Are educational disabilities increasing?

Ofsted’s report on special needs education, published today, wins few points for statistical excellence. It is an opportunity missed.
 
It suggests that the designation of special educational needs (SEN) by teachers is the result of perverse incentives. In some areas, funding depends on the proportion of pupils so defined, encouraging teachers to classify more and more pupils as being in need to win the bounty. Since 2003, while the numbers of severely disadvantaged pupils (those with statements of special education needs) has remained fairly constant, those assessed by their teachers as having SEN but without a statement have risen from 14.0 per cent to 18.2 per cent of all pupils in England.
 
This rapid increase in incidence may reflect real and growing educational disabilities, but what are the odds it doesn’t? The more pupils classified as SEN, the better the school score on the “contextual value added” league tables that have been in force since 2006. That’s a second perverse incentive.
 
How would you measure whether either of these incentives is influencing the numbers? First, by comparing growth of SEN pupils in areas that pay schools a bounty against areas that don’t. When two American researchers did that back in 2002 in schools in the District of Columbia, they found that paying bounties roughly doubles the growth in SEN designations.
 
As for CVA, a first step is to compare the trend in SEN designations before and after it was introduced.  Pupils with statements – the more severely disabled - show a decline since 2000, but those whose classification is decided by their teachers show a different trend (Table below).
 
Incidence rose in the late 1990s (not shown) to a peak in 2001, followed by a decline. Since 2003 there has been an increase, sharper in the years after 2006 when CVA measurement was introduced.  The table is not inconsistent with CVA having had an influence, though it proves nothing.
 

 
If SEN numbers are being artificially inflated by including pupils with no real disadvantages, then one would expect to find rapidly improving educational results in this group, which is just what the Ofsted report shows. While the proportion of non-SEN  pupils getting five or more passes at A*-C grades at GCSE rose by 27 per cent between 2005 and 2009, the rise was 109 per cent in statemented pupils, and 137 per cent in SEN pupils without statements.
 
They still did much less well than pupils not designated SEN, and the improvement may simply reflect better efforts by teachers to help those with disadvantages. But a simpler explanation is that many of these children don’t have special needs at all, and are being unfairly labelled by their teachers. An investigation by the TES earlier this year found 48 secondary schools in England claiming that more than half their pupils had SEN.
 
One example cited by the TES, St Marylebone C of E School in London, had 43 per cent of its pupils categorised as having special educational needs, but nevertheless 81 per cent of the schools’ pupils got five top GCSE results at GCSE, including English and Maths. 
 
Good teaching, or poor ascertainment of SEN? Teachers may say it doesn’t matter - only the results count. But they would be among the first to criticise anything that smacks of stigmatisation. It ought to be possible to provide the teaching without first having to tell children they are in special need, when many of them almost certainly aren’t.
 
That, at least, is what Ofsted concluded, but it could have made the arguments a lot more crisply with the help of some statistics.