Inspired by @prossertj and @p_surridge I've had a quick look at the pattern of public sentiment about Keir Starmer.

@BESResearch asked respondents about Starmer (and Phillips, LongBailey, Rayner) in their post-2019GE wave.

Here are the weighted spread of answers:

/1

Scale is a bit (!) tortured, but the main features are:

* if you're not a party leader in a GE, most people DK you well enough to feel anything
* Starmer was the best known of the LAB-LE2020 candidates (who happened to be in this question selection)

/2

For the minority who have feelings about Starmer, what drives (well, regresses against) them?

* eurefVote/2nd ref
* feelings about Labour
* feelings about LDs (and SNP/Greens)
* political attention
* trusting courts
* generalised propensity to like politicians

/3

What if we just look at the Labour membership (as of just before the 2019GE)?

* feelings about Phillips
* eurefVote
* euID (identification with Leave/Remain)
* feelings about Blair
* feelings about where LibDems stand on redistribution
* work in a company with 500+ employees

/4

For contrast, here's the same breakdown of regressors for sentiment about Jess Phillips among the membership

* Corbyn vs MPs I grasp
* really heavy regression with Swinson spec. ... more puzzling
* perc of how LDs prioritise environment/economy (relates to Green New Deal???)

/5

And for LongBailey

* Rayner/Corbyn/SNP feelings
* Brexit impact will be not so bad
* Thought Con would lose in 2019/Greens wd win constituencies
* Thought 2019GE was conducted unfairly

/6

It's tiresome to have to keep track of commonalities/flick between charts to make comparisons, so I'm finishing with a series of head-to-head comparisons (i.e. "likeStarmer - likeLongBailey") on *all respondents now* (not just membership).

Let the death matches begin!

/7

like Starmer more(/less) than LongBailey:

Leave --
Corbyn/Rayner --
LD vote in euros/Phillips ++
Political Attention +
Left (self-placement)--

Overall LD vs BXP split

/8

like Starmer more(/less) than Phillips:

I made dubious noises when @JadeFrancesAzim made a post about issues in Labour about electing a female leader (because N=3 doesn't seem adequate) ...

... but this Starmer/Phillips split is *interesting*

/8

Perfect (sentiment target) gender alignment up to LongBailey (where it's messy). I think I'd probably dismiss Corbyn as a confound ... but Rayner/Swinson/Moran/Davey kind of does feel like a point is being made.

/9

Could be people privileging men over women or women over men - de-tangling that would require some non-trivial work figuring out what the baselines "ought" to be.

But makes me feel like it's worth coding and tossing in a likeMalePol/likeFemalePol average variable.

/10

like LongBailey more(/less) than Phillips:

* Continuity Corbyn vs Dissidents
* LD vs BXP

/11

Old Leader vs New Leader:

First off, most significant thing is what *isn't* here - nothing to do with Brexit (in the top 20 regressors) - significant in both directions (not making the party support base more Leave-y or more Remain-y based on personal leader sentiment)

/12

Starmer has
* greater scope to reach beyond Labour partisans/people self-identify as Left
* more likely to fish among ppl who broke for LD/remain LD than BXP

No surprises - but probably the most significant thing is that he appears to do better with **older voters**

/Fin