Last week the High Court ruled that the Secretary of State was within his rights to force The Warren Comprehensive to become an academy
. This was an interesting court case, as the arguments were based around the differing analysis of the results of the Local Schools Network and the DfE.
The DfE could produce no evidence that academies perform better compared to similar schools once equivalent qualifications are stripped out. They won the case because the judge was persuaded that it was valid to use GCSE comparisons that included equivalents - even though the DfE and Michael Gove have both stated that they regard these as "devaluing" those results, and even though these.will mainly not be included from this year.
Academisation of The Warren was opposed by parents and staff, with the local consultation showing 85% against (out of 431 responses). In 2013 the GCSE benchmark figure jumped by 17%, from 39% to 56% achieving 5 GCSEs including English and Maths ("a dramatic rise" according to Ofsted) and is expected to go above the national average this year.
However the school is still in special measures and neither local opposition nor strong and improving results are, under the Academies Act, grounds to question Michael Gove's actions. To stop the conversion the school had to show that his decision was based on "a material error of fact".
In court the local authority (Barking and Dagenham) challenged the Academies Order on the basis that the claim that "the statistical evidence shows that sponsored academies are more likely to deliver attainment and improved progress" is a material error of fact. They relied on the LSN analysis
that shows that the increase in attainment is "broadly similar" in academies and non-academies, once similar schools are compared.
The DfE has never challenged this LSN analysis and did not challenge the LSN analysis in any of its evidence.
The DfE did respond with its own analysis for different periods of time. This found that "results in sponsored academies were marginally higher than in a group of similar local authority schools" (quote from High Court judgement).
Unlike what it says in its press releases, when giving evidence to the High Court the DfE was able to claim only "marginally higher" achievement for academies.
Under the Academies Act the demands on the Secretary of State are not great and a "marginal" gain would be sufficient to make his action legal. Counsel acting for the local authority challenged the DfE's research on the basis that this marginal gain was based on GCSE results that included equivalents, which Michael Gove himself described as "devaluing" the measures. The evidence is clear that academies make far greater use of equivalents.
Counsel contended that any marginal rates would evaporate if equivalents were stripped out. For instance the DfE's own document "Attainment at Key Stage 4 by pupils in Academies 2011
" found the marginally higher figure for sponsored academies (46% v 45.7% for similar non-academies) fell to 32.6% for academies v 35.6% for similar non-academies once equivalents were stripped out. Counsel pointed out that it was not possible to use more recent DfE analysis as these have not reported the statistics without equivalents.
DfE were not able to claim that academies do better than similar non-academies once equivalents are stripped out, and at no point put forward that claim
The DfE accepted that it had itself described equivalents as devaluing the GCSE measures and acknowledged that these would mainly be removed from 2014 onwards. Unable to argue that sponsored academies did better once equivalents are removed, it was left to argue that it was valid to use the figures that included equivalents as at the time of the analysis they were a "valid measure of a school's performance".
Counsel for the LA argued that whether to convert The Warren to an academy was a forward-looking question and so using figures that included equivalents was not relevant, and that it was irrational for the Secretary of State to use statistics based on equivalents that "in his view devalue the figures".
Ultimately the burden of proof under the Academies Act is very low and the judge ruled that to use one measure instead of another - even if it will no longer be valid in 2014 and beyond - does not represent "a material error of fact". Thus the local authority lost the case. However this battle of statistics was revealing.
When challenged in court, the DfE only claims a "marginally higher" improvement for academies over similar non-academies and is only able to do so with figures that include GCSE equivalents that the department itself regards as devaluing the measure. The implication is that, once equivalents are removed or once the new 2014 WOLF measure is used, there is no evidence of better performance from academies.
This is backed up by the first analysis
of the effect of using the WOLF measure, which shows that in 2013 academy results would have fallen by 7.4% - almost twice the level of the 3.8% of schools overall.
Postscript: For the full judgement, click here