Poll in the Dark

After the UK election several commentators have made a good point: the pre-election polls were wrong that Labour and the Tories were neck and neck, but parties' campaigns and voters' responses were based on and compounded the error.

This was a fine example of a means of measuring something becoming confused with and thus distorting the thing being measured.

Firstly, we should ban the publication of speculative pre-election polls for an agreed period before the vote.

Secondly, people in public-, private- and independent sector organisations, and especially those which annually rip up and replace last year's "robust" and "objective" measurement tools, should regularly ask themselves whether they are making the same kinds of self-perpetuating mistakes. If they think they are, they should say so.

Many people tend not to, you see.