There could, of course, be other reasons for the decline. For instance, the Government has set targets for reducing KS1 class sizes. However, it has done nothing to stop the seemingly relentless upward drift in class sizes for older primary schoolchildren.
In January 1996, only 30 per cent of KS2 classes had 31 or more pupils in the classroom on the census date. By 1998, provisional figures from the Department for Education and Employment appear to show that over one-third of classrooms had 31 or more children in attendance on the census date.
For the first time in many years, there were even a tiny number of classes with 41 pupils present, being taught by a single teacher. Because only pupils present on the census day were counted, the actual number of classes with more than 30 pupils on the roll is probably higher.
Ministers, DFEE officials, and Office for Standards in Education inspectors, who surely can't have failed to notice the rise in class sizes, will point out that there is scant evidence of any direct relationship between class size and performance. When the results for individual schools and local authorities are available, it will be possible to discover whether those LEAs where class sizes have worsened are also LEAs whose test results have deteriorated the most.
Unfortunately, the recent changes to local government boundaries and the shift in status of the grant-maintained sector will make comparisons more difficult.
Nevertheless, the key question is: what will the Government do to ensure that any gains resulting from limiting class sizes at KS1 are not lost either at KS2 or in the secondary sector? The proportion of classes with more than 31 pupils has also been rising in secondary schools, albeit at a slower rate than in the primary sector.
There must surely come a point where classrooms built since the 1960s are just not big enough to take 35 or more pupils. In such situations how does the teacher decide who does mental arithmatic in the corridor?
John Howson is a fellow of Oxford Brookes University and runs an education research company. Email: email@example.com