Some pollsters are on the pitch. They think it’s all over. Is it now? YouGov’s updated MRP model, which correctly predicted a hung parliament in 2017, has projected a Conservative majority of 28 – down from the 68 it predicted a fortnight ago.
The predicted scores on the doors see the Conservatives, who have a nine-point lead over Labour on 339 seats, up 22 on 2017; Labour on 231, down 31; the Liberal Democrats on 15, up three; and the SNP on 41, up six. It’s an altogether squeakier picture for the Conservatives than YouGov’s last round of MRP data provided, but it is broadly in line with what we would expect from every other poll: the Conservative lead is narrowing as Labour makes up ground from the Liberal Democrats. Even if every poll is wrong – which Stephen believes is unlikely – they are all conforming to the same pattern.
At a seat-by-seat level, the field is even tighter. There are 85 constituencies where the first-placed party has a lead of five per cent or less. In 25 of these, the Conservatives are in first place with Labour in second, and in 31 Labour are in first and the Conservatives second. The long and short of numbers like this is that, as YouGov admit, two scenarios are now very much within the margin of error: a Conservative landslide, or a hung parliament where Labour could well be in with a shot of forming a minority government.
But what should worry Boris Johnson is that just about every pollster shows a late swing to Labour. YouGov’s numbers suggest it is much more pronounced in Remain constituencies it won in 2017 (where the Labour vote has ticked up by an average of six per cent) than it is in Leave ones (where it is up by an average of two per cent). The net result would be that much of what for copyright reasons we must refer to as the red wall of Labour seats in the north and Midlands stay intact, with only two Labour losses in Scotland and two gains in Greater London.
Late last week, the Tories briefed that two things stood between Johnson and a majority: turnout and tactical voting. It looked to be expectation management – CCHQ has long feared that the perception that this election is a done deal could only hurt Johnson. But on this evidence, they have reason to fear both.
Higher than expected numbers of Remain or young voteA rs or tactical anti-Tory votes in a relatively small number of constituencies could very easily swing things in an unexpected way. That dynamic is evident in Conservative-held seats that now look too close to call, such as Esher and Walton, where Dominic Raab is only two points ahead of the Liberal Democrats, and East Devon, where the Tories are neck-and-neck with independent Claire Wright. We don’t know that it will happen. But this evidence suggests that it certainly can.
The opposite is also true, of course. First-past-the-post is a hell of a drug. The Conservatives could yet cruise to a majority of 28 or bigger. Yet they have more cause to worry than at any point in the campaign. We might pause to consider at this point whether this was necessarily the best week for the conversation to shift so decisively to the NHS, and for Johnson to look quite as unempathetic as he has in recent days.
But if one thing is certain, it’s that neither main party – or, for that matter, anyone in Westminster – can afford to ignore the 18 seats in Northern Ireland, as Great Britain’s pollsters do. As I wrote late last night, the overwhelming likelihood is that the next cohort of MPs from Northern Ireland will be considerably more diverse than the near-clean sweep of DUP and Sinn Féin MPs returned in 2017, and will want different things in a hung parliament: namely a second referendum. Who is Claire Hanna? How tall is Stephen Farry? What is Colum Eastwood’s net worth? We are just a normal-sized polling error away from the answer to those questions determining who forms the next government.