Bruce Charlton has noted that, based on many well-conducted studies spanning a long period of time, IQ varies with social class, both concurrently and prospectively ("Elite institutions' class bias simply reflects 'meritocracy'", 22 May). This is correct. He has reminded us that, in a meritocratic society, selective universities tend to seek brighter, more intelligent students who will make good use of their educations, and the more selective the university, the brighter they expect their students to be. This is also correct.

He has demonstrated that, given this situation, it is inevitable that the higher social classes will be disproportionately represented among university student populations and especially so at the most selective universities. This, too, is in essence correct.

As Charlton points out, the specifics of his analysis are based on several assumptions that could be challenged and adjusted, but the basic message would remain: when admissions are based even partly on IQ, there is no a priori reason to suspect that the observed social-class ratios among university populations are the result of overt discrimination or corruption. Even this is correct.

Charlton has skipped several steps. For example, most of the studies of IQ and social class have sampled adults whose social class is based on the occupations they actually perform. At the time of admission, university students generally have no occupations in their own right. Their social classes are based on the occupations of their parents, so we have to infer that there is some association between parent and offspring IQ if we are to buy Charlton's conclusion.

É«ºÐÖ±²¥

Moreover, universities generally do not have access to information about applicants' intelligence or even, for that matter, about their IQs. They rely on A-level exam grades, so we have to accept that A-level marks are good proxies for IQ scores, which in turn are good proxies for intelligence.

But there are also many well-conducted studies that substantiate the steps that Charlton skipped. IQ tests are the most valid and reliable psychological measures that have been developed, as reflected by both their predictive validity and their associations with biological measures of brain-system integrity. This does not mean that they reflect intelligence completely accurately, but it does mean that we know they get at something about it. And A-level grades correlate well with IQ scores.

É«ºÐÖ±²¥

Furthermore, the data are clear that IQ is transmitted from parents to offspring, both genetically and environmentally. We cannot specify exactly how, either genetically or environmentally, but we can be as sure as anything in social science that it is transmitted. There really is no point in arguing about anything Charlton has to say.

That does not mean, however, that we must accept the current state of affairs as inevitable. Having universities give priority to students they regard as brighter is a policy choice. Organising society so that income and social status vary widely with education is another policy choice. We can make different choices than the choices we have made if we decide we want to. They will have social consequences, but we can do a lot to predict what those social consequences will be so that we can evaluate whether we want them.

Other countries have made different choices in the past and are making different choices now. Belgium offers open university admission: in essence, anyone can attend. The introductory classes are swamped and many, many students who are not ready wash out. Sweden has very meritocratic university admission practices, but social class disparities are much smaller and less based on levels of education so the social impact of university meritocracy is not as great. The US has sports scholarships, legacy admissions, arguments about minority quotas, students who graduate facing mountains of debt and parents who sacrifice their retirement savings for their kids' educations.

There will always be individual differences in intelligence, status and material wealth, but we can choose how closely we want status and wealth to be tied to intelligence, we can choose the extent to which we want disparities in wealth across society and we can choose how we distribute opportunities such as education. The choices we make will accentuate or minimise the social impact of individual differences because the processes involved are developmental. There is much to be discussed about these kinds of policy choices, and this is the level on which the debate should be held.

É«ºÐÖ±²¥