Labour leadership hopeful Alan Johnson was embroiled in a row over spin last night amid claims that his aides tried to bury news of poor primary school test results.
A series of emails showed that civil servants had warned repeatedly that primary school figures should not be released on the same day as GCSE results, which dominate news bulletins every summer. But their concerns were overruled.
Officials were told the Education Secretary's special advisers "positively want(ed)" the two sets of results released on the same day. The primary school figures were published at exactly the same time (9.30am on Thursday August 24) as GCSE results. In previous years they had been released on separate days.
The primary school figures revealed falling standards in reading, writing and maths among seven-year-olds and missed targets for 11-year-olds.
In the emails, sent on Wednesday July 12, Tracey Beames, from the Strategic Communications Planning Unit, and Ruma Multani, a Department for Education and Skills Press officer, raised concerns.
Ms Multani said: "We can't publish the GCSE and KS1 and 2 (primary) stats on the same day.
"Last year we did the KS results on the 23rd Aug and the GCSE results on the 25th.
"Can't we do them a couple of days before the GCSE results?"
Andrew McCully, DfES director of school standards, wrote back: "Ruma, this has already been discussed with John and with special advisers. We positively want it to be 24th." Shadow education secretary David Willetts said: "Publishing all the results on the same day diminishes the scope for public scrutiny.
"The email correspondence reveals that it was special advisors, acting on behalf of ministers, and not statisticians, who determined the date of publication and pressed for a release of all three results on the same day."
A DfES spokesman said: "The Department's chief statistician has confirmed to the Statistics Commission that the correct procedures were followed. The chief executive of the commission has advised us that he is content that August 24 was indeed the earliest date that these statistical results could be released.
"The results received widespread media coverage, not least because we highlighted the data in a national Press release and ministers gave interviews to national broadcasters on the subject.
"Statisticians invited views on the date, a variety of opinions were offered, but they made the decision."