The bell is about to toll

Five minutes from now. That is when YouGov’s famed MRP poll is due to be published. Half a million interviews are said to have taken place online, a dizzying number compared to what any other polling company can do. It has acquired mythic status.

Still might be wrong though.

The letters “MRP” are not a magic spell. Here is an article dated May 12th 2017 in which Lord Ashcroft writes confidently about his own soon-to-be-tested MRP model for the election of 8th June. As Anthony B. Masters wrote in this article explaining the technique,

A curious case of selective memory has surrounded the use of MRP. In the 2017 General Election, people seem to recall YouGov’s accurate central estimation of a hung parliament. Using the same technique, the Lord Ashcroft model (which estimated a Conservative majority over 60) is sometimes forgotten.


And it’s out of the gate! A dramatic gain for Labour since the previous iteration: the predictions for twenty seats have flipped from Tory to Labour. But the Oracle of YouGov is still saying there will still be a Tory majority, of 28 seats. Then again, in the manner of oracles, there is a margin of error big enough to encompass both a hung parliament and a massive Conservative majority.

I hoped this would be the fat lady singing, but all she has done is clear her throat.

Maybe Icarus can land with a glide

“The Liberal Democrats misread the political mood. Yet perhaps not all is lost”, writes Martin Kettle.

Three conclusions follow. The first is that Brexit has not reshaped the electoral battle as comprehensively as some believe. That is not to belittle the fact that Brexit has done much to recast British electoral politics. This is still a Brexit election, because its results could mark a point of no return on this most all pervasive of current issues. But it is not year zero. The idea that the Lib Dems, by being clear on the biggest issue of the day, will automatically attract all remain voters en masse to their cause is being proved false. It’s as false as Labour’s equivalent fantasy that, by being clear on the need for a radical post-austerity political economy, it will automatically attract all the votes of those who agree with that policy. In both cases, belief in practicality and trust in the leader are crucial to making the sale.

I would add to that final point about trust something that Martin Kettle almost certainly would not. It is a line from an article published yesterday by Mr Kettle’s Guardian colleague John Harris: “Labour’s ‘red wall’ is looking shaky. But the problems started decades ago”.

Mr Harris writes,

Running through a great deal of what I heard was a point voiced time and again by all kinds of people: in the absence of Brexit being delivered, why should they trust politicians to do anything else?

Goodbye secure Tory poll lead, I hardly knew ye

A couple of days ago YouGov published the first set of results of their massive Multilevel Regression with Poststratification poll. This was big news, as the equivalent from the last election in 2017 was one of only two polls that came close to the final result. (The other was a conventional poll by Survation.) The YouGov MRP results showed the responses of the hundred thousand strong panel as of 27th November. If this election were a game of The Weakest Link, that would have been the point at which the Tories said “Bank”. The results suggested a sea of red seats turning blue. (Note for American readers: your colour conventions for parties are wrong. Wrong like the spelling “color”.) As ever, though, a poll is a snapshot not a prediction.

There is no bank. Today’s poll from BMG showed the steepest absolute rise for Labour – five percentage points compared to BMG’s last poll – that I have seen so far in this election, and add to that a decline of two percentage points for the Conservatives. Net change -7, giving Con 39%, Lab 33%. One might argue – Tories might pray – that it ain’t as bad as it looks because it was quite a while since the last BMG poll. Or perhaps it is just one of those spasms that all polls are subject to, poor lambs. But however you spin it, a Conservative lead in vote share as small as six points almost certainly means no Tory majority, which means a coalition between Labour and one or more of the other parties, which means a second referendum, which means, given Labour’s plan to allow foreigners the vote, no Brexit.

When the YouGov MRP came out some Conservative campaigners had been saying that they wished it had been tighter, as such a thumping great lead would make their side complacent. They have their wish now.

I assume from Labour being up five and the Lib Dems down five that we are seeing a straight tactical transfer of voting intention from the latter to the former. The Remain vote is consolidating just as the Leave vote did before it.

Doom and gloom

Fraser Nelson in today’s Telegraph:

Let’s not pretend that the Corbyn agenda is dismissed nationwide as a socialist calamity-in-waiting. His plans to nationalise water, railways, electricity and gas are supported by about half of all voters; his idea for British Broadband Corporation is backed by a margin of three to one.

Meanwhile, Boris Johnson is interviewed by the Spectator.

I remember having conversations with colleagues in the government that came in in 2010 saying I thought austerity was just not the right way forward for the UK.

Anyone wanting to reduce the size of the state has a lot of work in front of them.

Polly put the kettle on

Tim Shipman, Political Editor of the Sunday Times, tweets:

BREAKING: First big election model seat projection predicts Tory majority of 48

Con 349

Lab 213

LD 14

SNP 49

Plaid 5

Green 1

Speaker 1

Datapraxis ran 270,000 YouGov interviews through their own predictive MRP model (like the ones that predicted the last election) 10:00 PM · Nov 23, 2019

There have been no less than four other voting intention polls published tonight, from Savanta ComRes, Deltapoll, BMG and Opinium.

These four polls showed Tory leads over Labour of 10, 13, 13 and 19 per cent respectively. The average of the four gives a slight increase in the Conservative lead. Meanwhile Lib Dem hopes were getting bounced up and down: Deltapoll showed them up 5, but the other three were down a little or steady.

But the big news was the MRP – Multilevel regression with poststratification – poll. In the previous election YouGov’s MRP poll was one of only two (alongside a conventional poll by Survation) that did well. This one is slightly different, in that YouGov’s data has been used, but the MRP model is not their own but that of an outfit called Datapraxis.

Polls do not foretell the future. Things could still change. They did in 2017. But this is a blow to Labour’s hopes.

This same poll might not be so great for the personal hopes of some Conservative MPs. A subsequent tweet from Tim Shipman adds,

The Datapraxis MRP modelling predicts that 7 big beast Brexiteers are in danger of losing their seats.

Read the whole tweet to find out which MPs they are. Number 6 will worry you.

Making doubly sure Leave cannot win

It was always clear that the “credible Leave option” that the Labour party proposes to pit against Remain in a second referendum is a fake. The plan is for Sir Keir Starmer and Emily Thornberry to get together with their EU opposite numbers and, after the least arduous negotiations imaginable, to emerge beaming with a “deal” deliberately designed to be as unattractive as possible.

But if that wasn’t enough, the Labour manifesto plans to make assurance doubly sure, as Macbeth said when deciding to murder Macduff. I quote:

We will oversee the largest extension of the franchise in generations, reducing the voting age to 16, giving full voting rights to all UK residents, making sure everyone who is entitled to vote can do so by introducing a system of automatic voter registration, and abandoning plans to introduce voter ID which has been shown to harm democratic rights.

“It’s time for real change – The Labour Party Manifesto 2019”

The innocuous phrase “all UK residents” includes non-UK citizens from any country. As it says elsewhere in the manifesto, there are three million EU residents in the UK. They will get to vote on whether the UK remains in the EU. The majority by which Leave won the 2016 referendum was 1.3 million.