This article about the polygenic score was produced in partnership with The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education. This is part 3 of the series “Gifted Education’s Race Problem.”
Many factors boost a child's chance of success in school — like having wealthy parents who can afford tutors. But recent research has raised another possibility — one that is discomforting to many — the idea that scientists might someday be able to spot the genetic markers associated with academic performance.
To do this, researchers are turning to a relatively new genetic approach called the polygenic score, which assesses a person’s likelihood for a specific future based on a combination of genetic variables. It’s a research technique that some scientists are using to assess obesity or cancer risk, for instance. Now, researchers are exploring this approach in non-medical contexts, like academic or athletic success.
The scientists studying genetic markers in education are trying to untangle how nature and nurture together explain school performance. In principle, genetic screening might enable teachers to tailor their approach to groups of students. Educators might then more effectively instruct kids together in one classroom, rather than separating students into accelerated and low-level courses, which can deprive Black and brown children and children from low-income families of academic opportunities.
But some researchers fear this gene screening work could be misapplied and used to further racist or eugenic thinking, even though race is a social, not a genetic, classification. There’s an ugly history of proponents of eugenics, who believe in reshaping humanity by breeding “superior traits” and removing “inferior traits,” justifying their thinking with genetics. And there are debunked racist theories that have endeavored to falsely connect race and intelligence.
For now, the science is almost entirely based on data collected from people with European ancestry, which limits the conclusions that can be drawn from it, so researchers feel that they’ve at least temporarily sidestepped the issue.
But that doesn’t mean they aren’t worried about it — and about the other ways this research could exacerbate inequities in education. Screening is expensive, for instance, increasing the odds that privileged students will qualify for extra enrichment or support before their less privileged peers.
Indeed, the idea of predicting students’ academic performance based on their genes comes with such a raft of ethical questions and unknowns that scientists in the field are urging caution. “Polygenic scores are a potentially useful new scientific tool. At the same time, there are clear reasons to be concerned,” Stanford University social scientist Ben Domingue said. “We’re going to have the capacity, with a vial of spit, to be able to predict all these different things.”
Scientists and ethicists are also concerned about commercializing this work while the research is still evolving. Already, several companies sell reports to consumers that incorporate polygenic scores for health or various physical characteristics — despite the fact that the scores are not perfect forecasters of a person’s future.
Researchers in the field want to see more critical discussion of how their work could be applied in educational settings. “If we don’t pay attention now, systems will be created, constructed around us, responding to our genetic difference,” said Sophie von Stumm, a psychologist at the University of York, in the United Kingdom, who studies genetics and education. “It’s high time to have this discussion. Honestly, we’re late to the party.”
The polygenic score that could help predict academic performance aims to assess genetic markers related to educational attainment. In other words, it combines hundreds of common genetic variants that are linked to the number of years a person stays in school. In 2016, this score could explain about 5 percent of the variation in the level of education completed.
In 2018, researchers studied data from more than a million people across countries and found they could strengthen the polygenic score to explain 11 percent of the variation in educational attainment. That value puts the score on par with factors like a mother’s level of education attainment, which explains 15 percent of variation, and household income, which explains about 7 percent.
“There are genes that affect educational attainment — that is for certain now,” said Aysu Okbay, an economist at Vrije Universiteit in the Netherlands who contributed to the 2016 and 2018 studies.
The score’s ability to explain variation in years of schooling could improve with more data. Rough estimates indicate about 80 percent of the variation in educational attainment comes from environmental factors — the rest is genetic. With enough data, some scientists believe, the polygenic score could get close to explaining 20 percent of the difference in people’s level of education.
If so, the score would be an incredibly powerful single factor for making predictions about an individual’s academic future — even though the combined environmental variables still eclipse the role of genes. “It’s really not a puny predictor at this point,” Domingue said.
In February, Domingue and his colleagues found that the polygenic score could help identify which groups of high schoolers had been placed into advanced math classes. The score could also point to students most likely to stick with advanced math courses across all four years of high school.
But polygenic scores also come laced with caveats. So many, in fact, that Okbay and her colleagues published a massive list of public FAQs — including how the study was designed and whether the research could lead to stigmatization of people with certain genes — to help readers interpret their research.
For more of NBC News' in-depth reporting, download the NBC News app
Paige Harden, a clinical psychologist at the University of Texas at Austin — and a co-author on the math study — likens the polygenic score to a credit score. Neither the polygenic nor the credit score can really forecast what will happen to a particular person. Instead, they provide a rough sense of how people with that score will, on average, fare. The score is better at gauging a group’s overall performance than an individual’s performance.
Harden and others acknowledge that it’s still a mystery how the genetic variants behind the score contribute to how far a person gets in school. “We don’t know what the mechanisms are,” Okbay said. “We don’t know whether it’s causal or not.”
Some research suggests the genes associated with education are related to the nervous system and the brain, raising the possibility that they’re connected to cognitive functions — things like strong memory, creativity and perseverance — that serve people well in school.
But the relationship could be nuanced. Domingue pointed out that there could be genetic factors that make a person more likely to be a supportive parent, which, in turn, would correlate to better school performance in their children. Because the child and parent share DNA, the polygenic score could capture gene variants in the child that explain educational performance but actually reflect the parent’s behavior.
There is also an enormous shortcoming in the datasets used for this research: Virtually all are built with DNA from people of European ancestry. Although there are biobanks in the works in Asia and Africa that could address this omission, for the time being, the scores are essentially only applicable to people of European descent. “You’re basically developing a tool that’s only useful for one segment of the population,” Harden said.
Given all of these limitations, most scientists believe it would be unlikely, and inappropriate, for educators to use polygenic scores to determine student placement in specific classes or schools. “Will someone be mad enough to track or stream on the basis of genetic predispositions?” von Stumm said. “Fortunately, I think we’re far from that.”
There could be other ways of using this genetic information. Once genetic variants are better understood and enough data is in hand, for example, it might be possible to identify children with a predisposition to learning disabilities and intervene early. In May, von Stumm and her colleagues published a paper exploring whether a toddler’s polygenic score for educational attainment could identify children at risk for language or literacy delays later in life. The conclusion: We’re not there yet.
Critics caution that there is too much to establish ethically and scientifically before we confront those scenarios. “Someday we’ll understand the genetic contribution to educational success or to life success but it will be our grandchildren who understand it. It won’t be us,” bioethicist Arthur Caplan at NYU Langone Health said.
And even if we understood this information, it’s not clear how to best use the scores in schools. Last year, Stanford’s Domingue and two colleagues wrote about a hypothetical case study: What happens when a parent tries to use genetic data, like a polygenic score, to make the case that their child deserves additional classroom support?
“I don’t know that I have good answers to that,” he said. But the scenario hints at another serious concern: inequality. Not everyone will be able to afford genetic screening, even when there are meaningful scores for people across ancestries.
Still, researchers are already using the polygenic score to explore long-standing conundrums like why children with very similar advantages follow different trajectories in life.
“We are all subject to a big genetic lottery that corresponds to an environmental lottery,” von Stumm said. She added that research into the links between genetics and academic attainment could push people to examine “fairness” in meritocratic societies, given that some people may carry genetic strengths that give them a slight but significant academic advantage, that, in turn, improves other aspects of their lives.
Measuring a person’s genetic advantage (or disadvantage) also allows scientists to control for it in their studies. That is, they can better study factors that society can change, such as spending on special programs, compulsory education and school interventions, without having their results biased by a sample of students who are genetically advantaged or disadvantaged.
And researchers can use the polygenic score to assess whether a school has failed students with high potential — or if an intervention helped retain children who were otherwise likely to drop out. In the math paper published in February, Domingue, Harden, and their colleagues found that some schools better supported high school students with low polygenic scores than others, ensuring those kids stayed in school.
Harden hopes to see the science applied in ways that emphasize social justice and provide resources to programs that need them: “That’s how I think we should be using the polygenic scores — if we use them at all.”
Sign up for The Hechinger Report’s newsletter.