GCSE results published by the Department for Education (DfE) show enormous variations between the performance of English secondary schools. If we are serious about school improvement then it is imperative that we look at these data to identify the factors which have the biggest impact on children’s achievements, learning from the strongest schools to help the weakest. Understanding which groups of children are lagging behind must be a prerequisite to looking for systemic ways to accelerate their learning. While it might be difficult to look for fine detail, because the data are published at school rather than pupil level, we can still investigate some of the generalisations—or assumptions—that are frequently made regarding secondary school performance.

there is no statistically significant difference between academies and LA schools

The most striking observation when analysing the GCSE results, whether by raw grades, value added measures, or progress calculations, is that there is virtually no difference between academies and local authority (LA) schools. One specific group stands out: converter academies which perform well on all measures. This is unsurprising given that only schools which were initially high performing were able to convert to academy status. The group of sponsored academies shows a wider variation in performance than LA schools, but this might be expected since some of the lowest performing schools became sponsored academies. However, it is abundantly clear that there is no statistically significant difference between academies and LA schools: either the strategy of academisation as a tool for school improvement (and there could be other reasons for schools to become academies) does not work, or it needs more time before its effect will become apparent.

More important than the type of school appears to be location. There has been much discussion about a north-south divide in English education, and GCSE data show evidence of this, with schools in the north performing less well than those in the midlands, which in turn are less highly achieving than those in the south—but London schools outperform them all. Aggregating over LA regions (without ascribing any success or otherwise to the increasingly impotent LAs) to reduce between-school effects such as selection by ability, it is evident that London schools outstrip others by a significant degree.

Across all schools the data show that on average the performance of disadvantaged children is only about 70% of that of their peers, whether measured by progress or attainment—compare this with the well-reported gender differences that show boys’ achievements amounting to about 90% of the girls’—but the gap is smaller in London than elsewhere. Disadvantaged children in London schools have a 75% chance of equalling the performance of their more affluent friends, but in the north this reduces to almost 60%.

Where schools have a low score for attainment on entry, again the most marked difference is between London schools and the rest of the country, whereas high ability students can expect to make similar progress and achieve the same outcomes wherever they might go to school.

it appears to have been an embedded and sustained school improvement solution

It would be tempting to ascribe London’s success to the higher levels of funding that its schools receive, but investigation of this potential effect is difficult because the DfE does not publish financial data for academies (which account for 2/3 of schools). Costs are of course higher in London, so looking at how the money is spent in schools (proportion spent on teaching staff etc.) would be most instructive. The other substantial difference between London and the rest of the country has been the London Challenge (2003 onwards). This was never intended to be a political, single-strategy, quick-fix solution, but it appears to have been an embedded and sustained school improvement solution that has yet to be matched elsewhere. Reasons for the lack of transferability, such as those described in the Institute for Government’s review[1], relate to Tim Brighouse’s warning[2] that replication isn’t a simple recipe, urgent though it is: London Challenge activity was rich, deep and complex.

It would appear as though a decade of successful improvement in London schools ought to set the pattern for a nationwide raising of standards and reducing inequalities, or at the very least inspire some action in the “northern powerhouse” cities that currently sit towards the bottom of national league tables. More detailed pupil-level analyses recently published by the Centre Forum think tank[3] confirm the views expressed here, but at present the national strategy revolves solely around the one factor which, on its own, seems to make little difference: becoming an academy.

[1] http://www.instituteforgovernment.org.uk/sites/default/files/publications/Implementing%20the%20London%20Challenge%20-%20final_0.pdf

[2] http://londonleadershipstrategy.com/content/mitigating-consequences-london-challenge

[3] http://centreforum.org/publications/education-in-england-annual-report-2016/