Knife "fact sheet": The e-mail trail
A fascinating set of e-mails [283 Kb PDF] has just been published that covers roughly 24 hours around the release of that controversial knife crime "fact sheet" on the morning of 11th December last year.
The electronic exchange shows how it is possible for a special advisor in Gordon Brown's office to overrule the anxieties of senior statisticians.
I am told that senior civil servants are outraged that the Public Administration Select Committee - the all-party group investigating the affair - has chosen to put the e-mails up on its website. But to me, they provide important evidence of the power wielded by special advisors at the summit of government.
The first is a note from an advisor in Number Ten called Matt Cavanagh to another special advisor at the Department of Health, Mario Dunn. In it, Mr Cavanagh quotes a line on "knife wounding figures" that, he says "the PM would like us to publish tomorrow".
So, assuming Mr Cavanagh has checked, Gordon Brown himself is apparently keen to use this very specific figure on patients admitted to hospital with stab wounds. Downing Street knows that "these figs are provisional" and promises to make that clear in the release. (They never did.) Mr Dunn sends the request on to a Department of Health official assuming that this will be fine.
However, the health official has spoken to the statisticians involved in putting together the data. There is a "deadline" of just 45 minutes to provide clearance.
Notice how the original wording is altered. Downing Street wanted to call the category of hospital admission "knife or sharp instrument woundings" - stressing how these data are about people stabbed with a blade. Mario Dunn prefers a more accurate description: "assault by a sharp object (including knives)".
Then comes the bad news for Number Ten. Just four minutes before the 2pm deadline, the statisticians at the NHS Information Centre make it clear that their view is that the provisional figures "are NOT released". They explain how "they are potentially inaccurate and may possibly give the wrong impression".
But Downing Street are, apparently, unimpressed. In an e-mail from an official at the Department of Health, copied to Alan Johnson's special advisor, we learn that while the DH "would not seek to overrule the Information Centre on this matter... Number 10 are adamant about the need to publish this statistic".
As revealed on this blog last week, the e-mail goes on to say that Downing Street "are likely to publish the data irrespective of the concerns raised".
Forty minutes later, the chief statistician at the NHS Information Centre, Andy Sutherland writes back. He is not a happy man.
The last line of his missive may well come to haunt Gordon Brown. "This will look to observers as if the govt has cherry picked the good news and forced out publication for political ends - is this really what they want?"
Mr Sutherland then informs the Office for National Statistics.
The following morning at 8.35 - a few hours before the official release of the "fact sheet", but after some journalists have been sent it - Mike Hughes from the ONS reveals that the National Statistician, Karen Dunnell has been on the phone...
At just after 11am, Mr Hughes writes to an (unnamed) official at Number Ten making it clear that, as revealed here last week, the figures had been prepared on the understanding that "the figures aren't final and aren't for publication".
The official says he is "pursuing urgently".