Local Council by-elections 16th and 18th April 2024

[The Bishop]

Apr 19, 2024 7:14:04 GMT andrewp said:

Waverley

Res 32.5% (-17.7)
LD 29.6% (new)
Lab 23% (-11.7)
Con 14.9% (-0.3)

As seen previously, Labour only fielding one candidate last year perhaps slightly exaggerates their "real" fall now.

[Davıd Boothroyd]

One of the reasons why I now favour the 'total vote' approach over the 'averages' approach when working out vote percentages in multi-member seats.

[Pete Whitehead]

Apr 19, 2024 11:56:49 GMT The Bishop said:

Apr 19, 2024 7:14:04 GMT andrewp said:

Waverley

Res 32.5% (-17.7)
LD 29.6% (new)
Lab 23% (-11.7)
Con 14.9% (-0.3)

As seen previously, Labour only fielding one candidate last year perhaps slightly exaggerates their "real" fall now.

Not really in this case. There were two candidates each for Conservatives and Residents who seem to have got fairly equal levels of support. It's not as if there were also single Lib Dem or Green candidates which might have benefited a single Labour candidate. As a proportion of ballot papers issued I'd guess the Labour vote share would be even higher than the 34.7% implied there. You'd have had a lot of Labour plumpers I guess then the rest mostly giving their second vote to one or other of the Residents. Had there been two Labour candidates, obviously most of that latter element would have also voted for the second candidate. What exaggerates their vote share in 2023 (and hence their fall now) is the absence of any Lib Dem candidates then.

[Pete Whitehead]

And indeed modgov.waverley.gov.uk/mgElectionAreaResults.aspx?XXR=0&ID=403&RPID=18932947 the proportion of those who cast a vote who voted for the Labour candidate was 38.5% in 2023 and this was a first choice vote - it was not as if they were the beneficiary of spare second votes from another party which had undernominated

[The Bishop]

Thank you for that detailed look at the result

[minionofmidas]

Apr 19, 2024 12:01:00 GMT Davıd Boothroyd said:

One of the reasons why I now favour the 'total vote' approach over the 'averages' approach when working out vote percentages in multi-member seats.

total votes as a share of available votes (ballot papers with at least one valid vote times number of votes available). You still need to asterisk any incomplete slate, but at least what you're measuring are the actually extant votes.

[The Bishop]

Alerting jamesdoyle for his weekly scores on the doors

[jamesdoyle]

GWBWI

Con +53
LDm +45
Lab -11

A win for the Cons - with positive scores everywhere (albeit only +1 in Waverley).