by Stephen Tall on February 26, 2013
What will (I assume) be the final poll of the Eastleigh by-election was published this morning. Conducted by Populus and commissioned by Lord Ashcroft, it shows the Lib Dems sustaining a narrow lead — significant in itself, but the more so as Eastleigh voters were interviewed over the weekend during which the controversy over allegations against Lord Rennard were prominent. Figures shown are compared with the most recent Populus poll, published at the end of last week:
Mike Thornton (Lib Dem) 33% (n/c)
Maria Hutchings (Conservative) 28% (n/c)
Diane James (UKIP) 21% (n/c)
John O’Farrell (Labour) 12% (-1%)
What’s most striking is the stability in Populus’s two polls. However, the Lib Dems’ 5% lead is within the poll’s margin of error — and of course follows a Survation poll which showed the Tories ahead by 4%. To all intents and purposes, this is a dead heat contest, with the Lib Dem campaign organiser yesterday predicting that as few as 500 votes could separate the Lib Dems and Tories on Thursday.
Only one conclusion to draw from all this: it’s too tight to call! So let’s keep campaigning…
- Volunteer to help out in person;
- Make some phone calls from wherever you are;
- Donate to help finance the campaign.
Next Friday morning will be too late. There has rarely been a more important by-election for the party.
It’s up to us whether we remember it for the right reasons or not…
Finally, it’s worth highlighting Lord Ashcroft’s concluding point on the polling methodology used by Populus:
There is an intriguing methodological footnote to this series of by-election polls. The headline voting intention figures here, as with my previous poll, are based on the assumption that 30% of those who voted Lib Dem in 2010 but now say they don’t know how they will vote, or refuse to say, will vote for the party again on Thursday, compared with 50% of don’t knows and refusers who voted Labour or Conservative at the last general election. This has been a standard assumption in my polls, and is based on reliable national data. However, it is debatable whether this holds true in a Lib Dem stronghold like Eastleigh, where the party’s support is so entrenched that it ought perhaps to be treated on the same basis as that of Labour and the Conservatives. On this assumption, the Lib Dem share increases to 34%, a 6-point lead over the Tories. Even if it is true that half of guarded or supposedly undecided voters will vote for the same party as last time, a total of 27% of Eastleigh voters remain impossible to allocate – not least because nearly half of them refuse even to say how they voted in 2010. These people could yet produce a surprise.