations are not uniform. Take, for instance, a fact Harden herself acknowledges: two randomly selected people with majority African ancestry may have less in common genetically than do a randomly selected person with European ancestry and another with Asian ancestry. Thus, one continental ancestral population may be capturing far more diversity than another. This reliance on continental ancestry plagues genetics more broadly and is not unique to Harden. The risk is that geneticists who disavow race with one hand reify it with the other by using continental ancestry as their primary analytic category.

Despite her avowed antiracism, Harden entertains the possibility, raised by the likes of David Reich and Sam Harris, that racial inequality might be explained by genes. We are to imagine that at some point in the future we find out that people of European descent have genes that endow them with superior cognitive abilities. After assuring readers that such a finding is unlikely, she contends that genetic differences are inescapable and so “people’s moral commitments to racial equality are on shaky ground if they depend on exact genetic sameness across human populations.” Here again, she slips between “race” and “ancestry.” Her target is historian Ibram X. Kendi, who defines a “biological antiracist” as someone who believes that “there are no genetic racial differences.” But Harden is reading Kendi out of context. In How to Be an Antiracist, Kendi clearly rejects the view that biological, essentialist differences, whether genetic or otherwise, explain behavioral and social differences between Blacks and Whites. As he says: “Biological racism rests on two ideas: that the races are meaningfully different in their biology and that these differences create a hierarchy of value.” Kendi says that antiracism requires rejecting both views, not just the latter, as some would have it. Kendi does not claim that there are no genetic differences between people; indeed, he acknowledges genetic variation between individuals based on geography, but rightfully denies that this is racially meaningful. Harden, who rejects race as a biologically valid category, should agree. Her target is in effect a Strawman view, which no serious scholar holds, that there is no genetic variation across populations.

Harden inexplicably fumbles here while trying to make a point that serves as a major through-line in her book. She believes liberals reject the possibility for genetic explanations of social and racial inequality because they assume that this would justify that inequality. Instead of resisting the empirical possibility, however, she urges liberals to reject the mistaken moral inference. Harden’s vision for improving society is liberal-egalitarian, with John Rawls her philosophical touchstone. A major tenet of liberal egalitarianism argues that, since we are not responsible for the conditions into which we are born, it is unfair for accidents of birth to influence our chances for success in society. Accidents of birth include our genetic endowment. Since our genes do not justify social inequality, liberals need not antecedently reject the possibility of genetic difference.

We think Harden misunderstands the liberal resistance here. After all, it was the liberal egalitarian disability activists who fought for disability rights for the genetically, congenitally, and otherwise disabled in the 1970s. What liberals reject is any facile entertainment of the “cognitively superior European” thought experiment because it flirts with centuries of scientific racism. It is not an innocent example, and it is no surprise that the thought experiment isn’t instead a speculation about African cognitive superiority. Liberals balk because there is absolutely no data, as Harden herself acknowledges, suggesting the existence of biological differences that could provide an explanation for the appalling socioeconomic disparities created by centuries of racist policy. GWAS can’t provide such data (lack of diversity aside) because they can’t tell you if a SNP is, for example, making you smarter, or making other people treat you differently. This is a response Harden could have given to the Reich/Harris speculation that would have fit with her rejection of racial genetics. Unfortunately, then, an otherwise clear and helpful analysis of the illegitimacy of race as a scientific category and the fallacy of applying GWAS with European discovery samples to non-European ancestry groups backtracks on its own aims. It risks leaving the reader dangerously confused.

Genetics and Social Policy 

Just as Harden misdiagnoses the liberal opposition to the suggestion that individuals of European descent may have more “high-education” variants than people of African descent, so too does she misdiagnose the liberal opposition to her claim that “genetic differences between people cause social inequalities.” Harden suggests three sources for this opposition: we are ideologically committed to genetic sameness, we have bought into the eugenic assumption that differences caused by genes are immutable, and we are wedded to the idea of meritocracy and don’t want to believe that merit is allocated on the basis of genetic luck. None of these is the case. Rather, our problem with Harden’s focus on genes as the primary cause of social inequality is that it elides the role of power. It presents our existing social hierarchy as the result of (genetically driven) success or failure at the individual level, neglecting the policies and practices that systematically benefit some types of people (such as employers, landlords, and creditors) at the expense of others (such as employees, tenants, and borrowers). It renders invisible the ways in which the wealthiest members of our society get ahead through the active exploitation of the most vulnerable. Harden’s suggestion that some people are poor because their genes make them less capable of the kind of hard work that produces value to society willfully ignores the millions of people who work hard and produce value that isn’t shared with them. It isn’t shared because the corporations they work for lobby to keep the minimum wage stagnant, undermine union power, classify employees as contractors, and avoid paying taxes.

We are therefore unconvinced by Harden’s claim that social policy must take genetics into account in order to be effective. Harden draws an analogy between social status and vision: just as some people have genes that help them see better, some people have genes that allow them to succeed in a modern industrial society. Glasses can solve vision problems without genetic engineering, she tells us; so, too, can social policy ameliorate inequality. This is true. But this is also where the analogy begins to break down. Glasses are prescribed without knowledge of the specific genetic (and other) mechanisms that cause poor eyesight, whereas Harden claims that interventions aimed at promoting educational attainment must focus on those with low polygenic indices. Further, she claims that no existing intervention will work because none takes genetics into account. Yet she also demonstrates that wealthy kids with the lowest educational polygenic indices complete college at a higher rate than do poor kids with the highest educational polygenic indices. Effective interventions that are not genetically informed clearly exist, but they are only available to children of well-off parents. The polygenic index for educational attainment is therefore a crude indicator of which kids need help, and it is also limited to children of European ancestry. It is unclear how individual-level genetic data could be used equitably in the diverse classrooms of the United States.

Harden rejects universal interventions as too costly and inefficient, effectively proposing genetic testing as a new form of means testing without acknowledging the issues that plague existing means-tested programs. She expects that genetically informed interventions will be less politically vulnerable than universal programs, citing studies that show general opposition to inequality based on “luck.” Convincing conservatives that the “merit” they so value is actually the outcome of “nature’s Powerball,” she contends, will help them treat the losers of this lottery with more compassion. This critique of meritocratic logic, however, leaves intact the basic (and incorrect) premise that the labor market straightforwardly rewards people for the value they produce in the economy. Harden never challenges the notion that some people are wealthy because they contribute more while others are poor because they contribute less. Her intervention is simply to explain that the ability to produce value is the outcome of genetic luck rather than earned merit and therefore shouldn’t be the sole basis for resource allocation.

Such an argument elides the structural causes of inequality in our society and is unlikely to align conservatives behind redistributive solutions. Those who are poor probably won’t agree with Harden’s contention that their genes make them less capable of creating value. (Harden assumes her readers don’t fall into this category.) Wealthy conservatives may be convinced that, while they themselves have the genetics to succeed, their children may not be so well endowed due to the random meeting of sperm and egg, which is the real lesson of The Genetic Lottery. But recall that the polygenic index for educational attainment accounts for only about 10–15 percent of the variance in actual educational attainment, and that wealthy kids with low polygenic indices are more likely to graduate from college than are poor kids with high polygenic indices. In other words: the children of the wealthy, regardless of how they fare in nature’s Powerball, are likely to remain wealthy and do not need redistribution. Indeed, it is the current inequitable social order that allows the wealthy to pass their status along to the next generation, even when their kids don’t share their genetic endowment.

Harden envisions a more just society as one organized under John Rawls’s “veil of ignorance,” where everyone agrees on how primary social goods will be distributed before knowing the specifics of their identity or what their position will be. This is a nice idea in “a hypothetical situation in which everyone is on an equal footing and so can come to a fair agreement about the principles of justice.” We do not live in that world. Promoting social justice in the real world requires understanding the systems of power that structure our lives, and these can’t be reduced to genetic randomness.

In imploring us to treat the genetically unlucky with egalitarian charity, Harden identifies individual bodies as the primary source of inequality rather than the social institutions that allow those with structural power to get ahead by exploiting those with less power. Ultimately, her focus on genetics as a fundamental cause of social inequality reduces her version of social justice to benevolent paternalism.

¤

Brenna M. Henn is an associate professor of population genetics in the Department of Anthropology, the Center for Population Biology, and the Genome Center at the University of California, Davis. She is co-director of the Northern Cape Tuberculosis Project and published extensively on African genomics in Cell, Nature, and Proceedings of the National Academy of Sciences, among others.

Emily Klancher Merchant is an assistant professor of Science and Technology Studies at UC Davis. Her research focuses on the history of the quantitative social sciences in the 20th and 21st centuries. She is author of Building the Population Bomb and co-editor of Navigating Time and Space in Population Studies, and has published in such journals as Proceedings of the National Academy of Sciences, Population and Development Review, Social Science History, and the Journal of Policy History.

Anne O'Connor is a PhD Candidate in Sociocultural Anthropology at UC Davis. Her research focuses on gene drives for mosquito control. She is interested in the ways in which concepts of reproduction and change are reformulated with novel scientific techniques.

Tina Rulli is an associate professor in the Department of Philosophy at UC Davis. She specializes in bioethics and normative ethics, with a focus on genetic and reproductive technologies. She has published in venues including the Journal of the American Medical Association, Bioethics, Hastings Center Report,Kennedy Institute of Ethics Journal, and the Journal of Medical Ethics.