I have written about educational inequality many times over the years. That is hardly surprising. Differences between students remain one of the most-discussed topics in education policy, and I will gladly admit that it is also close to my heart. At the same time, it is one of the most difficult issues to talk about. As soon as we ask why some students are more successful at school than others, we quickly find ourselves discussing poverty, parenting, motivation, culture, school quality, intelligence or, for some people perhaps most controversially, genetics. We see these explanations are often as competing alternatives. But the reality is usually far more complicated. Two recent publications illustrate this rather well.

A recent report from the Education Policy Institute (EPI) and the Education Endowment Foundation (EEF) provided yet another sobering set of findings. The researchers tracked the development of the disadvantage gap in England and concluded that almost half of the attainment gap observed among disadvantaged students at age sixteen was already present by the end of primary school. In other words, a large part of educational inequality does not emerge during adolescence but much earlier. This is not a new finding. It aligns with work by James Heckman and many others. Nevertheless, it remains rather depressing news.

Education expert Dylan Wiliam responded to the report on social media with an interesting observation. He pointed out that the model did not include genetic factors anywhere. That is, in itself, a perfectly reasonable scientific observation. When researchers try to explain why students diverge in their educational outcomes, genetics is certainly one possible factor. If it is not included, part of the story may remain hidden from view.

At the same time, genetics is not exactly a factor that schools or policymakers can directly influence, whether at the school level or the national level. But there may be more to the story.

While thinking about Wiliam’s comment, I came across a very different study that appeared almost simultaneously in Science. The study does not contradict either Wiliam or the EPI/EEF report. Instead, it adds another layer to the discussion.

Scott Marek and colleagues examined the relationship between brain development and no fewer than 649 different variables in thousands of children. The researchers set out to identify which factors most strongly correlated to differences in brain structure and function. They analysed 649 variables across thousands of children and compared their associations with brain organisation. Socioeconomic factors emerged as the clear frontrunner.

In fact, socioeconomic circumstances produced the strongest and most replicable brain-wide associations in the entire dataset. The single strongest association was not linked to intelligence, personality, or mental health, but to the socioeconomic opportunities available in the neighbourhood where a child grows up.

The findings became even more interesting when the researchers looked at where these associations appeared in the brain. If socioeconomic differences primarily affected cognitive development, one might expect the strongest patterns to emerge in regions associated with higher-order thinking and reasoning. Instead, the researchers found something quite different.

Instead, the observed patterns more closely resembled patterns associated in other research with sleep, stress, and arousal. They showed notable overlap with patterns linked to sleep deprivation and with neurobiological systems involved in alertness and stress regulation.

This does not mean the researchers have proven that sleep deprivation or stress can cause educational inequality. Science does not work that way. What they do show is that socioeconomic circumstances may influence child development through mechanisms such as these. Growing up in an environment with more stressors, less stability, greater uncertainty, or fewer opportunities for healthy sleep could help explain part of the differences observed. This is also consistent with a considerable body of research that I have discussed previously on this blog.

This is where the EPI report and the new Science study intersect. Dylan Wiliam is right to point out that genetics is not explicitly included in the EPI model. Yet Marek and colleagues remind us that genetics is not the only factor that may be missing from the picture. Their findings suggest that stress, sleep, and the broader environment in which children grow up may also shape educational outcomes in important ways.

Together, these studies reveal just how complex educational inequality really is. They show that schools do not operate in a vacuum and cannot solve every problem on their own. Yet that is not an argument for lowering our expectations of education. If inequalities emerge early and are reinforced by a wide range of influences, then effective schooling becomes even more important. It remains one of the few tools society can deliberately use to counteract at least part of those disadvantages.