Latest YouGov data: Starmer now has a higher net leadership rating than Johnson
Starmer +23
Johnson +22
Starmer's ratings have considerably improved, while Johnson's declined by 18 points.
I won't hold my breath for Starmer's critics on the left to acknowledge they were wrong.
Additional data - Starmer's leadership ratings are in strong net positive territory in every age group:
18-24: +21
25-49: +17
50-64: +28
65+: +28
Importantly, Starmer is landing very well with the exact age demographics Labour did so badly in at the 2019 General Election
2/4
Jan 12, 2020 • 10 tweets • 3 min read
Some on the left are attacking Starmer, claiming:
1) He didn't prosecute re: Ian Tomlinson 2) He didn't prosecute re: De Menezes shooting 3) His campaign team is a cabal of lobbyists 4) He isn't a real socialist
Serious charges. But are they true?
A thread, with evidence
1/
Firstly, the decision not to prosecute over the death of Ian Tomlinson.
A serious claim indeed. And, it turns out, a complete lie.
The officer was prosecuted. It went to court in 2012 and he was acquitted by a jury.
We're going to hear a lot of crowing from centrists.
"Corbyn WAS the problem"
Their analysis is only partially correct - and not for the reasons they think.
Labour members should arm themselves with the facts, and logic, to win the fight for the soul of their party.
1/10
The centrist argument essentially goes like this:
"Corbyn was unpopular, therefore 'Corbynism' is unpopular. The solution is a return to centrism."
But like every centrist argument of the last five years, this 'analysis' collapses under even the most basic scrutiny.
2/10
Dec 13, 2019 • 17 tweets • 3 min read
A sobering fact.
Remain parties won 53% of the vote.
Leave parties won 47% of the vote.
How, then, did the Tories emphatically win?
And what does Labour need to do about it?
(and, yes, at the end, I will highlight why my polling analysis was wrong)
1/17
The Tories won due to:
1) Distribution of Leave vote 2) Ruthlessness 3) Splintering of remain
In terms of 1)
If you imagine Leave and Remain as political parties, Leave would win on 47% of the vote, because Leave is more evenly distributed. Remain piles up in cities.
2/17
Dec 11, 2019 • 21 tweets • 4 min read
These are the factors that will win or lose the election:
1) Turnout by age 2) Lab leavers 3) Tactical voting 4) Con remainers
And in terms of predicting the outcome we also need
5) The accuracy of the polls
Let's predict all five - and make a final call prediction.
1/19
Firstly though, tonight's polls.
Treat 'final polls' with the contempt they deserve.
The final polls are nearly always wrong. And they're always wrong in the same direction (too Lab or too Con, uniformly).
This shouldn't happen. But there's a reason it does: herding.
2/19
Dec 11, 2019 • 15 tweets • 4 min read
YouGov MRP turnout weighting. Oh dear.
I posted a quick analysis thread last night, and raised concerns about the turnout weighting.
In this thread, let's explore the issue further.
I'll try to give a simple, clear explanation of the problem.
Firstly, how do polls measure turnout (i.e. % who turn up and vote)?
After 2015, most except Kantar & YouGov MRP use 'self report'.
i.e. they believe you if you say you are 100% gonna vote.
They may however 'downweight' you if you are a first time or previous non-voter.
2/13
Dec 10, 2019 • 8 tweets • 2 min read
Quick analysis YouGov MRP
On a Lab vote share of 34%:
1) Lab are advancing strongly in the south compared to previous MRP and are positioned to make gains 2) Lab 'red wall' firming up ~15 seats saved 3) A LOT of tight marginals
What does this mean?
1/4
If Lab get to just ~37% on polling day (on current trends this seems plausible and likely) the following happens:
1) They could take ~10 seats from Con in the South 2) They save all but ~10 seats in the Red Wall 3) LD's take ~5 seats from Con 4) SNP take ~6 seats from Con
2/4
Dec 10, 2019 • 12 tweets • 3 min read
"What's the REAL Tory poll lead?" - UPDATE
In my original thread I outlined why the Con lead may be inflated.
I also invited feedback, and boy, did I get it - some of which has been very insightful, and I am grateful.
It is no surprise that Lab 'northern wall' seats show limited or no rise in voter registration.
2/6
Dec 9, 2019 • 8 tweets • 2 min read
We need to talk about the Brexit Party.
Their withdrawal from Tory held seats increased the national Con vote share from an average of 39% to 42%.
But this increase, in terms of gaining Lab seats, is worth nothing. And is likely to benefit Lab.
Why? Let's find out...
1/8
The withdrawal of the Brexit Party from Con seats achieved two things for Con;
1) It made retaining existing seats a bit easier
2) It increased their vote share in those seats, and thus increased their national vote share
It did not increase their vote in Lab seats.
2/8
Dec 8, 2019 • 8 tweets • 3 min read
Do the Tories believe they are going to win a big majority?
To answer that question, let's analyse where the Tories are focusing resources.
I've been analysing their constituency-level advertising, and they are not acting like a party expecting a large majority...
1/6
In the last few days, there have been two dramatic changes.
1) They've removed Boris Johnson's face and any reference to him on constituency specific ads
2) They have moved massively into defence mode
Let's explore the details...
2/6
Dec 8, 2019 • 18 tweets • 4 min read
What is the REAL Tory poll lead?
In my thread "are ALL the polls wrong?" I identified sources of error that inflated the Con lead.
In this thread, I go further.
I have re-modelled/weighted the data to try to correct these errors.
The real Con lead? Read on to find out.
1/17
The sources of error I will be attempting to correct are;
1. Turnout 2. Over-weighting of leavers 3. Under-weighting of GE17 non-voters and newly registered voters
I will be clear exactly how I am doing it, so anyone is welcome to tell me if I'm wrong, and how.
2/17
Dec 7, 2019 • 10 tweets • 2 min read
Last night's post-debate polls by YouGov showed a narrow Johnson win.
Johnson 52%
Corbyn 48%
Good news for Johnson, right? Well... no.
Let's look into the raw data and uncover some key points hidden by the headline numbers.
1/9
Firstly, who took part in the poll?
We have a sample of 1322 people, weighted to 2017 voting intention, EU voting intention, etc.
Looking at the numbers, Leavers however may have been over-weighted and 18-24's under-weighted. This will make the sample lean toward Con.
2/9
Dec 5, 2019 • 19 tweets • 4 min read
Could ALL the polls be wrong?
I've been looking at the data - and the short answer is yes.
We've had two elections (2015 and 2017) where pretty much every pollster was wrong.
So, what fresh polling disasters await us in 2019? Let's find out...
A thread, with data.
1/15
Looking at the data, there are three potential sources of error that could mean all the polls are wrong - AGAIN.
1) 2016 EU referendum weighting 2) Weighting (or not) to 2017 GE vote 3) New voter registrations
Let's explore these one at a time...
2/15
Nov 28, 2019 • 12 tweets • 4 min read
The YouGov MRP.
Terrible news for Labour? No.
Dig a little deeper, and all is not what it seems.
There are numerous trends and data points that indicate that Labour are, in my view, LIKELY to deprive Johnson of a majority.
A thread - with supporting charts/data.
1/12
Let's start with the YouGov 'house effect'.
Every pollster has a different approach to crunching the data.
YouGov have consistently reported the lowest Lab % in 2019.
See attached example from this year. They were the only pollster not reporting a Lab lead for MONTHS.
2/12
Nov 27, 2019 • 5 tweets • 2 min read
How confident are the Tories of winning?
The consensus is a 30+ seat majority. But their own targeted advertising paints a different picture.
In the last week, they've moved from offence to defence. Something has them spooked.
A thread (with supporting charts/data).
1/1
Analysis of ACTIVE constituency level ads has shown that the Conservatives have lowered ambitions.
A week ago, the focus was on Lab seats with larger majorities.
Now the focus is on ultra-marginals with a handful of Lab leave-heavy seats.
Attached: live ads in Eng/Wales
2/2
Nov 26, 2019 • 5 tweets • 2 min read
The new @KantarTNS poll shows the Tory lead dropping by 7%, and an 11% Con lead. The headline figures are weighted using Kantar's turnout model.
So what happens if we re-weight the data to 2017 General Election turnout levels?
In short - a Labour surge...
1/1
At the moment Kantar are using a turnout model that's hard to fully pick apart in their raw data.
We do know is that it is modelled based on 60% overall turnout. The 2017 general election turnout was 68.8%.
This matters. Occasional voters are more likely to vote Labour.
2/2
Analysis of ACTIVE constituency level ads has shown that the Conservatives have lowered ambitions.