Will the upcoming polling round make us any the wiser?

In the two weeks and two days since the Crewe and Nantwich by-election we have had just two national voting intention opinion polls and the fieldwork for both took place during school holiday periods which have, in the past, led to results being skewed. So the upcoming polling round, which hopefully should start with surveys for at least one of the Sunday papers and will certainly see the monthly Populus poll for the Times on Monday night, is eagerly awaited.

For as Sean Fear noted in his final weekly column here yesterday very small changes in voting behaviour can lead to massive differences in the outcome of the next election. Just look what happens when you put the latest figures from each of the five regular national pollsters into Anthony Wells’s seat calculator.

YouGov – Conservative Majority 250 seats
ComRes – Conservative Majority 106 seats
ICM – Conservative Majority 58 seats
Populus – Conservative Majority 20 seats
Ipsos-MORI – Conservatives 8 seats short of a majority

Since the London Mayoral election at the start of May ICM has introduced refinements to its approach which probably help the Tories by one or two points. Ipsos-MORI, meanwhile, is holding a major review of its methodology and will not be publishing any voting intention data until that has been completed, Its normal May poll was carried out but we will have to wait until the finalisation of the review before we see the figures.

The Populus survey, which has operated in a similar manner to ICM, will remain unchanged as, as far as I know, will ComRes. The last two, of course, carried out C&N polls which would have been much closer to the actual result than they were but for the application of their “spiral of silence” adjustors. This is where the views of all or half of those saying they will vote but don’t know or refuse are allocated to the party they said they voted for last time.

Populus also operate this system. When future surveys from the three firms are being analysed here I will attempt to publish both the headline figures from the firms and for comparison the unadjusted numbers before this final calculation has been made.

YouGov operate online and their surveys are restricted to members of its polling panel on whom they have a vast amount of data.

In summary the big issue is whether the new polling round will provide evidence to support what seems to have happened in the by-election – that Tory supporters who had simply stopped voting in the late 90s are now returning. If they are then talk of a landslide might be appropriate.

Anybody with deep enough pockets to play the commons seat spread markets needs to follow these developments closely.

Mike Smithson