View from
The Left

Review: “The Genetic Lottery: Why DNA Matters for Social Equality”

Now, given that genetics matters for these things, which, in turn, drive inequality, we should take genetics seriously if we are truly committed to egalitarianism.”

Books published by university presses do not, in general, generate significant discussion or media attention. Much like their shorter counterparts, peer-reviewed journal articles, books such as these put forward a set of ideas to be considered and discussed by other academics as opposed to being aimed at the public at large. This is why scientists also tend to write popular science books designed to make their ideas more accessible, books that often prove quite successful.

The Genetic Lottery: Why DNA Matters for Social Equality is one of those less common cases of an academic book generating considerable discussion. Written by Kathryn Paige Harden, a behavioral geneticist and professor of psychology at the University of Texas at Austin, and published by Princeton University Press, the book has already generated considerable media buzz. In the short period leading up to and after the book’s publication, Harden has been profiled by The New Yorker, appeared on the MSNBC show Morning Joe, and produced many conversations on social media and in other publications.

Much of this discussion has been, for better or worse, very polarized. For this reason, this is not intended as a regular book review but, instead, as a discussion of both the book’s contents and the reactions to it.

As Harden readily admits, genetics—and behavioral genetics in particular—is a controversial topic due to its historical entanglements with the eugenics movement of the early 20th century, which aimed to use genetic inheritance to naturalize and justify material inequalities between people and races and, in the most extreme cases, to advocate policies like mass forced sterilization and genocide. It should go without saying that these are not things that she advocates, so let us look instead at the basic arguments of the book.

The first part deals with the empirical and technical aspects of behavioral genetics and why, as the title suggests, it matters for social equality. For now, I will leave aside what “matters” means exactly in this context, and this is something to which Harden devotes an entire chapter. She focuses on why it matters in one particular social phenomenon—namely, educational attainment, while devoting some attention to related issues such as employment, income, and cognitive abilities. Now, given that genetics matters for these things, which, in turn, drive inequality, we should take genetics seriously if we are truly committed to egalitarianism. The second part of the book is dedicated to fleshing out this argument.

Part II of the book, entitled “Taking Equality Seriously” ought to be, in my view, the most uncontroversial, if one accepts at least some of the arguments in Part I. In a nutshell, what Harden wants is for society to treat genetic differences between individuals in the same way we treat family income—namely, as an accident of birth that unevenly distributes moral luck among different people. Some people, by no merit of their own, are born with the luck of a high-income family, which affords them a disproportionate amount of opportunity. This, of course, means interventions designed to level the genetic playing field are necessary in the same way cash transfer programs are used to mitigate the effects of family income disparities.

Despite this, some of the criticism has still focused on the perceived ethical implications of the book. For example, Tom Scocca, politics editor at Slate  wrote a review in his Substack blog entitled “The nice kind of calipers.” Now, while Scocca acknowledges that, contrary to authors such as Charles Murray, Harden actively contests notions of genetic racial differences in intelligence, Scocca still argues that both Murray and Harden are part of what he calls the American phrenological project. As he explains:

“Whatever higher purposes an individual researcher may have in mind, there is only one question the phrenology business has ever sought to answer: Isn’t it right that things are the way they are?” [Emphasis is the original]

The reason I point this out is that Harden anticipates this kind of reaction repeatedly throughout the book, while explaining why she wants the exact opposite. In Harden’s view, the fact that some people have genetic advantages while others have disadvantages cannot be seen as the reason to accept the unequal outcomes that result from these differences. Instead, we should prioritize the opposite: to have social arrangements such that they will not have an impact on an individual’s well-being.

In fact, at times it almost seems that Harden does not take her philosophical positions to their ultimate conclusion. She quotes often from the political philosophers Elizabeth Anderson and John Rawls, yet, at times, her position regarding egalitarianism seems more radical than that of the socialist theorist G. A. Cohen, who saw Rawls’s theory of justice as insufficiently egalitarian.

Before moving to the more controversial issues like educational attainment, Harden asks us to consider deafness, a trait that is indisputably rooted in genetics and a disadvantage in many aspects of life:

“The genetic lottery that produces deaf children or hearing children is a ‘natural fact’ that we can no more criticize as fair or unfair, as just or unjust, than we can be morally outraged that lightning struck in our backyard and not our neighbor’s. This ‘arbitrariness found in nature,’ however, need not—indeed, should not—be a contingency to which we are resigned. Deafness need not produce an unchangeable order, in which the Deaf are ascribed to a permanent underclass.”

And while Rawls does discuss biological inequalities, the language of luck and lotteries feels much closer to Cohen who, in addition to being a socialist, was part of the so-called luck egalitarian philosophical tradition. Here is Cohen describing his conception of socialist equality of opportunity in his 2009 book Why Not Socialism?:

“When socialist equality of opportunity prevails, differences of outcome reflect nothing but difference of taste and choice, not differences in natural and social capacities and powers. So, for example, under socialist equality of opportunity income differences obtain when they reflect nothing but different individual preferences, including income/leisure preferences.”

But as Harden argues elsewhere, non-cognitive character attributes such as grit and hard work, motivation and curiosity are also part of genetic luck. And while grit and hard work can easily be classified under the natural capabilities that Cohen mentions, it is perfectly possible to consider curiosity and motivation as part of someone’s preferences. If that is, indeed, the case then Harden is asking us to go beyond Cohen’s socialist equality of opportunity.

That is not to say no one is trying to use genetic research in order to naturalize inequality or, as Scocca puts it, to argue that it is just right that things are the way they are. Charles Murray and Richard Herrnstein are quite explicit about this in the last chapter of their 1994 book The Bell Curve. They suggest that rather than committing ourselves to the kind of egalitarianism inherited from the French Revolution, we might look elsewhere, to the rigid social structures of Confucian China or the caste system of India. Since, in their view, society is already stratified by intelligence, some of which is genetic (and racial) we might as well make it official. In other words, yes, it is right and just that things are just the way they are.

But, as I hope I have shown thus far, this is not what Harden argues. In fact, she argues the opposite. Yet it seems the mere suggestion that some differences in abilities are biological is enough to put Murray and Harden in the same category despite very clearly opposed moral commitments.

This is a problem because it suggests that the mere act of looking for natural causes of certain inequalities is enough to naturalize those inequalities. Of course, the two are related, and the line that separates them is a thin one, but they are not the same. Harden manages to make the distinction clear. The former (i.e., looking for natural causes) is purely descriptive and non-deterministic. The latter is always deterministic and often prescriptive as well.

All of this, however, only makes sense if the more descriptive first part of the book, entitled “Taking Genetics Seriously” makes its case successfully. For the most part, it does. It only falters, in my view, regarding claims of causality. This is a subtle point, but it is one that matters and likely drives some of the negative reaction to the book, both justified and unjustified.

I have no background in biology or genetics, so I attempt to make no claims whatsoever about the technical aspects of DNA research that are presented in the book. For this reason, I want to analyze the idea of causality advanced by Harden from the point of view of the philosophy of science. After all, “causation” is not a purely empirical concept, and it requires some kind of metaphysical commitments and assumptions.

The reason why genetics matters for social equality, in Harden’s argument, is because genetics causes us to be different in certain characteristics such as cognitive abilities, motivation, preferences for hard work, among others. These, in turn, are one of many of the drivers of social inequality. As stated before, Harden is clear that not only does this not imply any kind of moral hierarchy, but it is something that can and should be socially ameliorated. She does, however, insist that this can be changed, even if genes cause these inequalities strictly on an individual level.

Harden argues that the first hint that genes have a causal relation with these individual attributes is that they are correlated. She then devotes an entire chapter to explaining exactly what she means by causation and why. This, of course, is necessary since, as any undergraduate will recite, correlation does not imply causation. We need to know more than the fact that two things are observed together regularly in order to know that one causes the other. Instead, what we need is to know whether one thing makes a difference in whether another thing happens. To quote Harden, “[T]o say that X causes Y is to say that, if X had not happened then the probability of Y happening would be different.”

Harden then explains why causation can be understood in different ways. Specifically, she introduces two concepts, “thick” and “thin” causation. The former is the one most people probably have in mind when they think of the word causation. This means that X is the cause of Y, if we can say with certainty that Y would not happen if X did not happen either, and, moreover, we know the specific mechanism by which X influences why. As an example of this, Harden mentions Down’s syndrome. In this case, it is known what specific mutation of a particular gene causes it, and through what biological processes. This makes it “portable,” which is to say, we can infer that this will remain true in other contexts, rather than the specific one we observed.

On the other hand, thin causation does not require that we know the mechanism. All that it needs is that we can reliably establish that in the absence of X, the probability of observing Y will be reduced. As an example, she cites a study in which a group of researchers essentially established a fully functioning foster care system for the thousands of orphans that had been placed in state-run orphanages at the end of the 20th century in Romania due to the country’s abortion ban. The researchers randomly assigned a group of children to foster care families so that the only difference between the two groups was an adoptive family rather than, for example, gender. Through this, the researchers observed that the foster children scored significantly higher on their average IQ score.

Harden next discusses how the actual mechanism has only been hypothesized but never fully determined. Some hypotheses have attributed it to the greater stimulation provided by a family, others to improved nutrition, amid other possibilities. In addition to showing the importance of the environment for cognitive development, this is an example of thin causality. While the specific mechanism was never established, it can be reliably said that one variable made a difference in changing the other.

Harden is clear that this is the only kind of causality that can be established between genes and variables like educational attainment. This is done through something called a Genome-Wide Association Study (GWAS) in which the variable of interest is measured, in this case through years of formal schooling, and that is correlated with a polygenic index. This last part is a number that is assigned to a person’s genome based on the specific configuration of certain genes of interest.

The author’s argument is that GWAS and polygenic indices allow us to establish thin causality between these genetic configurations and, say, cognitive ability or educational attainment. Here it is important to clarify that other researchers in the field disagree. Eric Turkheimer, Harden’s former doctoral advisor, maintains it is all just correlations, as he is quoted in Harden’s profile in The New Yorker.

So, even if we put aside the possibility that there are racial differences due to genetics, there are, in some sense, three possible worlds. The first one is Turkheimer’s, in which we can observe certain correlations between genes and educational attainment, among other variables, but there is no causality. Then, there is Harden’s world, in which there is thin causality. The effects of genetics are small but noticeable, and environmental interventions have the capacity to level the playing field. Lastly, there is the world of authors like Charles Murray and Robert Plomin in which not only is there causality, but it is fully deterministic, and trying to level the playing field is essentially futile.

The third possible world seems to be essentially impossible, according to what Harden herself argues in the book. Yet, here again it seems that the mere suggestion of causation is enough to make Harden’s and Murray’s world one and the same. Another review published in the Los Angeles Review of Books opens by stating that Harden’s thesis “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.” This, again, is a mistake because just the difference in determinism should be enough to make it clear that the two arguments are almost opposite because neither of the two theses work if the way they approach determinism is changed.

The problem with this kind of lumping is two-fold. First, what I have called “Harden’s world” is one which ought to be seriously investigated. But just as importantly, there are, in my view, perfectly valid criticisms of Harden’s descriptive argument that do not require equating it with the race science of Murray or Plomin.

What I see as Harden’s main problem comes down to her definition of causality. As stated previously, she is clear that she is referring only to “thin” causality. Under that definition, the argument holds. However, the issue is whether her definition of thin causality is something that raises to the level of causality, as something we would understand in plain language. The evolutionary biologist Richard Lewontin disagreed, for example, as Harden acknowledges:

“The evolutionary biologist Richard Lewontin, who was a vociferous critic of behavioral genetic studies of human behavior, alleged that scientific results that have a ‘historical (i.e., spatiotemporal) limitation’ and that do not give information about ‘functional relations’ (i.e., mechanisms) are ‘no use at all.’”

In essence, for Lewontin causality is either thick, or it is useless. I do not think this is necessarily the case. The question, however, remains of what exactly can we remove from the concept while still being correct in calling it “causality.” An instructive case for this comes the German physicist, mathematician, and philosopher of science Grete Hermann. Her work revolved around the emerging (at the time) field of quantum mechanics (QM), which seemed to present a challenge to the very possibility of causality with its apparent randomness.

Hermann believed otherwise. As she concludes in her 1935 paper “The Foundations of Quantum Mechanics in the Philosophy of Nature”:

“The theory of quantum mechanics forces us to resolve this mixture of different principles in the philosophy of nature, to drop the assumption of the absolute character of knowledge about nature, and to deal with the principle of causality independently of this assumption. Quantum mechanics has therefore not contradicted the law of causality at all, but has clarified it and has removed from it other principles which are not necessarily connected to it”

Her main argument is that, before the emergence of QM, the concept of causality necessarily involved the possibility of predicting the outcome of an interaction through the knowledge of the mechanism involved in it. QM makes prediction impossible, due to the multitude of possible outcomes. Yet, as she explains, QM remains a causal theory because it is always possible to observe the outcome and then retroactively infer the cause and the mechanism by which it happened.

This is interesting because, in some sense, it is like a competing definition of “thin” causality, as it could still be argued that the pre-quantum concept remains the “thick” version. But note that, what Hermann keeps and removes from the concept are almost opposite of Harden’s version. Harden’s thin causality rests on the predictive power and the regularity of the cause-outcome combination restricted to its context, but she does away with the generalizability (or portability) and the knowledge of the mechanism. Hermann, instead, does away with regularity and predictability, given outcomes have a degree of randomness. However, she keeps generalizability and, importantly, knowledge about the mechanism (i.e. particle interactions through the electromagnetic force).

I see Hermann’s version as more akin to the plain language understanding of causality and, thus, a more appropriate “thin” definition. The problem is that Harden’s thin causality—because it lacks specific functional relations—can always be reduced to non-causality in the relation we really want to explain. To illustrate, consider the following clearly false (but hopefully useful) examples.

We design a study to measure whether genes cause differences in educational attainment. To do so, we measure the polygenic scores of a large sample and compare it to their formal years of schooling. We observe a clear relation of higher polygenic scores to additional years of formal schooling. A few years later, however, it is observed that the mechanism is that the additive effects of the genes included in the polygenic scores increase the production of a pheromone that makes those people much more likeable, and that is what explains their success.

In this case, we can technically still say that genes increase educational attainment, but it certainly would not be through the kind of mechanism that one expects when reading that genes improve educational attainment. In fact, people with those particular genes could even have lower cognitive skills and still have higher educational attainment.

Now, Harden does explain that, while specific mechanisms are still unknown, the genes included in the study do act in specific regions of the brain which make them interesting. So consider instead the second example.

The setup and results are the same as in the first one, but this time it is discovered that the mechanism is the following: The people on the lower end of polygenic scores have genetic configurations that cause their brain to be worse at visually making out shapes such as letters and numbers. This makes it so that taking exams is considerably more difficult, so they complete less years of formal schooling. This is still consistent with Harden’s version of soft causality of genes over educational attainment—and this time even with the requirement of genes acting in the brain. Once again, though, it does not have anything to do with cognitive skills, hard work, resilience, or any other attributes that we would intuitively associate with educational attainment.

This is simultaneously a problem for Harden, but not so much, depending on one’s perspective. If one is looking for descriptive precision, then I argue it is an issue for the reasons just described. But from a policy standpoint, it might not be. Note that even the pheromone hypothetical example, which was clearly far-fetched, would still have implications for social equality, and since it does not even involve anything that could be called a “talent,” it would mean a lot more people would readily accept it as an unfair advantage.

Lastly, it should be clear that nothing said so far should be understood as a dismissal or a reason to ignore the very real history of the misuses of ideas like inheritance in order to justify inhumane policies like forced sterilization, segregation, or any other. Harden is clear about this and, in fact, stresses the need to involve more diverse people in genetic research which, until now, has been almost exclusively done on people of European ancestry, which constrains the kinds of conclusion one can draw from it.

In fact, it is interesting that her call for the inclusion of diverse and culturally competent researchers in genetics could have been replaced with a call to decolonize behavioral genetics, and the meaning would have stayed the same. This, in turn, could have been framed as part of the broader movement to “decolonize science.”

In short, while I do think there are legitimate criticisms of Harden’s work, especially with regards to her use of the concept of causality, the ethical questions she raises and the answers she provides are important and useful. I think she does provide enough on the descriptive side of things that we should consider her arguments seriously. And then, even if it turns out, when our understanding of genes advances, that the causal mechanisms are not what she suggests, we still have reasons to consider her ethical arguments. All things considered, The Genetic Lottery is a welcome addition to the public discourse.

Néstor de Buen holds an M.A. in social sciences from The University of Chicago. He has previously written at Quillette.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.