How much will the polls themselves influence how people vote?
Is this why by-election polling can be so challenging?
When the three Old & Sad polls were coming out on Saturday evening an astute observer on UKPR wrote that we won’t really be able to compare these surveys against the actual result because the fact these three polls have been published could itself change people’s votes. That’s spot on and why all three parties were very jumpy as news of the numbers trickled out.
The Labour strategy in the days leading up to the weekend was to talk up the Conservatives and we saw one national commentator after another buy their line that the blues could be runners-up. If Labour’s data did indeed suggest that the Tories were in second place then what does it say about the veracity of their numbers?
This was, however, a sound approach for as I’ve been arguing for some time the real threat to red team comes if the “pro-coalition” vote does not split but polarises round the candidate best placed to take on Labour.
Well all three surveys have certainly done that and the question in the final days is how much of the Tory vote will switch to Watkins on Thursday and will that be enough to overhaul what are with ICM and Populus pretty hefty Labour margins?
For although newbie pollster Survation had Labour just one point ahead the established firms – ICM and Populus – each reported leads of 17 points. The red camp must feel pretty comfortable.
We’ve yet to see the detailed data from ICM but when that’s out I’ll carry out a detailed look to see if there is anything else within the surveys that might help.
At the end of the day it could all be down to turnout and the effectiveness of the red and yellow GOTV (get out the vote) operations.