Two of the country's most respected educational statisticians have savaged the Labour party's wide-ranging literacy proposals. They claim they are misleading, statistically unreliable and mark a major policy shift away from scientific study towards the Conservatives' "crude" league tables.
Harvey Goldstein, the country's only professor of educational statistics and a a former Labour party adviser, condemned the numerical basis of the document, published this week, as "complete rubbish".
His angry views are particularly notable as he works at London University's Institute of Education - alongside the document's author of and chair of Labour's Literacy Task Force, Professor Michael Barber.
The row points up a major uncertainty in Labour's policy-making machinery. For years Labour has been attacking crude statistics as inadequate. But now it has decided that raw figures might be a powerful motivating force in success.
"What they're proposing here is league tables: more of them than we have at the moment," said Professor Goldstein, a regular adviser to Jack Straw, Labour's former education spokesman. "Such league tables are inaccurate and misleading, with no notion of value-added analysis."
Proclaiming "zero tolerance" of failure, Labour's report recommends a co-ordinated national campaign of target-setting for schools, spot inspection of literacy standards, re-training primary school teachers, and a promise that all 11-year-olds would be reading at the appropriate level by the end of a second Labour term in office, 2006.
But the document has caused anger among academics who believe that years of research into what makes schools work has been ignored in the interests of publicity.
A Labour party insider this week said that the party has moved decisively away from the "value added" approach towards the "zero tolerance" of failure adopted by HM Chief Inspector Chris Woodhead: "This leaves Labour critical of crude secondary school tables, but all in favour of them for primary schools. "
This was also the approach noted by Professor David Reynolds in his Worlds Apart report, in which Taiwanese schools were shown to be successful by insisting that every child reached minimum standards.
The Labour document starts on the basis that there are "unacceptable differences" between schools, but this was attacked by Professor Goldstein as a statistical "sleight of hand". "It's just rubbish. When you analyse the data properly, most schools can't be separated. If you base a league table on crude results, they're unreliable. And not to tell people is to deny them a basic democratic right. It's wrong to pretend that league tables mean something they don't."
He said that the long-term target, of getting all 11-year-olds to national curriculum level 4 is "silly". "It assumes that you can know what a level 4 is over a long period of time. It flies in the face of all the evidence of all the research that you simply cannot maintain that standard over time. Level 4 will change.
"It couldn't have been written by anyone even half-familiar with the evidence. If you put resources into improving literacy, there's a good chance you'll get good results. I have no objections to special programmes. What I object to is an attempt to legitimise this with research evidence that doesn't exist. "
Professor Carol Fitzgibbon, of Durham University, a leading analyst of school performance, was also sceptical, saying: "I'm very much against simple- minded comparisons of similar schools. There are no similar schools, only similar pupils. Such comparisons are unfair to teachers, unfair to schools and misleading to everybody.
"I'm quite confident that value added is here to stay and will be here in 10 years' time. If the Labour party is moving away from value-added analysis, it is very, very foolish of them to do so."
A spokesman for David Blunkett said Labour still favoured value-added analysis as a useful tool: "No one has advocated it more strongly. But that is not a substitute for setting demanding targets," he said.