Are soaring Coalition parties about to crash on landing at Eastleigh?

According to conventional wisdom the upcoming Eastleigh by-election is a close-run two-horse race between the partners in the Coalition Government – and two opinion polls in the last month seem to confirm this. Much is made, in the slavish corporate-media – and with many a flourish – about the fake fall-out between the Liberal Democrats and the Tories. However, in the detail of the opinion poll conducted by Lord Ashcroft, the data indicated a very large percentage of undecided voters who could enable UKIP to reel in the Establishment parties should they decide to lodge a protest vote (or even move their permanent support to the party). The performance of UKIP in last year’s Corby by-election in relation to another Lord Ashcroft poll conducted there – which also showed similar numbers of respondents who didn’t know who they would vote for, or “don’t knows” – was spectacularly well above expectations. The same could very well happen again in Eastleigh, and with dramatic effects.

Earlier this month Lord Ashcroft’s polling company conducted a poll of the voting intentions of the people of Eastleigh, and the headline figures that were published were as follows:

Conservative 34%, Labour 19%, Liberal Democrat 31%, Other parties 16% (UKIP 13%).

These figures were generated after some considerable manipulation. In the unweighted, basic responses, only 51% of respondents named a party; the rest either didn’t know for whom they would vote (29%), refused to answer (8%), or stated a preference for not voting (12%). In this particular data set, the raw poll results actually looked very different to the headline figures, and clearly demonstrated the significance of those undecided votes:

Conservative 15%, Labour 11%, Liberal Democrat 17%, UKIP 6%, Greens 1%.

As the author tirelessly tries to point out, the purpose of opinion polls is to project the status quo into the future by astonishing acts of statistical manipulation that shape the perception of the consumer of the poll and cause a belief that LibLabCon dominance is inevitable. But once again, as can be clearly seen here, the stated level of what is presented as committed support for all the Establishment parties in the headlines is very different indeed from the raw data.

Under the calm exterior of the corporate-media, which has been presenting the Eastleigh by-election as a formality, there is recognition of the raw data reality, and the prospect of the “don’t knows” turning to UKIP. In the Guardian, there was guarded allowance for the possibility of a shock result, with the key revelatory indicator words being “most bets” and “off”:

Let’s be absolutely clear, as ever, that the consensus may be wrong. In the national opinion polls that have reported this week, the Lib Dems are on 12% or less, which means they are at least 12 points down on the 24% they scored across Britain in 2010. In the same national polls, Labour is at least eight points up on 2010 and the Tories at least five points down. And one national poll has Ukip on 14%. If opinion shifts on that scale were reflected at Eastleigh, and especially if there were any sort of a late surge towards one of the challengers, most bets would be off. And a lot can happen in a week.

In the Independent, there was some more clarification. The “challengers” mentioned in the Guardian piece weren’t going to be Labour:

Voters said they saw the contest as a Tory and Lib Dem fight with the protest vote was going not to Labour but to Ukip.

And that is the real wild card in this election. Driving around Eastleigh Ukip looks to have the second largest number of campaign posters up in gardens and windows and growing support from the ’none of the above’ protest vote that used to benefit the Lib Dems in by-elections.

There is something in the idea that UKIP would benefit from people abandoning the Lib Dems. In the last week alone, Lib Dem controlled Eastleigh council bitterly disappointed many people by giving the go ahead for hundreds of new houses in rural parts of the borough (with UKIP being the only party who are opposed to building to accomodate immigration). A former Liberal Democrat mayor even defected to UKIP. Survation’s opinion poll of 10th February showed the switching of allegiance to UKIP being greater from the Lib Dems than from the Tories. Indeed, a fight between the Lib Dems and UKIP might benefit the Tories, but the reality on the ground tells a different story. In what is being called an embarrassment when the corporate-media bothers to cover it, when Boris Johnson came to town looking for that 15% who would definitely vote Tory, he couldn’t find them (see link above).

The only question that remains to be asked is what the extent could be of the move in support to UKIP? It’s hardly scientific, but the results of, and the opinion polling approaching the Corby by-election could be examined to obtain a general idea; Corby was also subject to a Lord Ashcroft poll. As this site discovered at the time, it was in the detail of the Lord Ashcroft poll for Corby where a more representative figure could be found in terms of the actual by-election result. Could the equivalent data for Eastleigh similarly point to something more akin to what could be expected in the by-election there?

The data that needs to be looked at in this context is the percentages after the “don’t knows” and other non-responses have been taken out; for Eastleigh, this gives these figures (which could be weighted with a factor representing certainty to vote):

Conservative 33%, Labour 21%, Liberal Democrat 29%, UKIP 15%.

The “don’t knows” for Eastleigh are of a similar percentage as they were at Corby (27%) where UKIP’s actual result saw an improvement of 133% upon the relevant polled figure. Applying the same to Eastleigh, UKIP could be on course for 20% of the vote, which would be a victory whatever else happened. And if the trend is repeated in Eastleigh as it was in Corby, whereby support for the other three patries falls off in the translation between this raw/semi-processed data and the headline figures, then UKIP could very well beat Labour – and Survation has predicted as much (see link above). Moreover, UKIP could move very close – within single digits – to the Liberal Democrats and the Tories, and if drop off of support for those two parties has accelerated since Corby, things could get very interesting indeed.