Last week, after months of cajoling from David Lammy MP, Oxford and Cambridge released data on their admissions from 2010-2015. For those of us who work in education, the stats will not be surprising but they are stark nonetheless. At least one Oxford college had recruited no black students in five years. Cambridge took more students from four Home Counties in those five years than from the North East, North West and Yorkshire combined. Over 80 per cent of Oxbridge students had parents in professional or managerial roles, compared with 30 per cent of the population.

Oxford and Cambridge have, to their credit, acknowledged the need to do more to improve access, and these statistics do, remarkably, represent an improvement on previous years. But both they, and their vociferous defenders in the media, have also fallen back on the tired argument that there simply aren’t enough young people in these various categories of disadvantage with the academic ability to get in.

It is, of course, true that there are fewer pupils from these backgrounds with top grades across the board; and we also know that much of this is to do with inequalities within the school system and in wider society. It is ridiculous, though, to suggest that three A/A*s at A level is the sole valid signifier of academic ability.

‘Let’s put academic ability in context’

Let’s compare the journey of two young people. Our first subject attended the most selective grammar school in the country; she also benefited from private tutoring in all her A-level subjects. Her parents, an academic and a hospital consultant, have dinner every evening with their children and discuss the news of the day, as well as the books they’re reading. Her parents both went to Cambridge and all her friends are applying to a Russell Group university. She received three A*s in her A levels.

Our second subject attended a comprehensive school in Middlesbrough. She had some incredible, committed teachers, but her school was seriously underfunded, and a series of poor inspection results meant that there were four different headteachers while she was at the school. Behaviour in the school was poor and her teachers struggled to give her much time in lessons, even though she was clearly a strong student. She lived with her mum but her stepdad was abusive and drank heavily, so she tried to stay with friends as often as possible. When at home she had to share a room with her two much younger sisters. She went on to an FE college to complete her A levels and received an A* and two Bs.

Which of these two has greater academic ability? Student one has higher grades but has had vastly more help in achieving them. There’s no obvious way to tell, yet only one would have much chance of getting into Oxbridge. Those who argue for contextual admissions are typically accused of favouring “social engineering” rather than letting the real cream rise to the top - as if selective schooling, private tutoring and parental input weren’t powerful tools of social engineering themselves. Parents will, reasonably enough, always try to secure advantage for their children, so it’s equally reasonable for universities to counteract that advantage when making decisions.

Most universities, including Oxbridge, do make some use of contextual data when making decisions, but this tends to be at the margins, where candidates otherwise have very similar academic records. If we’re serious about solving this problem, far more emphasis needs to be given to context. King’s College London and Lady Margaret Hall in Oxford have taken this approach furthest by offering degrees with a foundation year for those with weaker A levels - hopefully others will follow their lead.

Of course, other parts of the system need to do their bit. Schools need to make sure they’re giving pupils the best advice on the pathways to make elite higher education viable. Government could be more robust on linking tuition fees to access. It should also develop a central widening participation fund by top-slicing universities to scale-up access schemes that have a strong evidence base.

But universities can’t avoid responsibility for the lack of diversity on their campuses. They need to get a lot more radical, and fast.

Sam Freedman is executive director of Teach First, and a former policy adviser to former education secretary Michael Gove

Teach First is running a conference on social mobility at Wembley Arena on 24 October. Free tickets are available here

Want to keep up with the latest education news and opinion? Follow Tes on Twitter and Instagram, and like Tes on Facebook.