Yesterday, Daniel Finkelstein took up two suggestions for educational reform. The first, from Malcolm Gladwell’s book Outliers, was that kids should spend a lot more time at school in order to improve skills in core subjects like mathematics. But also, given the book’s aim to give an account of the production of geniuses (the account is roughly, that geniuses are not produced by individual exceptionalism, but by sheer hard work and support), the suggestion was that kids spend more time at school in order that there be more geniuses. The second, from the Chief Executive of the RSA Matthew Taylor, in his blog, was that kids spend less time at school; that older kids in their final year of schooling have one day a week where they supervise themselves in independent study.

Finkelstein suggested that because these ideas contradict one another, there must be something at fault. That turns out to be ‘survivor bias’ – both Gladwell and Taylor are guilty of focussing on the cases that support their idea and ignoring the (perhaps numerous) cases where the idea has not been borne out by reality. So for every pupil that thrives on longer or shorter hours spent at school, there will perhaps be fifty that don’t. In other words, both ideas are vouchsafed by pointing to the cases of success and ignoring  the cases of failure. We resolve the contradiction then, by saying that one idea serves some people better, and the other some other people better.

Finkelstein goes on to suggest, that in light of this, we keep an open mind about what does and doesn’t work in education. I have no problem with that. But I feel that survivor bias is differently weighted with regard to Gladwell and Taylor’s ideas.

First, we need to take into account the degree of behavioural adjustment each idea requires. A school in New York with long hours might work if it draws on super apirational parents and their expectations. Or it might work culturally in China and Korea for various reasons I don’t feel competent to comment on. But it might be a disaster in many other (say) parts of America and the UK. Conversely, the suggestion that kids in their final year spend one day unsupervised requires far less behavioural adjustment. The wider culture it is being introduced to already values self-reliance, so the suggestion is not alien. But perhaps more important, it is only one day a week in the final year. If it doesn’t work for some kids, it only doesn’t work for them for one year of their schooling.

Second, the Gladwell idea presumes the idea of education is to produce really smart kids. Which of course it is. But long hours slogging in a classroom produces a particular kind of smartness – kids who are really good at maths. Are we sure that is a powerful enough reason to massively change the education system? Won’t there be cons as well as pros to such a change?

Finally, Taylor’s suggestion is in response to pressures such as demographic change and lack of resources to fund public services. Thus it has other reasons in its favour than mere educationalist dogma. Gladwell responds to pressures also: the apparent falling behind Asia of the North Atlantic world. I’m not going to comment on which pressure is more powerful, but in analysing the merits of each idea we should at least take the distinct kinds of pressure into account.