Opinium/Observer Graph thread - 20th June (and 13th June)

Voting Intention - Local Regression:

yeah things have basically stabilised out at a 5-6 point lead for the Tories, which I think you'd take rn if you were Labour considering Labour were 22 points behind in April 1/

Remember, this is 5-6 points behind *before* the government has to start making hard decisions on the economy and (probably) has to go through an inquiry into 40-60k deaths...

2/

Personally, I think the fact that BLM has halted the Tory decline highlights how "culture war" is this government's home ground.

3/

So where are those extra Labour voters compared to the GE coming from then? Conservatives? Lib Dems? Greens?

4/

Here's an Alluvial-Sankey diagram showing how voters are moving compared to 2019

We can see that there's not much movement between the main parti....SHIT LOOK AT THE LIB DEMS
https://public.flourish.studio/visualisation/2916296/

5/

tbh I don't like Sankey diagrams for voting intention. They're often cluttered, require deconstruction and are awful at showing net flows, which is what matters in the context of an election.

So let's look at something better imo:

6/

Here's a chord diagram of the same data. Chord diagrams are basically circular Alluvial-Sankey diagrams. They're less numerical but more proportional - and show net flows better.

https://public.flourish.studio/visualisation/2790503/

7/

Basically, each of these arcs (chords) represent a flow between parties. The two endpoints of each arc show the proportion of 2019 voters moving FROM one party to other (the arc end-points represent LOSSES not GAINS)

8/

So here we can see that, proportionally, Con are losing the most voters to Lab, BUT some Lab voters are also moving Lab->Con, so the net movement isn't that much.

Similarly, Lab->Green and Green->Lab are both basically cancelling out.

9/

However, here we can see how stark the net movement is from LD->Lab. About HALF of LD voters polled are moving to Labour, and virtually no-one is moving in the opposite direction.

This is where the biggest chunk of Labour's +7% compared to GE2019 is coming from.

10/

(there's still turnout and non-voters to worry about - I'm working on a way of visualising that soon)

11/

Unusually for this thread, we're now gonna look at a negative for Starmer:

His net approval is going down (-8 in a month)

12/

This is being driven by his disapprovals trickling up - now over 20% for the first time - whilst his approvals remain roughly constant.

This isn't surprising for a politician tbh; more exposure = less liked (generally speaking), but the movement is fairly small...

13/

I did wonder if those disapproves might have been coming from some disgruntled Labour members (gut feeling was that Keir hasn't had a great few weeks with progressives), but nope, clearly coming from Tories deciding "actually, maybe I shouldn't be liking the Lab leader..."

14/

(probs some non-voters cooling a bit from Keir as well, but I didn't have time to graph this)

15/

Johnson's movement is negligible - still a slight net-negative: (-5)

Possibly some "neutrals" moving to approve, but tbh I'd want to see that sustained for a few more weeks before calling it a mini-rebound.

16/

Keir is still tantalisingly 1 point off Johnson in Best PM.

This is 100% because I cursed it by saying it was only a matter of time a few weeks ago. Sorry, everyone...

17/

I got no more here because there's really no more to say. Everything's flattened out now in terms of public opinion.

As we've said before though, the next months are worth watching as the Gov moves to trickier ground

18/ends