If leader ratings are indeed a good guide to electoral outcomes then Clinton should do it on Tuesday
Gallup shows her with a stable and distinctive edge on favourability
As PB regulars will know I’ve long been of the view that leader ratings are as good an indicator, and quite often better, to electoral outcomes than voting polls. With these those sampled are asked what they think not what they will actually do and there’s an argument for saying that you get a more reliable response. A precursor to actually voting for someone, I’d suggest, is having a favourable view of them.
In 1992, for instance, John Major was always well ahead of Kinnock in the satisfaction ratings although the voting polls pointed to something different.
In May last year the ratings were a far better guide to Cameron’s Tories doing well compared, as we all know, with the voting polls. Ahead of the EU Referendum in June Cameron saw a dramatic decline in his ratings which is a pointer we should have given more weight to.
The question format I like best is favourability and I’ve been very taken by how that is running in the US election. Clinton figures have remained pretty solid in spite of the FBI move a week ago. The Gallup numbers above point to a very solid and consistent gap on this measure over the past month.
Gallup, the world’s first opinion pollster, is not carrying out voting polls for this election after its WH2012 performance when it overstated the Republican, Mitt Romney, and is putting all its daily polling effort into questioning like this.