Whenever inequality is discussed, there comes a point in the conversation when the debate invariably turns to the supposedly innate qualities of the rich. The elite are where they are, so it is argued, because they possess both the natural ability and the drive to succeed.

The grammar school debate is a good example of this. New Prime Minister Theresa May is expected to open the door to new selective schools in England in the autumn. Beyond the many deleterious effects grammar schools have on the life chances of poorer pupils (which I documented here recently), the desire to turn the clock back to the age of selection at 11 rest on the assumption that there is a cognitive elite who must be separated from the rest like apples plucked from a tree. If a disproportionate number of the rich are the product of an elite upbringing – well, that is simply because talent is distributed throughout society unevenly and is often heritable.

The former Mayor of London Boris Johnson articulated something along these lines in 2013 when he told the Centre for Policy Studies that inequality was the product of human beings who are 'far from equal in raw ability'. Johnson was not talking about grammar schools in this instance (though he does support them), but the logic of his argument can easily be marshalled to defend selective schooling: once you start attributing nearly everything to heritability the focus quickly shifts to helping only those capable of being helped. The sheep must be ruthlessly sorted from the goats at an early age lest they hold back their genetically superior peers. I'm simplifying things, but not a great deal.

The sheep must be ruthlessly sorted from the goats at an early age lest they hold back their genetically superior peers.

It's important to concede that intelligence is to some extent heritable. Experiments with identical and fraternal twins (the former share the same DNA and the latter don't) consistently show that the IQs of identical twins are far more similar in their scores than fraternal twins. As Stuart Richie, author of Intelligence: All That Matters, puts it, "The only possible reason for this genetic: after all, the only thing that differs between the two types of twins that would make them more similar – so long as each pair is raised in the same family – is the percentage of genes they share." Richie goes on to say that, on average, around 50 per cent of the reason people vary on intelligence scores is genetic.

The debate over how intelligence is formed has been reignited by Theresa May's schooling proposalsiStock

This is the reasonable end of the debate. At the controversial end is the 1994 American book The Bell Curve by the political scientists Richard Herrnstein and Charles Murray. The central thesis of the book is that intelligence is responsible for the trajectory of a person's life, and that a deficient IQ is typically the best explanation for poverty. The book proved controversial when it came out, not least because of its reiteration of discredited eugenicist arguments from the nineteenth century – that the 'cognitive elite' was being out-bred by a multitude of halfwits – but also because it claimed (spuriously) that different ethnic groups possessed varying levels of intelligence.

The strong whiff of racism discredited those sections of The Bell Curve that dwelt on the supposed innate difference in intelligence of different ethnic groups; yet the book remains influential in American conservative circles. And, from the political perspective of the largely white middle class elite, it is easy to see why: a thesis that purports to justify economic and racial privileges as neatly as The Bell Curve contains some obvious attractions for those who are doing well out of existing inequalities.

While most of us can acknowledge that intelligence plays a part in achievement, its influence is often wildly exaggerated. Environmental mechanisms exist that can turn a dull child into an average one and an average child into a successful one. The inverse is also true. Heritability interacts with a person's environment; there is evidence to suggest that underprivileged environments can hamper a child's intellectual potential regardless of that child's innate IQ. Poverty steals time and saps focus. A child who spends the greater part of the day hungry will already be at a disadvantage when compared to his or her affluent peers. As Richie points out, studies show that heritability also tends to be higher in wealthier families.

In an influential 2003 article, Eric Turkheimer, a sociology professor at the University of Virginia, noted that many of the twins who had hitherto been used in studies of IQ had come from middle-class families. As a result, Turkheimer went in search of data applicable to twins from a broader range of families. Looking at the appropriate mixed study of 50,000 American children, the professor found that among the poorest families the IQs of identical twins contained just as much variety as the IQs of fraternal twins. Put another way, poverty had an appreciable influence on what were previously thought to be innate characteristics. As Turkheimer would later tell the New York Times, "If you have a chaotic environment, kids' genetic potential doesn't have a chance to be expressed. Well-off families can provide the mental stimulation needed for genes to build the brain circuitry for intelligence."

In the face of evidence like this, proponents of grammar schooling wish to pretend that environmental factors play no role in what we know as intelligence. Yet socialisation plays a significant part in the forming of a child's IQ – most experts recognise that as much as half of the reason people vary on intelligence scores is environmental - which is why it is so important not to have a cutting off point age 11, beyond which a minority of children (disproportionately the well-off) are socialised in an educational environment of a far superior standard to their peers.

It's fine to admit that some youngsters may be cut from a different cloth; but it isn't the state's job to compound that advantage and write off the rest.