Home Office in the naughty corner once again

The Home Office continues to rely on statistics on drug seizures that the UK Statistics Authority believes are “highly selective” and which, in defiance of UKSA recommendations, have still not been put into the public domain.

The chair of the authority, Sir Michael Scholar, has written to ministers and officials in the Home Office reiterating his view that this is unsatisfactory. But the Home Office has so far ignored his earlier letter, reported by Full Fact on 15 November last year, in which Sir Michael expressed concern about the statistics, released a few days before official statistics emerged telling a different story.

The issue hangs on the creation by departments of what I call “statistics-lite”. (An alternative name might be “plastic statistics”, which has a pleasing rhythm.) These are figures put together quite legitimately from administrative sources, but issued in an irregular way, to a selected audience, not always made available more widely, and  – in this case – published in an apparent effort to act as a spoiler for the real statistics due to be released shortly afterwards.

In brief, the statistics-lite on this occasion showed the Border Agency to have increased seizures of drugs in the period April to September 2011 compared to the same period the year before. The National Statistics, which covered an earlier period (2010-11 versus 2009-10) showed a decline in seizures. The Home Office press release contianing the Border Agency data, which is not accessible but which is included in a letter from Sir Michael to Home Office Minister Damian Green on the UKSA website, does not say that National Statistics on the same subject are imminent, nor explain how the data it reports were compiled.

They also cover a different geographical area – the UK including offshore seizures, compared to England and Wales for the National Statistics.

Mr Green has denied the data were issued as a spoiler, but one may be entitled to entertain doubts. Both sets of data are produced from the same management information, but the National Statistics undergo data processing and quality assurance before release. So the Border Agency and the Home Office would have known that the National Statistics shortly to be published would paint a discouraging picture. To anticipate them, it published its own press release with more up-to-date but unchecked data, knowing that having written one story the media would be unlikely to write another telling a different tale so soon afterwards.

So far, so masterly. The spinners must have been proud of themselves. But the Home Office may now have gone a step too far.

In parliamentary answers to Diana Johnson, Labour MP for Kingston upon Hull North, on December 12, the Home Secretary dismissed a question about drug seizures by referring to the statistics-lite issued by her department. Theresa May said:  “Seizures have gone up in the past six months. If the hon. Lady is as concerned about drugs as she appears to be from her question, I look forward to the Opposition welcoming the drugs strategy that the Government have introduced.”

Later, the Speaker responded to a point of order raised by Ms Johnson. It was outside his powers, he said, to insist that ministers make available the statistics they rely on when answering questions. “That is the minister’s responsibility” he said. Ms Johnson wrote to Sir Michael complaining.

His reply, sent at the end of last week, acknowledges that the situation is unsatisfactory. Authority officials have been unable to find the disputed press release either on the Home Office or Border Agency websites, even though the Statistical News release issued six days later refers directly to it. The possibility for confusion “is not helpful”, he says, “and we will draw this to the attention of Home Office statistician”. His letter has been copied to Damian Green, the Permanent Secretary at the Home Office, and the National Statistician.  

There are several potential breaches in the Code of Practice, such as the failure to release statistical reports into the public domain in an orderly fashion, taking actions that might undermine confidence in the independence of statistics, and issuing statistics together with political comments on their significance.

UKSA has earlier crossed words with the Department of Work and Pensions over its statistics-lite production, and with good effect. These analyses are now made available promptly on the DWP website.

However, it’s no surprise to long-term Home Office watchers that that particular department should prove more tenacious. And, sadly, nothing much seems to have changed since the Home Office acquired a distinguished statistician, Professor Bernard Silverman, as its chief scientific adviser almost a year ago. It’s business-as-usual in Whitehall’s most recalcitrant department.