Just when Scotland’s politicians – and its psephologists – thought it was safe to lie in the sun and forget about the referendum for a while, up popped the Sunday Times last weekend with its Panelbase July poll. So apologies that it has taken us a little while to catch up with what was the first regular public poll of referendum voting intentions to be published for no less than two months.
Still, perhaps the politicians might as well carry on lying in the sun for a while longer. For if Panelbase are to be believed, Scotland continues to remain utterly unmoved by their efforts to persuade them to vote one way or the other.
Panelbase’s headline figures were: Yes, 37%, No 46%. However, as I have argued previously, to see what eventual outcome the polls are really pointing to it is best to take out the don’t knows and won’t says. Once we do that the Yes share in the poll is 44%. That figure is down just a point on Panelbase’s previous poll in May, exactly the same as in March and perfectly in line with the average of all of Panelbase’s polls of referendum voting intention taken since the beginning of 2012!
Political campaigns crave momentum, the sense that public opinion is shifting in their direction. Yet the truth is that despite attempts by both sides to claim they have it, so far there is no independent evidence that either of them has managed to advance their cause one jot.
Even so Panelbase’s latest poll still leaves us with a major uncertainty. For as I pointed out in my previous posting, Panelbase’s polls have so far been the odd one out psephologically in this campaign, proving to be much more optimistic for the Yes side than those of any other company. While Panelbase consistently put the Yes vote at 44%, the figure in other companies’ polls is just 37%.
If Panelbase are right then although changing minds might be proving difficult, the Yes side are at least within sight of their goal. But if everyone else is right the Yes camp would appear to have a very formidable challenge on their hands. Unraveling why this gap exists is now clearly central to understanding the prospects for the referendum and thus will shortly be the subject of another posting.
In the meantime we should note that Panelbase made a small change to their methodology in this poll. Hitherto, Panelbase’s estimates of referendum voting intention have been based on the responses of those who said their likelihood of voting in a Scottish Parliament election was at least eight out of ten. This time they asked people separately how likely they were to vote in the referendum and based their headline figure on those saying at least eight out of ten in response to that question instead.
However, the two questions got much the same response. Of the 1,001 respondents in Panelbase’s latest poll 825 gave a mark of at least eight out of ten to their chances of voting in a Holyrood election, while just forty more, 865, said the same when it came to the referendum. While the change might represent a welcome tidying up of Panelbase’s procedures it can have made little or no change to the overall distribution of their responses.
Still, intriguingly, Panelbase reported separately the voting intentions of those who indicated they were not that likely to vote in the referendum. Unsurprisingly nearly half failed to say whether they would vote Yes or No, but those who did split no less than three to one in favour of No. Previous Ipsos-MORI polls too have suggested that those less committed to turning out are more likely to vote No if they do. It is perhaps fortunate for the No side that so far the polling evidence suggests that in practice most people will make it to the polls.