Lies, damned lies and statistics

Employment Minister Chris Grayling deliberately overrode the advice of his own department's chief statistician to make claims in the Daily Telegraph that government figures showed what Grayling called "benefit tourism" and "fraud."

After Grayling published the figures alongside his own scaremongering about "fraud" and immigration, the minister was rebuked by the government's top statistician, Sir Michael Scholar, who said Grayling's use of figures was "highly vulnerable to misinterpretation."

Documents I obtained through Freedom of Information show that Grayling was also advised in advance against this attempt to "spin" the figures in the Telegraph by his own departmental statistician.

Grayling ignored this professional advice. But when Scholar criticised Grayling his boss, Iain Duncan Smith, claimed that "ministers followed professional advice" on the way they handled the figures.

The emails I obtained show this is not true.

This January Grayling persuaded civil servants at the Department for Work and Pensions (DWP) to compile a statistical report on benefit claimants born abroad. He then arranged for the Daily Telegraph to get the statistics first, alongside his own commentary.

This stood to breach strict rules on official statistics. The government code of conduct on statistics bans ministers from "issuing a political commentary on the statistics ahead of their publication."

They are not allowed to "spin" statistics before figures are released. The numbers are supposed to be issued for everyone to discuss at the same time, without ministers trying to force an interpretation.

These rules are designed to preserve the integrity of official figures.

Grayling came up with a plan to call these "ad hoc" statistics so they could avoid the rules, which only affect "official statistics."

When the plans to give the Telegraph a preview of the stats were first aired, the DWP chief statistician Tim Knight emailed his manager to say he was "not happy" with this "early briefing" which "would put a coach and four" through normal practices.

A day later, on January 18, Tim Knight agreed with David Frazer, the head of profession for statistics, and DWP communications director John Shield to write to the minister advising against the plan.

Knight sent an email to Grayling copied to the permanent secretary. In it Knight rejected the "ad hoc" statistics trick.

He told Grayling: "The controversial nature of the statistics will mean that they are likely to attract considerable media attention (and thus could be designated as official statistics even though they are being released as a one-off) and it is therefore desirable that they are released in accordance with the spirit of the code, that is, in accordance with best practice.

"This is why the head of profession for statistics and I advise you that they be released for the first time at 7am on the pre-announced date, January 20, on the DWP website, thus ensuring equal access for all, while allowing the minister to comment on them early in the day."

Knight was advising against "the proposed course of action - releasing the text of the Daily Telegraph article under embargo until midnight and then publishing the statistics on the ad-hoc website at 7am."

Knight said Grayling's Telegraph plan "may attract criticism from some elements of the media and/or from the UK Statistics Authority."

Knight also warned Grayling that the Statistics Authority "may take the view that they should have been released as official statistics and may write publicly to the minister to say so, or at least recommend that any future release on this subject should be designated as official statistics."

Knight did accept that Grayling wasn't breaking the law. He said that "in short, our advice is that the proposed release plan is lawful, but may attract criticism."

It is easy to see why Knight opposed Grayling's plan - the Telegraph published Grayling's article with the figures on Thursday January 19 at 9pm. The paper didn't bother waiting until midnight.

The figures came in an article by Grayling accompanied by a video of the minister, articles by Telegraph staff and an editorial.

In his Telegraph article and video Grayling said: "This is all about dealing with benefit tourism."

He talks about "the level of benefit tourism," "the fraud issue" and the "chaotic way of controlling foreign benefit-claimants" under the Labour government.

Under the title Labour Didn't Care Who Landed In Britain, Grayling wrote about "a natural instinct that says that no-one from other countries should receive benefits at all."

This story was picked up by other national news media. But the actual statistical report never uses the phrase "benefit tourism," because the figures do not show any benefit tourism.

They show workers who may have been in Britain for many years and who paid national insurance, but who later legally claimed benefits.

As leading economist Jonathan Portes explained the figures show that foreign-born British workers are less likely to claim benefits than British-born workers.

As Tim Knight warned, official statistics chief Sir Michael Scholar soon criticised Grayling.

Within a week Scholar wrote to Employment Secretary Iain Duncan Smith rebuking ministers for commenting on government statistics before releasing them.

Scholar said the numbers should have been prepared as official statistics because of the obvious public interest, saying: "Many users have treated them as official statistics, and have assumed that they should have been published in accordance with the code of practice, which would, amongst other things, have prevented government ministers from issuing a political commentary on the statistics ahead of their publication.

"These statistics are both highly relevant to public policy and highly vulnerable to misinterpretation," he wrote.

Duncan Smith wrote back to Scholar: "Ministers followed professional advice throughout this process regarding the status of this release from the head of profession for statistics."

However, the emails show that this was simply not true.

Grayling has been criticised over statistics before. In November 2010 Scholar told Grayling that his use of crime figures was "likely to damage public trust in official statistics."

It seems that he doesn't care.