Since last Thursday's local elections I've been berating some people for the their London-centric analysis of the results. After all, my adopted hometown accounted for 41 per cent of the seats contested and – it felt like – over 90 per cent of the national media coverage, despite being home to just 12 per cent of the UK's eligible voters.
But there is a technical difference between the capital and the provinces. London was actually polled.
In the final stages of the campaign, YouGov and Survation both published local election voting intention numbers, giving Labour 22- and 20-point leads respectively. With the votes counted, Labour actually beat the Conservatives by 15 or 16 points, depending on how you account for multimember wards. What should we make of that?
Well, first of all, I wouldn't jump on the "let's bash the pollsters" bandwagon. Polling local elections is notoriously difficult and rarely done. But Labour did underperform the polls, so it does raise the question of whether methodological changes post-2017 have taken things back to the pre-2015 problem of too many Labour voters.
I would be very cautious about that narrative, too. Not only do people vote differently in local and general elections, but being a bit off in London isn't the same as being a bit off everywhere. Case in point, in recent years London surveys have actually been fairly accurate – it's the rest of England and Wales that have caused problems for GB or UK-wide polls.
And secondly, while we're on the subject of London, Heidi Alexander's resignation as Lewisham East MP sets up a by-election in South East London. Her majority of 45 points over the Conservatives in 2017, with everyone else losing their deposits, screams Labour hold, but last week's local election results suggest the Lib Dems and even the Greens might have something to say about that (though note that the council result is slightly distorted by the Greens running a lone candidate in a multimember ward). So Vince Cable's troops should be looking for a solid second-placed finish, given that as governing party the Tories are likely to drop back, and that this area was about 66 per cent Remain by our estimates (Chris Hanretty has 65 per cent).
By-election in store for Lewisham East. Here’s how its wards voted on Thursday (top candidate method, %) and GE 2017 result for comparison:
And there’s even a good omen for the Lib Dems. Although not the most useful historical comparison, and although there have been other by-elections in what’s generally considered South East London, the last two Westminster by-elections in constituencies primarily within the SE postcode area were Greenwich in 1987 and Bermondsey in 1983, both of which saw a Lib Dem predecessor party come from third place to oust Labour…
Returning to the local elections, plenty of great analysis has already been written. Recommended reads include those from Steve Fisher, John Curtice and Rob Fordon the overall picture and Andrew Teale on the Greater Manchester area. Habib Butt (aka TSE of PB) assesses the projectons and Keiran Pedley has a bonus edition of Polling Matters.
From the activist blogs, Mark Wallace has a campaign debrief on Conservative Home, while Luke Akehurst goes through the results in detail on LabourList, and Conor Pope eloquently makes a point I’ve been trying to get across – or Labour’s electoral strategy to succeed, it needs to be making the historic gains in order to often losses in its traditional heartlands.
And finally, Anthony Wells has an important reminder: Don’t scoff – you’d be lost without us pollsters.