The polls are ludicrously everywhere and anywhere in this election. This is due in part to what I discussed a few days ago: the polling houses have not settled on a consensus fix for 2015. Again, it all seems to hang on which likely voter model you favour, and this post has an excellent overview of the two general approaches: self-reporting and socio-demographic modelling based on previous elections. YouGov in particular seem to be going well out on a limb this year. If they’re right, they’ll look brilliant (and Labour in Plymouth — Sutton & Devonport at least — will have a positive rather than negative reason for intoxicating ourselves following the declarations).
One interesting note on the YouGov model is that in the preceding week, I’ve noted while Labour’s fortunes nationally are improving with each successive daily estimate, here in Plymouth our two constituencies haven’t tracked the national trend. To wit, a week ago YouGov estimated Sutton & Devonport to be “safe Labour”, and since it’s descended into “likely Labour”. As of yesterday, the point estimate is 49% L (error band range 41-57%), 38% T (error band range 31-45%). Labour haven’t held this seat since the 2010 election, and while we like the encouragement, given we lost it by 523 votes in 2015, it will be a hell of a lot closer than an 11% Labour win. For example, the more conservative Britain Elects forecast model only gives Labour a 49.5% probability of victory in Sutton & Devonport.
Survation, like YouGov, uses a self-reported likely voter model. Yesterday, they try to explain just what the hell they think is going on here. The interesting bits are on new registrations, with the 18-24 cohort having 33% more new registrations for 2017 than at this point in the 2015 General Election. This in part maps our own experience here in Plymouth, where word is that we have over 6,000 new registrants contrasted with the EU referendum last June. What does this mean? It could be an encouraging sign, as the 18-24s are mobilised. It could also mean nothing at all.
Here in Plymouth, the City Council appears to have seriously screwed up the delivery of postal votes, with 1,510 not delivered. That said, while some on the hard left are predictably screaming conspiracy (because everything is a conspiracy, of course), the bland reality is that the snap election caught the city as off guard as everybody else. Staffing in the relevant department was experiencing significant turnover in what was supposed to be an off-year for elections in Plymouth (the Council here is elected by the so-called “thirds” method, where elections are held three out of four years with one third of the council up in any given year).
Finally, tomorrow will be a long day. The past several polling days I’ve done, I’ve clocked between 25 and 30km on foot working GOTV. Then it’s the count. That said, once the declarations have been made, speeches given, and all the other errata with these events, it will be time for this bad boy, stocked with some decent bourbon (and that will only be the beginning). I’m not going to wait until we get to the pub with a pre-arranged late license for us Labourites.