Bioethics Forum Essay
Will Sociogenomics Reduce Social Inequality?
In her new book, Kathryn Paige Harden is full of hope that insights from genetics will become powerful tools for advancing a left-leaning political agenda. Her hope rests on the argument that if people understand the extent to which their socioeconomic status is influenced by their luck in the genetic lottery, they will understand that they do not merit that status and will become committed to social policies that favor redistributing resources. That is, she hopes the field of sociogenomics, in which social scientists use genetic data to predict outcomes as complex as educational attainment, will be used—not to justify—but to reduce social inequality.
Moreover, Harden is full of enthusiasm about the potential of sociogenomics for creating concrete social interventions that can reduce social inequality. Her enthusiasm rests on the argument that, if social scientists begin to control for genetics when they seek to understand social outcomes, they can start to create social interventions that actually work and contribute to reducing the obscene inequalities that plague our society. Not only is Harden, a behavior geneticist and psychologist at the University of Texas at Austin, aghast at the opportunity costs associated with doing social science without controlling for genetics, she is full of dismay at people on the left who, in her account, let their fear of the evil political potential of sociogenomics get in the way of seeing its potential for good.
I salute Harden for so deftly bringing into the public square her philosophical observation about the compatibility of genetic findings and a left-leaning political agenda, as well as for her scientific observation that controlling for genetic differences could be used to create social interventions aimed at left-leaning political purposes. I share with her not only a progressive political orientation, but an intellectual desire to understand how all variables—genetic and environmental—can help to explain why human beings behave as we do.
But I think Harden’s arguments do not justify her enthusiasm about, and her hopes for, the transformative left-leaning political potential of sociogenomics. (Full disclosure: Kathryn Paige Harden is a member of a working group on sociogenomics that I co-lead.)
Eugenicists Versus Anti-Eugenicists
Harden is acutely aware of the sinful and catastrophic history of people on the right using putative or real insights regarding genetic differences to advance their political views, which have so often reeked, and reek today, of racism. To counter what she understands will be resistance from progressives, she draws a fundamental distinction between “eugenicists” and “anti-eugenicists.” In her account, eugenicists believe that “there is a hierarchy of superior and inferior human beings, where one’s DNA determines one’s intrinsic worth and rank in the hierarchy.” Anti-eugenicists believe that DNA can be one factor in explaining where we end up in the social hierarchy and that where we end up has nothing to do with our intrinsic worth.
The figure on the political right who looms over Harden’s book, and from whom she most energetically seeks to distance herself, is Charles Murray. In 1994, together with Richard Herrnstein, Murray wrote The Bell Curve, in which they professed agnosticism about whether, on average, “Whites” are genetically superior to “Blacks” with respect to intelligence, while simultaneously making clear that they believed that was the case.
To her credit, Harden is also clear about the scientific claims on which she and Murray agree. Like Murray—and like social scientists as hostile to Murray as Eric Turkheimer and Richard Nisbett—Harden takes “general cognitive ability” or “intelligence” to be among the most valid and reliable of the phenotypes that psychologists study. They also agree that “intelligence tests measure an aspect of a person’s psychology that is relevant for their success in contemporary educational systems and labor markets.”
Moreover, like Murray, Harden believes twin studies established that genetic differences can help to explain—in a given population, in a given environment, at a given time—virtually all observed or “phenotypic” differences, including differences with respect to what psychologists call intelligence. Harden emphasizes that twin studies did not show that behavioral traits—or the social outcomes associated with them—are determined by genes alone. Again, they showed that, to varying degrees, genetic variants can partly help to account for the observed differences.
In practice, it has been exceedingly difficult for molecular geneticists to discern which genetic variants are making a phenotypic difference. Harden and Murray are equally aware of the limited success that molecular geneticists had when they used Genome Wide Association Studies (GWAS) in efforts to identify single genes with large effects on the phenotypes studied by the twins researchers.
And like Murray, Harden is hopeful about the new strategy favored by sociogenomicists: creating “polygenic scores,” which relies on the ability of GWAS to detect single base pairs—single nucleotide polymorphisms (SNPs)—with tiny effect sizes. The basic idea of creating a polygenic score is simple: researchers assign weights to the associations between traits and hundreds or thousands of SNPs and add them up to make a score. That idea surely has critics, but it has taken hold throughout mainstream medical genetics and among sociogenomicists, who hope that it will eventually be possible to make predictions about the likelihood that a given individual will exhibit phenotypes as complex as “educational attainment.”
Polygenic Scores and Educational Attainment
The study that forms the backbone of Harden’s book was conducted in 2018 by James J. Lee and colleagues. To get the statistical power needed to detect the tiny associations between SNPs and traits, researchers need DNA samples from gigantic numbers of people. In their effort to create polygenic scores for educational attainment— simply the number of years that someone went to school—Lee and colleagues analyzed DNA samples from 1.1 million people of European ancestry.
Lee and his colleagues found that people with low polygenic scores for educational attainment were far less likely to complete college than people with high polygenic scores: only about 13% of the people in the lowest quartile of polygenic scores completed college, compared with 52% in the highest quartile. Harden emphasizes that the correlation between family income and rates of college completion aren’t much bigger. Although there is a long way from a result such as Lee’s to making a policy intervention of any kind, she makes a fair point when she argues that, if you think family income is important enough to factor into analyses of social outcomes, you should think genetics is too.
Below I will take up Harden’s central example (regarding sex education in Texas) of a genetic study that is closer to the world of policy than is Lee et al.’s. But first I want to consider her boldest theoretical claim, about how starting to recognize the role of genetics in explaining all observed outcomes can help to transform nothing less than what we think justice is.
Sociogenomics and Justice
At the start of the book, where Harden introduces her argument that sociogenomics can be a tool to advance a left-leaning political agenda, there is an odd sentence. She has just announced her own left-leaning political sympathies and expressed her wish that readers with politics very different from hers will nonetheless appreciate the questions she raises. Then she writes, “I invite my conservative readers to remember that justice was an idea that also preoccupied the ancient Greeks, the authors of the Bible, and the Founding Fathers.” That is, she is informing her conservative readers that being concerned with justice is not just a preoccupation of people on the left today. Justice is something that should occupy conservatives, too.
The problem is that she is assuming, not just in that sentence but throughout the book, that justice has one authoritative meaning. Specifically, she assumes that authoritative account was offered by the 20th century political philosopher John Rawls, in his Theory of Justice, which she describes in some detail.
According to Rawls, because people do not deserve their good or bad draws in the social and genetic lotteries, society should be organized to benefit those with the worst draws. This, he argues, is what a reasonable person would recommend if, in advance of creating a society, they found themselves behind a veil of ignorance that prevented them from knowing whether their draws would be good or bad. In this view, where we do not merit our socioeconomic status, gross social inequalities are not only unfortunate. They are unfair. In the left-leaning view, it is not fair to organize a society based on the assumption that people deserve the socioeconomic status they end up with. Moreover, it is fair for the government to tax people for the sake of redistributing resources and thereby reduce inequality.
For those on the right, the most obvious alternative to Rawls was articulated by his colleague at Harvard, Robert Nozick. Like Rawls, Nozick is aware that luck is a central feature of our lives, and that luck in the social and genetic lotteries is a central feature.
The difference between Nozick and Rawls is in how they view those lotteries. For Nozick, the basic idea is that individuals are responsible to play the hand they were dealt as best they can. Contra those who lean left with Rawls, the resulting inequality is not unfair. In Nozick’s view, what is deeply unfair is for the government to take anyone’s winnings in the form of taxes. Taxation, to Nozick, is a form of theft. It steals from citizens what they deserve, based on how they played the hands they were dealt.
Contrary to what is assumed by Harden’s sentence inviting conservatives to think about justice, conservatives have not forgotten that justice is an idea warranting their concern. The problem is that they have a different conception of justice than do we on the left—whether we identify with Rawls or Amartya Sen or Charles W. Mills. And Harden does not grapple anywhere in her book with the reality that when the sorts of facts produced by sociogenomicists fall into the political world, the human beings who interpret them will not all share the conception of justice she takes to be authoritative.
Empirical Data Supporting Harden’s Hope
To be fair, Harden does not rely only on armchair speculation when she suggests that, if people come to understand the extent to which their draws in the genetic lottery influence where they end up in the social hierarchy, they will recognize their moral obligation to redistribute resources. She also appeals to a handful of laboratory and survey studies conducted by Norwegian behavioral economists, which suggest that most people are “more likely to support redistribution when they see inequalities as stemming from lucky factors over which people have no control than when they see inequalities as stemming from choice.”
Harden is well aware of the large social psychology literature that suggests that people use whatever new information comes their way to advance the conclusions they already hold, and she is aware that political conservatives “are reluctant to say that luck plays a role in people’s success.” But her hope for people on the political right is that once they recognize the role of genetic luck, they will move leftward.
Sociogenomics and More Effective Interventions
Her hope for many of us already on the left is that we stop speaking as if we know which social interventions would actually reduce social inequality. To warrant that hope, however, she has to elide the difference between the huge structural changes we have in mind and the sorts of specific interventions she does.
We have in mind changes that would create decent jobs for all, would create good schools in safe neighborhoods, and would make medical care, dental care, nutritious food, clean air, and clean water accessible to all. But Harden has in mind improving specific programs and interventions within our current system. As she points out, a recent review of randomized controlled trials evaluating interventions to improve educational outcomes found that “the large majority of interventions evaluated produced weak or no positive effects compared to usual school practices.”
Not only is she worried about our ignorance regarding what specific policy interventions actually work, but she is appalled by the opportunity costs entailed by programs that we think work but might not. She describes at length the the idea promulgated by the Clinton Foundation (and then taken up by Barack Obama) that the poorer educational performance of low-income children was at least in part caused by the fact that those children did not hear as many words as children born to high-income parents. If, she argues, we can come to understand what is really causing different outcomes in children, we will be better positioned to propose interventions that are effective.
Her central example for using genetic insights to craft interventions that really work—or at least to avoid ones that do not—begins with some of her own research. In Texas, the Education Code mandates that, in her words, “students are legally required to learn that having sex as an unmarried teenager causes emotional trauma.” Moreover, she suggests, the developmental psychology literature seems to back up the state’s claim. Adolescents who have sex at younger ages “don’t go as far in school, report more psychological distress and higher rates of depression, are more likely to use alcohol and other drugs, are more likely to engage in delinquent and criminal behavior, and, in girls, are more likely to have patterns of disordered eating.”
According to Harden, however, the fatal flaw in the logic of the State of Texas is in imagining that early sex causes the bad outcomes with which it is correlated, and the state’s inference that abstaining from sex will “prevent these bad things from happening.” To explain where Texas goes wrong, Harden describes her own clever twin study which showed that, although bad outcomes are correlated with early sex, early sex isn’t causing them.
But the case of sex education in Texas reminds us of the real-world political context into which findings from sociogenomics research will fall. Does anyone imagine that the Texas State Legislature would abandon its concerns about premarital sex upon learning of genetics research showing that early sex is correlated with, but does not cause, psychological problems?
Sociogenomics, Race, and Ancestry
There is one issue, however, that is almost enough to dampen Harden’s optimism about the left-leaning political potential of sociogenomics. She observes that, much as when people hear the word bed they cannot help but hear sleep, when they hear the word genes or intelligence they cannot help but hear race. Specifically, people in the U.S. cannot help but be reminded of the original evil claim upon which the catastrophic history of racism in the U.S. rests: that whites and Blacks are distinct “races,” and that whites are, by nature, more intelligent and morally superior to Blacks. Again, her book’s central rhetorical challenge is to establish the vast gulf between the eugenic pseudoscience of the past and present, which was and is preoccupied with race, and the new anti-eugenic science of sociogenomics, which has no truck with the concept of race, much less any desire to establish the superiority of one “racial” group over another.
In addition to emphasizing that race is a biologically meaningless concept, Harden also establishes distance between herself and Charles Murray by addressing the logic of his infamous argument that if genetic differences help to explain why some individuals do better in school than others, and if, on average, the group of people called whites in this country does better in school than Blacks, there is reason to speculate that the group called “whites” has a genetic advantage over the group called “Blacks.” Specifically, Harden points out that Murray commits the common fallacy of moving from data concerning genetic differences among individuals within one group or population to making a comparison between groups or populations.
Although Harden emphatically rejects the concept of “racial group” as biologically meaningless and thus of no practical value for sociogenomicists, she also explains why she accepts the distinct concept of “ancestral group” as biologically meaningful. Researchers use the concept of ancestral group (or population) to increase the chances that, when they detect genetic differences, they are detecting variants that are ultimately making an observed difference—as opposed to being causally uninteresting variants that accumulated as groups of humans migrated out of Africa and procreated in regions separated by geographical barriers. The precise biological meaning of ancestral group is far less clear than Harden or anyone doing genomics would wish, but it is the term she uses in her book, and I am trying to take her on her terms.
Moreover, Harden speculates that if it ever becomes possible to compare ancestral groups, the average differences between them will likely be minimal. Further, she rightly demands that anyone who wants to talk about group differences should remember the scientific fact that if any ancestral group has a genetic advantage with regard to educational attainment, it won’t necessarily be the ancestral group or groups that currently appear to. Her hunch and demand are fair enough, if not sources of enormous comfort to those of us trying to wrap our minds around the political and social ramifications of making scientifically legitimate comparisons between groups.
From Differences Between Ancestral Groups to Differences in SES
Although Harden makes clear that it is scientifically meaningless to attempt to compare racial groups, and explains why currently it is not scientifically possible to compare ancestral groups, she does make the case for comparing a different sort of group: SES. Whatever might be the scientific problems with comparing SES groups, Harden uses a study that compares them to press her case for the progressive political value of sociogenomics. Here I’m talking about a paper by Daniel Belsky et al., which is based on the data collected in the Lee et al. educational attainment study that serves as her book’s central scientific exhibit of sociogenomics done well. Belsky leans left politically and his study purports to offer insights only regarding people of European ancestry.
The Belsky paper reported that people who are born into low SES environments and have a high polygenic score for educational attainment often do not go as far in school as people with a low polygenic score and high SES. In other words, as Harden observes, Belsky et al.’s analysis shows that a huge amount of “genetic talent” is wasted when people with high polygenic scores do not receive the opportunity to express their talent.
Harden is, however, silent about another feature of the same data. On average, people from the high SES group seem to have slightly higher polygenic scores than people from the low SES group. That is, on average—according to these data—people with high SES have slightly more “genetic talent” than people from low SES. The very same sociogenomics study presents data that can be—and have been—cited to support a very different message from Harden’s. Conservatives and libertarians can find support for what we might call their “talent rewarded” message, which presupposes that our society rewards people according to what they do with their God-given talent.
So, yes, the findings from a sociogenomics study such as the one conducted by Belsky et al. are compatible with the left-leaning view that we in the U.S. have created structures that waste huge amounts of genetic talent. But findings from the study are also compatible with the right-leaning view that the rich are rich because they have taken their lucky draw in the genetic lottery and earned their place in the social hierarchy. Findings from the study are compatible with the view that we urgently need to restructure our society and the view that our society reflects the natural order of things.
Affirming Individual Differences
To be fair, Harden’s focus is not on using genetics to compare ancestral groups or SES groups. Rather, it is on using genetics to contribute to understanding phenotypic differences among individuals. But even if we could bracket or ignore all attempts at using genetics to make comparisons between groups, there remains what can be for us on the left the anxiety-producing fact that genetics always entails attempting to understand how genetic differences help to explain why some individuals are more likely than others to exhibit traits that are highly valorized in this society.
As someone who experiences that anxiety but who is also committed intellectual transparency, I want to acknowledge a question that might have occurred to others, in light of my comments above about the social psychology literature which suggests that people use information to advance whatever view they already hold. If I was right above to be skeptical about the positive impacts of genomic information that Harden hopes for, shouldn’t I be more skeptical about the negative impacts I fear? I think that’s an important question. But that question does not figure in Harden’s strategy for reducing the anxiety of her colleagues on the left.
Rather, Harden’s anxiolytic strategy is to hammer home the profound point that moral equality does not depend upon phenotypic sameness. Always the optimist, Harden offers the recent history of Deaf culture to make the case that we needn’t be so anxious about the prospect of sociogenomicists helping to explain why people embody exceedingly complex traits to different extents. As Harden observes, Deaf people have succeeded in helping people who communicate with vocal cords to recognize that there is nothing “inferior” about communicating with ASL. Genetic differences, whether associated with intelligence or hearing, don’t have to be imbued with value or moralized. We don’t have to think of people who have lots of “genetic potential” for intelligence to be morally superior to people who have less, any more than we have to think of people who hear in the typical way as being morally superior to people who don’t. What we should do, Harden wisely counsels, is recognize that our genetic differences are entirely irrelevant to our equal moral or human worth. Her fundamental and noble aspiration is to affirm, as she puts it, “difference without hierarchy.”
Nowhere in her book, however, does Harden mention that people on the right also profess a commitment to the moral equality or equal worth of all persons. More importantly for evaluating her foundational opposition between eugenicists and anti-eugenicists, nowhere does she acknowledge that Charles Murray himself explicitly says that we need to recognize that people are phenotypically different, but “of equal human worth.”
Harden and Murray
To establish her book’s foundational opposition, Harden inadvertently trades on the fact that the terms superior and inferior can be used in a descriptive or prescriptive sense. They can refer to an observation about how persons, in fact, compare with respect to some trait, and they can refer to an estimation of the relative moral worth of persons.
To shore up her claim that Murray thinks that some people are morally or intrinsically superior to others, Harden points to a passage in The Bell Curve where he and Herrnstein complain that “it has become objectionable to say that some people are superior to other people.” She reads him as saying that some people are morally or intrinsically superior to others. But I read him to be lamenting exactly what Harden laments throughout her book: “people on the left” find it objectionable to say that some individuals are superior to others with respect to what psychometricians call intelligence. Harden and Murray are both battling the same real or imagined foe: people on the left who deny that genetic differences have anything to do with differences in all traits, including highly valorized ones like intelligence or educational attainment.
As much as Harden needs Murray to hold the view she attributes to him, he explicitly disavows it in his books, nowhere more explicitly than in last year’s Human Diversity. I do not know what is in Murray’s heart. And I do know that Murray brings to his arguments in The Bell Curve and to Human Diversity the same profoundly racist assumption that the environments of “whites” and “Blacks” in the U.S. are equal enough to raise the question about the role of genetics in explaining the observed educational achievement gap. But intellectual honesty demands that we recognize his explicit rejection of the view Harden attributes to him. In the final chapter of Human Diversity, he explicitly states that what we need is to come up with “a secular understanding of the truth behind the old formulation, ‘We are all equal in God’s eyes.’”
Murray does not only share with Harden an explicit commitment to the equal moral worth of all. Like Harden, Murray also explicitly states his commitment to us learning how to celebrate a diversity of talents rather than deny their existence. Perhaps most importantly, nowhere does Harden acknowledge that she and Murray are in fundamental agreement that none of us deserves or merits our draw in the genetic lottery.
The differences between them are not nearly so simple as that he is a “eugenicist,” who believes that some human beings are intrinsically or morally superior to others, and she an “anti-eugenicist,” who believes the opposite. The political world into which findings of sociogenomics will fall is staggeringly more complex than the one Harden seems to imagine when she describes the positive and transformative power of such results.
Human beings have been aware forever that Luck or Fortune or Fate is an utterly central feature of our existence. It is Panglossian to imagine that, once we lay over what we have always known with a genomic gloss, many of us will be moved leftward. One can appreciate Harden’s achievement in forcing people on the left to recognize that sociogenomics is not necessarily a tool of the right, without signing onto her extravagant hope and enthusiasm about sociogenomics becoming a powerful tool for reducing social inequality.
Erik Parens is a senior research scholar and director of the Initiative in Bioethics and the Humanities at The Hastings Center.