Stories + Views
Academies: The Evidence of Underperformance
The massive release of data by the Department for Education (over 200 pieces of data on each of over 5,000 secondary schools) makes possible a thorough analysis of how well different types of school have performed. The evidence is clear and overwhelming: Academies have not been the success story that their supporters have claimed. Instead there is a clear record of under-performance.
The overall figures have long been clear, for the key measure of % achieving 5 A-Cs at GCSE including English and Maths:
The data now includes a figure for the % achieving 5 A-Cs including English and Maths but without counting non-GCSE qualifications like Btecs. Here the difference is even more stark:
The gap is huge but this is an unfair comparison. We know that the raw % pass rate (though currently Ofsted’s favoured figure) is closely related to the ability of the students at entry. We know that schools in disadvantaged areas tend to achieve lower % for 5 A-Cs and we know that the early academies were more likely to be in disadvantaged areas. So does this explain the discrepancy?
The answer is a resounding no. To analyse this, I split the data into five comparison groups according to the % of students on free school meals. The first group, the most advantaged, is of schools where less than 10% are on FSM and so on up to the most disadvantaged where more than 40% are on FSM. Academies still perform worse than comparable non-academies.
The figures with GCSEs only show a bigger difference:
The message is clear. When academies are compared to comprehensives with the same level of disadvantage, their results are worse.
The same is true for % of FSM students acheiving % A-Cs (with English and Maths), with equivalents and without. The same is true for % FSM students making expected progress in English, and also this % in Maths. Indeed Academies seem to underperform virtually however you analyse it. The one exception I’ve found is Best 8 Value Added (the value added in a student’s top 8 GCSEs) where Academies rating is slightly better, though within the margin of error. However even here the difference is explained by the use of Btec and other equivalents. On the rating for value added in English and value added in Maths, Academies are slightly below average.
This government claims that Academies have such a strong proven track record that every school could convert to them. They quote schools like Mossbourne and Burlington Danes in support. However this is policy-making by anecdote not by evidence. Both those schools are outstanding but they are clearly, from the data the DfE release this week, not the norm for academies. It is likely to be other factors than their academy status that is the cause of their success. If government education policy was genuinely evidence based, then they would now be looking at converting academies to LA-supported non-academies, in the hope that this would raise their results.
Data Notes: The Academies figure, here and throughout this post, refers to the category of sponsor-led academies, of which there are 249. It does not include the ‘converter academies’, of which there were just 25 at this point. Non-academies include those classified as community, foundation, CTCs or voluntary aided schools, 2,681 in total. Special schools are not included.)