The DfE has released a table of provisional GCSE benchmark results for every secondary school in England. The good news is that previously underperforming schools continue, on average, to improve their results significantly. This year when non academies are compared to sponsored academies, the non academies have improved at a consistently faster rate.
This is an important finding. At the moment the Education and Adoption Bill is heading for the Lords, and forces “inadequate” and “coasting” schools to become sponsored academies. The government’s argument is that only by becoming sponsored academies can their improvement be guaranteed. This data, from the Department for Education, shows that the reverse is true. The evidence of the last three years shows that schools will improve at a faster rate if they stay in the maintained sector.
My initial analysis – for 2011, 2012 and 2013 results – found that sponsored academies improved no faster than maintained schools, but also no slower. This changed in 2014, when maintained schools grew at a faster rate. This seemed to be because of the removal of “GCSE equivalents” from the benchmark figure, as sponsored academies tended to make more use of these.
The 2015 benchmark figures, for the % of students achieving 5 GCSEs including English and Maths, are directly comparable to the 2014 figures. And, for the second year, there is a clear difference: Maintained schools, on average, have improved at a faster rate than similar sponsored academies.
The crucial point here is the comparison with “similar” schools. Previously underperforming schools tend to increase their results at a faster rate. If sponsored academies are compared, as government ministers like to do, to schools overall then their results will look impressive.
However this is misleading, as the schools have very different starting points. For the analysis below I have split schools into quintiles (or fifths) by their GCSE benchmark for the base year. This means there are roughly the same number of sponsored academies in each quintile. I then compare the average growth for maintained schools and sponsored academies in each quintile.
2014 to 2015: Maintained schools improve at a faster rate
For the lowest quintile, those with 33% or less achieving the GCSE benchmark in 2014, the average increase for sponsored academies was 4.7%, compared to 7.3% for maintained schools. Indeed in all five groups, maintained schools saw their results rise at a faster rate, or fall at a slower rate, than similar sponsored academies.
(Analysis based on 440 sponsored academies and 1,361 maintained schools.)
2013 to 2015: Maintained schools improved at a faster rate
Due to the change in the GCSE benchmark in 2014, most schools saw their results fall from 2013 to 2014. Again in each of the five groups, comparing similar schools, those in the maintained sector grew faster or fell slower than sponsored academies from 2013 to 2015.
(Analysis based on 358 sponsored academies and 1,362 maintained schools)
2012 to 2014: Maintained schools improved at a faster rate
Over three years, the picture is similar. For those in the lowest quintile the benchmark for maintained schools rose, on average, by 8.5% compared to just 3.6% for sponsored academies. Again in all the five categories the results of maintained schools rose by more or fell by less than those of sponsored academies.
(Analysis is based on 297 sponsored academies and 1,355 maintained schools.)
Time for Review
Sponsored academies are the preferred solution of the government for schools seen as "underperforming". However the data conclusively shows that schools will improve more if they do not become sponsored academies. The effect is remarkably consistent across all three periods and across all groupings.
This is significant as the Education and Adoption Bill is set to force hundreds, or even thousands, of schools to become sponsored academies. Analysis of Ofsted data has already indicated that this will make it more likely that “inadequate” schools will remain “inadequate” (see here and here). This data indicates that it will also slow their progress in terms of GCSE results.
This is a problem unique to sponsored academies. Converter academies do not show the same difficulties. This is possibly because the problem lies with the multi-academy trusts (or “chains”). The DfE’s own analysis has indicated that the vast majority of the chains (17 out of the 20 largest ones) underperform in terms of value added. To place more schools with the chains, as the Bill proposes, seems to be the opposite of what is needed.
The DfE data is obtainable from here. The figures are provisional and so the final figures for any individual school may vary.
The analysis takes the provisional figures for 2015 and links them (via a VLOOKUP on the unique URN number for each school) to the 2012, 2013 and 2014 results in the 2014 GCSE table, available from the same download page.