Tuesday 18 March 2008

The 'purity' of statistics

Ben Brogan relates the valedictory whinge of the Statistics Commission, one of the most self-important, prissy creations of Gordon Brown's years at the Treasury. They are upset that ministers dare to sully their pure statistics with comment (published separately from the statistical releases, as it happens) as they are published. The implication is that their statistics are sacred facts, whereas politicians' interpretations are vile propaganda. In fact, government statisticians collect rather too many irrelevant facts that are of little use to anyone; a cull of such information should be a priority for the Commission's successor, along with the collection of more genuinely useful data. As I have previously blogged, some of the figures collected by statisticians have been simply ridiculous: such as counting as a 'class of 40' an assembly that happens to coincide with the annual class size census.

In the real world, unless government politicians attempt some explanation, they won't get a look in as the media put far more outrageous spin on the statistics than any minister. Of course, the statistics should appear in full, and independently in statistical releases for those that want to read them (and I am one of those who does). But the idea that a minister should not separately and simultaneously highlight the good news when everyone else will alight on the bad is as silly as trusting the Daily Mail to provide a fair and balanced presentation of the data. Perhaps the new UK Statistics Authority could busy itself ensuring that government statisticians collect more genuinely useful and reliable information rather than worrying too much about how politicians debate the results. That was the job the Statistics Commission was supposed to do.

No comments: