In this poll Labour are 'losing' more votes to the Lib Dems than to Reform and the same number to the Greens. (I say losing but it is probably partially a tactical unwind at this point). The Reform lead is at least in part an enthusiasm gap very few 2024 Reform voters are undecided. - ThreadSky

psurridge.bsky.social • 4 days ago

In this poll Labour are 'losing' more votes to the Lib Dems than to Reform and the same number to the Greens. (I say losing but it is probably partially a tactical unwind at this point).

The Reform lead is at least in part an enthusiasm gap very few 2024 Reform voters are undecided.

Reposted from YouGov

Latest YouGov Westminster voting intention (23-24 Feb)

Ref: 25% (-2 from 16-17 Feb)
Lab: 24% (-1)
Con: 22% (+1)
Lib Dem: 16% (+2)
Green: 8% (-1)
SNP: 3% (=)

yougov.co.uk/politics/art...

Comments

duncanstott.bsky.social•4 days ago

The same tactical voting pressures on the left will return come the next election. That's nothing for Labour to fret over at this stage.

psurridge.bsky.social•4 days ago

I tend to agree, and a significant Reform challenge probably heightens them if anything but plenty of time for any/all the parties to upset different parts of the electorate.

chrisworthing.bsky.social•4 days ago

Increase in tactical voting

mariosrichards.bsky.social•4 days ago

Probably significantly different - but really hard to tell how/how much different it could be at next from this polling right now.

gsds3.bsky.social•4 days ago

In sight of the tories being in 4th place. Not wholly implausible at this point. Quite the triumph for Kemi

martynomad.bsky.social•4 days ago

None of the major parties are really addressing the same issues arising from Reform as the AfD threw up in Germany.
Unregulated misinformation thro media, bombardment of social media - print media chat with shills and bots and a dodgy spiv who needs to be checked on financially.
It won't wait

t0nyyates.bsky.social•4 days ago

Can you explain the last point. Is this alluding to weighting by likelihood to vote - a point you made previously?

mariosrichards.bsky.social•4 days ago

I think simpler than that, this is the case of "what do you with people who say 'Don't Know'/'Would Not Vote' *and at starkly different rates by 2024 vote*?".

psurridge.bsky.social•4 days ago

Yes this - and I think @yougov are modelling them rather than just excluding them - which may be the right approach or might not be but you have to dig a fair bit to know that

mariosrichards.bsky.social•4 days ago

> I think @yougov are modelling them rather than just
> excluding them

I did not know that - feels like an unfair burden to put on a polling firm.

I can see the pressure for that in the run up to a GE - but I feel like this far out it would be better to present data as raw & clear as poss.

psurridge.bsky.social•4 days ago

They retained the MRP element for the headline VI - it notes it in the tables but it isn't immediately obvious. I guess once the code is set up dropping new data in isn't a huge task.

mariosrichards.bsky.social•4 days ago

Hmm - I feel for their situation!

An understated reason for switching to PR is that it would make politics significantly more transparent.

mariosrichards.bsky.social•4 days ago

As in, close to a GE, you can meaningfully hold advanced Yougov complex models to account (with the GE result) - so there's a logical and business for trying to be extra clever (at the cost of some transparency).

mariosrichards.bsky.social•4 days ago

But at this point in the cycle, these projections will never be tested against reality - so why use a complex model whose failure modes will never be tested?

mariosrichards.bsky.social•4 days ago

So, say you have the situation where your big online panel tagged by 2024 vote shows Reform voters 90% saying "I'd vote Reform" but your Con/Lab/LD/Green voters are 50% saying "I'd same/adjacent party" and 50+% DK.

mariosrichards.bsky.social•4 days ago

If you drop DKs, Reforms vote doubles even though strictly speaking the number hasn't risen.

If you include DKs, what are people supposed to do with that? Build a giant model to infer how those DKs might shift over the course of 4 years?

mariosrichards.bsky.social•4 days ago

It's legitimate for people to want a projection that doesn't require a PhD in statistical inference - and enthusiasm gaps are meaningful information - but people are putting these raw numbers into seat calculators and talking about adding them together to build an automatically winning coalition ...

kaynells.bsky.social•4 days ago

Rely on the "theyll come back to Labour because they have no choice" at your own peril...

You can only really rely on the far right keeping coming.

We need real ambitious politics to stop them. Labour have blown their first impression.

mariosrichards.bsky.social•4 days ago

I remain strongly of the opinion that the correct format for these charts should include DK and WNV (lumped/separately).

It's too much for consumers to expect polling firms to 'extrapolate past that' - and too much to expect consumers to dive into tables to reconstruct that.

robpountney.bsky.social•4 days ago

Interestingly the only polls that I ever come across that do this are polls on Welsh independence, a clearly political choice as it makes support for independence look a fair bit smaller, they wouldn't get away with it in Scotland, which is presumably why they don't try...

columnist.bsky.social•4 days ago

💯

Comments

Posting Rules

Reply