The Tory lead is more solid than you might think

The Conservatives last won a general election in 1992. That was also the year when the opinion polls met their Waterloo. The results of 50 nationwide surveys were published during that campaign. All but six showed Labour ahead, and they all suggested that the outcome of the election would be a hung Parliament, with Labour probably the largest party. They were all wrong. The largest Tory lead reported by any poll during the campaign was only a single percentage point. In the event, the Conservatives’ lead over Labour approached eight points.

To this day, no one knows why the polls came a cropper in 1992. Indeed, no one knows for sure why the polls went on to perform with only indifferent success in 1997 and 2001. To be sure, they forecast comfortable Labour victories on both occasions, but most of them — as in 1992 — exaggerated Labour’s share of the vote. Only last time, in May 2005, were the poll forecasts in line with what actually happened.

The obvious inference to be drawn is that, even now, opinion-poll findings may not be entirely reliable. Maybe the Tories are not as far ahead as they seem. There is, however, a less obvious inference to be drawn: that anyone wanting to know what is really going on should not rely too heavily on the headline voting-intention figures.

Even in 1992 there were signs that all was not as it seemed — signs that caused some of us at the Daily Telegraph to counsel against the paper’s organising its post-election coverage on the assumption that Labour, not the Tories, had won.

For example, although some (not all) of Gallup’s surveys for the Telegraph showed Labour ahead, the responses to the same organisation’s principal question on the economy told a different story.