Bioethics Forum Essay
A Narrow Path for Optimism that Social Genomics Can Combat Inequality
In his recent piece, “The genes we’re dealt,” Erik Parens puts his finger on cause for concern with what he calls social genomics: while progressives can use insights from this new field to justify combating inequality, conservatives can use them to justify the existence of that same inequality. This pessimistic conclusion—which Parens argues convincingly for—follows from a focus on insights at the societal level, that of a whole population. But there are grounds for optimism by focusing instead on potential insights from social genomics derived from local-level comparisons between different environments. Such insights could point to interventions that progressives and conservatives might just be able to agree on.
The central tool of social genomics is the use of polygenic scores. These scores sum up the small effects of many genetic variants associated with a trait of interest: educational attainment, IQ, income. The techniques deployed are the same as those developed and in use for studying the genetic variation linked to disease. Issues with polygenic scores are legion, including that it is not clear how to interpret them, that they are currently massively less predictive in non-European ancestry populations, and that even within an ancestry group predictive ability is variable. Any consideration of use of polygenic scores needs to proceed with these caveats in mind.
It’s useful to distinguish between three levels of potential insight from the use of polygenic scores in social genomics.
At the individual, or personalized level, and as Parens points out, in direct analogy to personalized medicine, is personalized education. You genotype a child, calculate their polygenic scores for a range of traits deemed relevant to their education, and use this information to tailor individualized educational interventions, such as early reading support for a child whose genetics indicate they may struggle in this area. There are many good reasons to be skeptical of this idea, not least because while polygenic scores may capture some of the variance of the trait in a large study sample, for any individual the scores are (at least currently and in the near-term future) not that predictive.
At the societal level there are potential insights that feed into the old nature-nurture debate. How does the hand one is dealt in the genetic lottery make a difference to success in life? This is the level that Parens focuses on. Social genomics tells us that genetic variation can explain some of the variation in traits such as educational attainment. Conservatives can focus on this to make the point that some differences in outcome are unfortunate, or, equivalently, that not all unequal outcomes are unjust. Social genomics also tells us that much of the variation in these traits can be explained by other sources of variation, for example socioeconomic status, even when we control for genetic variation. Progressives can focus on this and make the claim that some unequal outcomes are unjust, or, equivalently, that not all unequal outcomes are solely unfortunate.
We know that progressives and conservatives are going to disagree about where the line between unjust and unfortunate unequal outcomes lies. Indeed, this can be seen as a—perhaps the—core difference between conservatives and progressives.
But this level of insight from social genomics only rules out positions at both extremes. And these are positions that no one actually holds. Most progressives will argue for some version of “society owes everyone some basic set of stuff” (and will disagree as to what exactly this bundle looks like). It is unjust for individuals to not receive this basic set of stuff; other inequalities are unfortunate. And conservatives will see some legitimate role for society to buffer against the largest inequalities of outcome, particularly when it comes to children. An example is providing a certain level of education to all.
The opportunity for progressive social genomicists is to draw attention to situations where conservatives can agree that society is failing to adequately buffer against the largest inequalities of outcome. To do this, social genomicists need to draw attention away from the societal level insights to focus instead on local-level insights that might influence how we make choices to protect against large inequalities. This is the third level of analysis and application of social genomics. Sitting between the personalized and societal levels are local, comparative claims made between groups of individuals that experience different environments.
A candidate insight comes from Kathryn Harden, whom Parens identifies as a leading progressive proponent of social genomics. In a paper from earlier this year, Harden and collaborators used the polygenic scores of individuals as a “molecular tracer” to understand the flow of students through high school math courses. They found that students with lower polygenic scores who attended socioeconomically advantaged schools dropped out of math less often than those with low scores in disadvantaged schools. They also found that students with higher polygenic scores who attended advantaged schools started upper-level math courses earlier than those with high scores in disadvantaged schools. These findings point to a next step: social genomicists can work with educators to understand which school policies and practices are responsible for these differences. These insights, enabled by local-level comparisons that highlight the different structures of local environments that shape everyone’s success, can then have broad application. Note that to act on these insights, it is not necessary to genotype individuals.
At this point the progressive skeptic can point out that we’re just going to learn what we already know we should be doing, but which we lack the political will to implement. The counter from the progressive proponent of social genomics is twofold. First, to question our confidence in which policies are going to work, pointing to the graveyard of failed attempts. And second, to demonstrate that these local, comparative insights can be used to develop political will with conservatives, and actually enact change to help those worst off.
Parens highlights a relevant similarity between conservatives, as represented by Charles Murray, and progressives, as epitomized by the disability rights movement: the acknowledgement of human diversity, and that we each need different things to flourish in our own unique ways. Parens goes on to say that recognizing this similarity doesn’t reduce the profound differences between the underlying visions. My hope is that there is in fact a lot of common ground to be built on here. The challenge becomes, how do we provide environments that allow for these diverse ways of flourishing? Again, the potential for alignment is a focus on local, comparative insights on how the structure of local environments shapes everyone’s success.
In his book Shaping Our Selves: On Technology, Flourishing, and a Habit of Thinking, Parens emphasizes the benefits of binocular thinking: by applying both an optimist’s and a pessimist’s lens, we can hope to bring an issue into sharper focus. Here is my attempt at strengthening the optimist’s lens: if social genomicists focus on comparative, local-level insights, they can increase the chance of identifying interventions that promote individual flourishing in a way amenable to both progressives and conservatives. This is a tall order. There are good reasons to be skeptical, but though the path may be narrow, it need not necessarily head in the direction the pessimist fears.
Anna C.F. Lewis, PhD, is a research associate at the E.J. Safra Center for Ethics at Harvard University.