Beware of attempts to undermine Feinstein’s graph

Agreement over education policy may be rare, but the need to address persistent gaps in attainment, and vastly differing life chances, between children from different backgrounds has attracted significant all party attention. The remedies may be disputed but no one would now seriously question the need to eradicate what is undeniably a scar on British society. Or would they?

One piece of academic research has been particularly emblematic of the ‘gap’ problem. Written in 2003 by Leon Feinstein, then a researcher at the LSE, it crunched the results of the 1970 Birth Cohort Study, which tracked the development of British children at 22 months, 42 months, 5  and 10 years.

Feinstein’s conclusions were stark.  Children’s test scores at 22 months could predict, though not determine, educational qualifications at 26 and were related to family background. The offspring of educated or wealthy parents who scored poorly in the early tests had a tendency to catch up, whereas the low achieving children of worse-off parents were unlikely to. Early high achievers from disadvantaged backgrounds were gradually overtaken by early poor achievers from advantaged families.

Over the years Feinstein’s graph, illustrating the last point, has been heavily used by the left and right. It featured in David Willetts’ controversial speech repudiating the idea that selective education was socially progressive, because gaps in attainment by ten meant any 11 plus test would always be skewed towards the better off.

It is quoted frequently by Michael Gove, most recently, and crudely, to MPs in defence of the hasty passage of the 2010 Academies Act.  Inequality, he explained, was so entrenched that “rich, thick kids” achieve more than their “poor, clever” peers even before they start school.

Then earlier this year, the graph took pride of place in Nick Clegg’s Social Mobility Strategy and a lot of policy initiatives now appear to ride on it;  investment in early years; the Tory conversion to comprehensive education; proposals to re-balance access to top universities using contextualised data and re-pointing  internships away from the haves to the have nots.

Maybe not surprising then that a recent, rival piece of research, taking issue with the graph’s findings, has, according to reports in this paper, found a hearing amongst restive Tory MPs,  The alternative argument, made by Anna Vignoles and John Jerrim, suggests that Feinstein’s work is flawed, because over time, and with repeated testing, the children’s scores would have regressed to the mean, implying the bright poorer children were probably misclassified initially, thus didn’t lose ground as they got older.

Feinstein, on loan to the Treasury, has not yet responded publicly but in private is sticking robustly by his original paper, insisting that the statistical basis is sound ,that his fellow academics over-estimate the extent to which repeated testing would level out results and that their work makes spurious assumptions about ability, leading to a dangerous  conclusion that ability gaps are  “innate”, which may account for its appeal to the Tory right.

While it may not be wise to categorise any child as either “bright” or “thick  on the basis of a few tests, the longer term outcomes of the children in the original study appear to back up Feinstein’s deeply held view that the potential of the least privileged in society is routinely squandered. Anyone familiar with his later work will know that he, and others, have also probed more deeply why it is that some children fall back in relation to what they are capable of achieving.

Schools do have a part to play in breeding academic success, vital personal skills and attributes. But money, parenting, relationships, culture and community also matter. Bright children can easily lose competitive advantage if they are battling with poverty and chaotic home lives in poor neighbourhoods. Others can flourish; develop ability and confidence if they have aspirant supportive parents, good schools, cultural capital and (often) private tuition.

None of this is rocket science to those working with children from a range of backgrounds.  The real question is why there might now be a political class searching for evidence to undermine such a powerful and widely accepted basis for more progressive policies.

My own theory is that we are moving into an era in which what David Willetts once described as the ‘parental arms race’ will be increasingly fierce and give rise to a mean, rancorous streak. The better off already fight like tigers to protect what they perceive to be their children’s right to a place at the top. What better way to up the ante than by sneakily reintroducing the idea of  “innate “intelligence, subtly linked to family background, restoring the idea of a natural order which social policy cannot, and should not, interrupt.

First published in Guardian Education