For several years, the idea of using genetic information about students as the basis for education policy has been gaining traction in the UK. The scientific work underpinning claims of the relevance of genetics to education policy and practice is, however, significantly contested, as colleagues and I detailed in a new paper. In this post I focus on a specific controversy over the idea that students have “genetic predispositions” to learning outcomes that the genetic sciences can uncover with advanced technologies and methods. This is an important issue as genetic science is becoming more proximate to education policy centres in the UK.

Including genetic sciences in contemporary education policy discussions and decision-making may seem unlikely given the long and ugly history of eugenics in education. However, a number of publications and developments indicate growing interest in the application of genetic science and data.

In 2019, researchers published an article titled “Viewing Education Policy through a Genetic Lens,” proposing that insights from the field of behaviour genetics could be incorporated into “interpretations of education and policy data.” Only a few years later, the lead author made a similar argument in a policy report for the UK Early Intervention Foundation, particularly in relation to special educational needs identification and intervention. The Department for Education in England even published a “futures report” detailing the implication of genomics for education in 2024, and is establishing a major policy research centre that will incorporate insights from genetics experts.

These texts are just a few examples of what has been termed a growing “geneticization of education” that has been taking shape in the US and Europe in recent years.

Polygenic scores

It’s in this context of growing advocacy for the use of genetic science and data in education that a new study about the genetics of educational outcomes in the UK was promoted this month on the website of the FFT Education Datalab, an independent source of data aimed at shaping policy debates. Authored by education and social statistics experts from the UCL Institute of Education, the study focused on how genetic data could be used to understand the biological factors involved in differential levels of school achievement. On the Education Datalab website, the authors wrote:

Discussions of educational inequality usually focus on family income, school quality and neighbourhoods. Much less attention is paid to genetics, often because the topic feels complex and uncomfortable. Yet recent research allows us to ask new questions about how children differ in their educational development and what this does – and does not – mean for schools.

As further detailed in the full preprint paper, the study focused on “socio-economically disadvantaged young people with the strongest academic potential.” The authors then explore how that “potential,” described elsewhere in the paper as “high-ability,” could be identified using “polygenic indices.”

Often referred to as polygenic scores, these are measures of the genetic basis of a social outcome or behavioural trait. Polygenic scores or indices are summary statistics that capture minute variations in thousands of single nucleotide polymorphisms (SNPs). Though no single SNP can exert much influence on an outcomes or behaviour, it is argued that the interaction of SNP variations can affect outcomes such as educational attainment or achievement in standardized tests.

Polygenic scores or indices, then, constitute the underpinning scientific methodology for claims about the genetic basis of educational outcomes, such as inequality in attainment. This is specifically what the study promoted by the FFT Education Datalab proposes.

Genetic predispositions

The study specifically highlights that polygenic indices can be used to identify a student’s “genetic predisposition” to educational attainment. The authors are very careful to highlight that genetic differences are only one influence among many others, including family resources, schooling and wider environments. Nonetheless, they refer in the full paper to a polygenic score as a “single estimate of an individual’s genetic predisposition towards the trait under study.” It is therefore clear that “genetic predisposition” to educational achievement—with an emphasis on identifying high-achieving children—is their main object of analysis.

It is this construct of “genetic predisposition” that is contentious. As the critical criminologist Callie Burt has previously detailed, polygenic scores are often discursively framed as measures of the genetic basis of, influence on, or disposition to outcomes such as educational achievement, even though genetic factors may be massively outweighed by other non-genetic factors.

Just the same month that the genetic predisposition paper appeared, too, the behavioural scientist Remy Furrer and behaviour geneticist Eric Turkheimer wrote a short, sharp commentary paper with the blunt title “Polygenic Risk Scores Are Not Genetic Predispositions.” They argue that it has become common to describe polygenic scores as indicators of an individual’s likelihood of a particular social outcome, “as though it were an intrinsic property that resides within individuals.”

Against this genetic essentialist way of describing them, Furrer and Turkheimer instead argue that a polygenic score “is not an internal predisposition waiting to be realized; it is a risk estimate derived from a model constructed from population-level DNA associations without clearly defined biological mechanisms.”

The use of the term genetic predisposition reifies a statistical construct and encourages genetic essentialist interpretations in which associations are treated as intrinsic, identity-defining properties of individuals.

In an accompanying blog post, Turkheimer put it even more bluntly: “A PGS measures nothing. It is cobbled together from bits of DNA with no regard for biology.” The point is that the language of genetic predisposition is profoundly misleading and risks distorting public understandings about the biological basis of behaviours and outcomes. A polygenic score is a statistical construct not an innate biological property. As such, a polygenic score does not provide any explanation for how biological mechanisms may influence behaviours or outcomes.

The elision of polygenic scores with innate biology is a significant risk in the application of polygenic scores in educational research, policy and practice. As discussed at a recent workshop on the ethical risks of educational genomics, much work in this field tends to represent findings in ways that overstate genetic effects on educational outcomes and downplay external factors. Whether genetics make any meaningful contribution to highly complex social outcomes like educational achievement remains highly disputed, with some scientists and critics claiming that the direct effect of polygenic differences is near zero once social and environmental factors are fully accounted for.

The real danger of focusing on innate genetic predispositions is that this leads to “genetic distractionism,” as we argued in our new paper about polygenic scores and education policy. Genetic distractionism is “the risk that attention and resources devoted to genomic research will distract policy attention from other interventions to address social problems, foregrounding biological data over other kinds of expertise.”

Genetic distractionism is a particularly risky prospect in contexts where policymaking authorities are paying increasing attention to genetic arguments about the biological bases of stratifications in student achievement.

Institutionalizing educational genomics

Indications are now appearing that genetics will be a key focus area for the new £3million Educational Neuroscience Policy Research Centre currently being established as a national source of policy-relevant scientific expertise and advice by the Department for Education in England. Led by academics at UCL Institute of Education (including one of the authors of the genetic predispositions paper), the centre will combine expertise in “educational neuroscience and  cognitive psychology, … applied classroom practice, genetics, and AI.”

The consortium members will interrogate questions relevant to learning and education today, from how cognitive and biological factors may enable us to identify children with special educational needs early on, to the impact of AI-enhanced technology in home and school learning environments.

The Educational Neuroscience Policy Research Centre therefore represents a significant institutionalization of the neuro and genetic sciences as the basis for education policy. Several members have strong associations with educational genomics research and have published widely on the genetic basis of educational outcomes.

Though the full research agenda of the centre is not yet publicly available, it seems reasonable to speculate from its staff expertise that studies of students’ “genetic predispositions” to educational attainment and achievement will feature, despite the potential of such framing to mischaracterize the contribution of genetics to such outcomes.

As educational genomics is now being brought into proximity with educational policy authorities, it will be critically important to trace how scientists communicate the findings of research and frame and construct its policy relevance and implications for practice in schools. The real risk here is of misleading the public, parents, and students themselves to think of inequalities in educational achievement as the result of innate biological predispositions, and to distract policy attention away from social explanations.