There is something very curious about new Labour's education policies.
As Alan Smithers, one of the country's most respected educational researchers, says in a new report that the government took responsibility for delivery and "began treating schools as the branches of a large company, setting performance targets against which they would be judged". Yet, having imposed rigid controls, Labour devised complex new ways of allowing schools more "independence" and "flexibility".
The Smithers report (Blair's Education: an International Perspective, published by the Sutton Trust) is sceptical about the effectiveness of target-setting. First, the rise in reading and maths scores for 11-year-olds merely continued a trend that started before 1997. After 2000, the rising curve flattened out.
Second, the Statistics Commission reported that the tests in English for 11-year-olds became easier between 1996 and 1999. Third, England's rise in the international league tables has not been as clear-cut as ministers suggest. The position of our year-nine pupils did not change significantly, with England still well below nations such as Singapore, Korea, Japan, Holland and Hungary. Our ten-year-olds did surge up the tables, reaching third in reading and sixth in maths but, as Smithers notes, we tend to exclude more children, on various grounds of incapacity, than other countries.
Fourth, the testing and targeting regime appears to have turned many young people off education altogether: truancy from secondary schools is up nearly 20 per cent and the number of 17-year-olds who are in neither education nor training is up almost 10 per cent. Fifth, if the recent UN Children's Fund report (An Overview of Child Well-being in Rich Countries) is to be believed, our children are now the most miserable in the developed world.
Smithers believes that schools do better if they are independent and he supports new Labour's creation of academies and trust schools under private sponsorship. Even after allowing for home background, he argues, private schools, including those dependent on public funding, are superior across the world. I find his evidence unconvincing; he casually dismisses the superior performance of government schools in Japan, Switzerland, Italy and Scandinavia as "exceptions".
It is simpler, surely, for ministers to stop micro-managing publicly funded schools than to enlist second-hand car salesmen and religious crackpots to run them instead.
Smithers has a second, sounder proposal. Labour has used a single set of tests to do different things: monitor national progress, set targets, assess schools, grade children and inform parents. The first function at least should be separated. If ministers stake their reputations on reaching targets, they will not collect and present national results dispassionately. More important, tests used to assess children and schools cannot tell us if standards are truly rising or falling. For obvious reasons, the same tests cannot be given every year and schools will narrow teaching to concentrate on the targeted areas.
Effective monitoring entails testing a representative sample of children, not all of them. To do this, Smithers proposes an independent body similar to the Bank of England's monetary committee.
I agree. Politicians face a crisis of public trust. Credible independent bodies are needed, in this and other areas, to report on the policy outcomes. They should have budgets to publicise their findings widely, through press advertisements and household circulars.
Politicians might then adopt better policies and voters recognise genuine achievement.
The Smithers proposal, however, would relieve pressure on schools only a little. I would also stop publication of league tables covering test and exam results. More detailed information than exists now - results in individual subjects, for example - should be available to prospective parents at each school. But the results should not be published centrally. Apart from truancy rates, the education department releases no other information on individual schools and it implies that what is testable overrides everything else in importance.
Then we wonder why our children are bored and unhappy.