So just to tie up some loose ends from the weekend – the Panelbase datasets have been published, and contrary to some previous suggestions it turns out that the standard independence question was asked after all. So the question wording isn’t an alibi for the slight drop in the Yes vote, but what is a partial alibi is the rounding of the results. The raw numbers show Yes on around 52.4% and No on 47.6%, so they were within a smidgeon of being rounded up to 53-47 rather than down to 52-48. In the overall scheme of things it’s not really that far away from the two Panelbase polls in the summer showing Yes on 54%, which at the time was a record high for the firm.
With Don’t Knows left in, Yes leads by five percentage points even on the rounded numbers (Yes 49%, No 44%). That contrasts with last week’s Survation poll for this blog, in which the equivalent gap was just two points (Yes 45%, No 43%). I’d suggest, then, that the Panelbase poll is somewhat better for Yes than it may have looked at first glance. It also contained a multi-option question about the various constitutional options, and incredibly, independence is more popular than the two other options (the status quo and greater devolution). Gone are the days when we knew that a multi-option referendum would kill independence off by making a win for Devo More inevitable.
