So now we know for sure, thanks to the permanent secretary at the Department for Education, who really ought to order in some document folders pronto. Jonathan Slater slipped up outside No 10, accidentally revealing a briefing note, and thereby confirming that Theresa May’s government does indeed intend to open new selective schools – although this is only to be pursued “once we have worked with existing grammars to show how they can be expanded and reformed”.

At one level, this is building on David Cameron’s ambiguous stance on selection. Last October the then education secretary Nicky Morgan gave the go-ahead for a new grammar “annexe” in Kent a full 10 miles from the main school. May’s strategy, with its further pledge to open new grammars, and possibly seeking to overturn the 1998 law banning them in order to do so, is even more radical, and will face correspondingly greater resistance. “I simply can’t see any way of persuading the Lords to vote for selection on any other basis,” the note concludes.

Condemnation came swiftly from Labour, the Liberal Democrats and the National Union of Teachers; but opposition to an expansion of grammar schools has united leading figures in education and on the political left and right, including the influential thinktanks Policy Exchange and Bright Blue, and the outgoing Ofsted chief, Michael Wilshaw.

Yet a vital dimension is missing from the debate already raging within and beyond Westminster. The government is proposing a fresh crop of grammar schools as a way of boosting so-called social mobility, but May and her advisers seem to have forgotten that the government already has a social mobility strategy, produced in 2011, and it doesn’t make a single mention of grammar schools. In fact, though it identifies four “critical points for social mobility”, age 11 is not one of them.

The evidence on grammars is by now crystal clear. The educational advantage received by those selected for these schools is more than outweighed by the drag effect of the remaining secondary modern pupils, who perform disproportionately badly. Only 3% of grammar school pupils receive free school meals, and even these will gain only a marginal uplift in GCSE grades. As the Daily Telegraph’s Jeremy Warner says, grammars offer “segregated education for the middle class”. They are elitist institutions that entrench, rather than disrupt or disperse, privilege.

But the evidence is equally clear on what does improve educational outcomes: high-quality support in the early years. According to Michael Pavey, director of Labour Friends of Sure Start, “Neuroscience shows that a child’s brain is approximately 25% formed at birth, and that by the age of three it is 80% formed. It is during this crucial time that gaps open up between children from different backgrounds.”

The government should understand this very well. Independent reports commissioned by the coalition government from Frank Field (2010), Dame Clare Tickell (2011) and Graham Allen (2011) all emphasised the overwhelming importance of the early years. And as the government’s own social mobility strategy declared: “Children’s life chances are most heavily influenced by their development in the first five years of life. By the time children start school there are already wide variations in ability between children from different backgrounds.” So how does testing 10- and 11-year-olds help?

There’s been talk of introducing a small number of targeted grammar schools. But if the government is looking to test interventions, it should concentrate on the early years, targeting maternal health, school readiness, home environment and parenting skills.

Maybe this sounds familiar: yes, Sure Start children’s centres, those much-valued community-based institutions whose budgets have been halved during Theresa May’s time in government, forcing over 800 of them to close.

An independent evaluation of Sure Start in 2012 noted that “children’s centres have been found to be immensely popular with parents, and have been successful in reaching the parents who are likely to be the most disadvantaged. Beneficial effects for parents persist at least two years after their last contact with Sure Start.”

Three years later, the independent charity 4Children found that 90% of parents reported that their children’s centre had a positive impact on their child and 83% believed it had a positive impact on themselves; 80% said life would be harder for their family without their children’s centre.

The government will never improve the educational of lower-income families through the societal guillotine of the 11-plus. Grammar schools belong to the postwar world, when most young people tended to leave school at 15 and only 4% attended university. Not only do most people in education know that, so now does a substantial chunk of the Tory party.

Equally, although Labour’s pledges – to support and bolster a comprehensive system and restore the education maintenance allowance and student grants – are welcome, they still address only the symptom rather than the cause. To truly start to shift profound inequalities, we need to revolutionise support for the early years. In one respect, at least, secondary education is, well … secondary.

This piece was initially co-written with Michael Pavey, steering group member on Comprehensive Future, and first appeared in the Guardian in early September.