Happy new year, and welcome to the first Number Cruncher Politics briefing!
I’ve been getting quite a few requests for “something like” this, so here it is. Because of the festive period, this first offering might read a bit more like a catchup than a typical briefing, but normal service will follow.
There were four published GB voting intention polls in December – a simple average of them gives the following figures (and changes from November):
CON 40.3 (-1.1)
LAB 28.0 (-0.9)
LD 10.0 (+1.7)
UKIP 12.0 (+0.6)
GRN 3.5 (-0.4)
First up, the NCP 17 for 17, a chartstorm looking both forward and back. See how the parties have fared over the last 19 months, how Theresa May’s honeymoon and personal ratings compare historically and why 2017 should be tougher for the Tories, even before we start talking about the various policy- (and politics) related challenges.
It also contains analysis of the lost Labour votes since the Blair years and the challenge facing the centre-left right across Europe, together with a look at UKIP, the Greens and the developing #LibDemFightback.
If you were away from social media over the festive period, you may have missed a very interesting and detailed paper from UK in a Changing Europe called Brexit: Six months on, with contributions from a number of academics, many of which will be of interest to poll watchers.
But what about those polls and predictions? Following recent mishaps, I argue in a Bloomberg article that data isn’t dead, but it does need to adapt. Or alternatively (I didn’t even try to get this one past the editor) “My Christmas message to the Turkeys that want to ban opinion polls: Get stuffed”.
Some data is useless though, namely the “polls” that don’t even attempt to get a representative sample. Some local newspapers’ open access (“voodoo”) polls found apparent swings towards Remain in areas that voted heavily Leave. When the Independent covered this I explained to Charlotte England why I’m not putting any weight on them. Anthony Wells sets out the case for the prosecution in even more detail.
Mark Pack’s polling database (the source for most of my historical comparisons) now includes a number of by-election polls. There haven’t been many of these recently – in the last parliament they reasonably accurate (much more than General Election constituency polls). The one that’s been published in this parliament was for Richmond, which showed Zac Goldsmith 27 points ahead, however that was five weeks out. Some Lib Dems think they really were that far behind at the start of the campaign, others don’t.
Talking of by-elections, the Copeland Westminster seat about to be vacated by Jamie Reed will probably be contested a couple of months from now – see my initial thoughts here, with more detailed analysis to follow. And for those of you that like the Thursday night madness every single week, Harry Hayfield has a roundup of recent council by-elections here.
That’s all for now – there’ll be more to come as and when relevant content is published. Polls are often a bit slow to get going again after new year, but we’ll hopefully get at least one this week.
If you haven’t already signed up, you can do so using the form below: