Saturday, 8 January 2011

Constituency polls being unreliable as ever?

This is a quick blog about the polls that have come out about Oldham East & Saddleworth; ICM & Populus have very similar polling levels whilst Survation have a very different level of Labour/Conservative support.
I think much of the reason for this is the combination of the scale of survey, the timing of the survey and much of the calculation.

Populus had a survey sample of 1,500 whilst ICM and Survation only had 500 (the difference between ICM and Survation here is that Survation only had about half their calls producing results). Populus and ICM also operate much faster than Survation who took 10 days to complete their survey of 500 people so their poll was less of a snapshot of opinion.  I do find the idea of a 20 point swing away from the conservatives a bit extreme considering the national outlook.

We should be aware, however, that constituency polls are notoriously unreliable, especially when it comes to the Liberal Democrat marginals; tactical voting tends to mess things up a bit for polling companies.
So taking these polls with a pinch of salt they all display something very significant... the conservative tactical vote is effectively saving the lib dems from falling into third, populous included questions on the reasons why people were voting the way they were. 10% of Lib Dem voters (who claimed to vote conservative at the election) claimed to be voting tactically ‘to keep labour out’. If this is true it would be clearly apparent that the Conservatives haven’t been fighting this seat at all; indeed it seems that the Lib Dems have been exploiting the tactical vote significantly. It would be ironic for tactical voters if they are the ones saving the Lib Dems from 3rd place.

My conclusion? These polls are inconclusive, there’s still everything to play for in Oldham ;-); however if ICM/Populus are accurate then this will be a spectacular no vote for the coalition.

No comments:

Post a Comment