Gallup Says Labour By 30!

Since the beginning of the election the British Election Study Gallup poll has been asking c. 150 people each day a number of questions relating to the election, including how people intended to vote, their certainty of voting, interest in the election, issues of importance, liking and disliking of political leaders, etc. Up until Friday, I had not read about it nor had been aware of its design.

On Friday, one of my young colleagues discovered the results on the University of Essex web site, and I reported the results of the voting intention up to the 16th in my report yesterday, together with the findings reported from the Labour Party's private poll.

These two polls, until then unreported in the media, are highly significant. One, the BES Gallup poll that found 54% for Labour, 29% for the Conservatives, and 12% for the Liberal Democrats, a similar result to the MORI poll published in the Economist on Friday. The Labour Party's private poll, reallocated to equate it to the published polls, indicated a 52% Labour, 28% Tory, 14% Liberal Democrats finding.

I made the point that from the sketchy data provided on Friday (three days only of a seven/eight day rolling poll design) that at the rate that the Tories were falling and Labour was rising that the final day's poll (of only c. 150 interviews remember) would have had to be of the order of 60% for Labour and 25% for the Conservatives. I reported the Gallup findings on the Powerhouse (Channel 4) programme yesterday morning. I also mentioned it on my regular BBC 5 Live Drive Time slot, 5:30 p.m. Mondays.

And on Monday, yesterday, the updated results were published on the Essex website. Dynamite. A 30 point Labour lead over the Tories, 55% to 25%.

The design of the Gallup poll is one used by GMTV in the 1992 General Election, and proved highly 'bouncy', using a three-day rolling poll design. The Gallup poll has a more stable design, but can be forecast in the same way as we did in 1992. With the first of the seven day's fieldwork (14th) dropping out tomorrow, to be replaced by today's, 21st, it is clear that the bad news for the Conservatives will be continuing. For on the 14th, a 53% / 32% / 11% result will drop out, to be replaced by yesterday's findings. Over the week, there has been a six-point drop in the support for the Tories, with a two-point increase for Labour and three for the Liberal Democrats.

The most recent finding, 55% / 25% / 14% would project on a uniform swing to a majority of over 350 seats for Labour, with the Tory seats in the House of Commons dropping from 165 down to about 80, with Labour having over 500 Members in the House of Commons.

By no means is this the only finding of interest in the survey, which I will be reporting on over the next few days.

• A quarter of the public, 26%, say they are 'very interested' in the election; three in ten are either 'not very' or 'not at all' interested.
• Likelihood of voting is static, with a mean likelihood of either 8.0 or 8.1 on every day of the seven we have access to, and no discernible trend.
• Four people in ten said they have not yet decided for sure for whom they are going to vote.
• By more than two to one, c. 45% to c. 19%, people say their reasons for supporting the party they do is for its policies rather than their chosen party has the best leader.
• Blair's 'like'/'dislike' average is 5.7 out of 10
• Hague's 'like'/'dislike' average is 3.9 out of 10
• Kennedy's 'like'/'dislike' average is 5.1 out of 10

But when asked who would make the best leader, Blair is chosen by three times as many people as Hague, 60% to 19%, with 14% choosing Kennedy.

And 65% think Labour is most likely to win.

The British Election Study (BES) for the 2001 general election, funded by the Economic and Social Research Council (ESRC), is based at the University of Essex, and directed by David Sanders, Paul Whitely and Harold Clarke. The details of the findings can be found at www.essex.ac.uk/bes.

16 days and counting.