2021-22 Local government by-elections

[Terry Weldon]

I have just learned that Cllr Peter Isherwood (Hindhead ward, Waverley BC) passed away last night.

[jamesdoyle]

Over the course of the municipal year, 2021-22, the Conservatives defended 99 seats.

This shows the week-by-week distribution of the defences:



I was interested to see if there was a pattern to the losses, so this image shows the same defences, but now with losses indicated (red for losses to Labour, gold LibDem, green Green, yellow SNP, white Independent). To make it easier to see, I've dropped the losses 'below the line'. Apart from the appallingly bad period around early December 2021, the losses seem reasonably evenly distributed - certainly not clumped to an extent that any useful info could be extracted based on such a small sample.



As part of my Good Week/Bad Week index, I calculate a value for how 'safe' a seat is for the defending party: this is based on wins in the ward, not vote share. A ward where the defending party has won every seat over the past four cycles gets the maximum score of 100.

The next chart shows the defences separated into five quintiles, based on this Safe Ward Index - so the first band is SWI 0-20; then 21-40, 41-60, 61-80 and 81-100 (effectively rock solid wards).



Unsurprisingly, there are few defences in the lower bands (because if they are being defended now, they have been won in the most recent cycle); also unsurprisingly, these were largely lost: 2/2 for SWI 0-20; 5/6 for 21-40.
In the middle band (SWI 41-60), the Tories lost just over half (6/11).
Bearing in mind the small sample numbers, this makes sense, losing almost all the weak seats, and roughly half of the central band.
There are considerably more defences in the higher bands, but the losses are still significant: 16/30 for SWI 61-80; 18/50. This is roughly half of the strong seats, and over a third of the rock solid seats. This is more than I would expect, although I'd like to compare to other years to see if this is a change from previous history. I'd also like to look at these solid seats and see if the losses correlate to 'tactical alliance' elections where one or more of Lab/LDm/Grn stand down in favour of another.

I hope this is of some interest.

[gwynthegriff]

Jun 3, 2022 11:00:35 GMT jamesdoyle said:

Over the course of the municipal year, 2021-22, the Conservatives defended 99 seats.

This shows the week-by-week distribution of the defences:



I was interested to see if there was a pattern to the losses, so this image shows the same defences, but now with losses indicated (red for losses to Labour, gold LibDem, green Green, yellow SNP, white Independent). To make it easier to see, I've dropped the losses 'below the line'. Apart from the appallingly bad period around early December 2021, the losses seem reasonably evenly distributed - certainly not clumped to an extent that any useful info could be extracted based on such a small sample.



As part of my Good Week/Bad Week index, I calculate a value for how 'safe' a seat is for the defending party: this is based on wins in the ward, not vote share. A ward where the defending party has won every seat over the past four cycles gets the maximum score of 100.

The next chart shows the defences separated into five quintiles, based on this Safe Ward Index - so the first band is SWI 0-20; then 21-40, 41-60, 61-80 and 81-100 (effectively rock solid wards).



Unsurprisingly, there are few defences in the lower bands (because if they are being defended now, they have been won in the most recent cycle); also unsurprisingly, these were largely lost: 2/2 for SWI 0-20; 5/6 for 21-40.
In the middle band (SWI 41-60), the Tories lost just over half (6/11).
Bearing in mind the small sample numbers, this makes sense, losing almost all the weak seats, and roughly half of the central band.
There are considerably more defences in the higher bands, but the losses are still significant: 16/30 for SWI 61-80; 18/50. This is roughly half of the strong seats, and over a third of the rock solid seats. This is more than I would expect, although I'd like to compare to other years to see if this is a change from previous history. I'd also like to look at these solid seats and see if the losses correlate to 'tactical alliance' elections where one or more of Lab/LDm/Grn stand down in favour of another.

I hope this is of some interest.

A bloody impressive piece of work!

[The Bishop]

Yes, that is a genuinely interesting way of presenting outcomes. And of course the same could be done for other parties.

[jamesdoyle]

Jun 3, 2022 12:38:34 GMT The Bishop said:

Yes, that is a genuinely interesting way of presenting outcomes. And of course the same could be done for other parties.

I haven't checked the figures but there are far fewer losses for the other parties, and part of my impulse to do this was to see if governmental/national issues showed up.
I might have a look at the others, as all the data is in my spreadsheet now , so it's just a matter of plotting it.

[jamesdoyle]

Labour, in the same formats as for the Cons above:

Defences:



Holds and Losses:



Quintiles by ward safeness:

[jamesdoyle]

And the LibDems:

Defences:



Holds and Losses:



Quintiles by ward safeness:

[jamesdoyle]

And for completeness (not a lot here), the Greens, SNP and PC:

Defences (Greens in dark green, PC in bright green, SNP in yellow):



Holds and Losses:



Quintiles by ward safeness:

[ricmk]

Given how the Tories have started 2022-23 those 5 losses at the end of 2021-22 now seem a bit of a pointer, the rot started well before mid-May.

[jamesdoyle]

Jun 5, 2022 14:15:21 GMT ricmk said:

Given how the Tories have started 2022-23 those 5 losses at the end of 2021-22 now seem a bit of a pointer, the rot started well before mid-May.

I think the pointer might well have been the losses the first three Thursdays in December: 21 defences, 15 losses, and 9 of those 15 coming in wards where the Cons would have expected to defend easily. Not only that, but they lost to four different parties.

It could be that the opposition are co-ordinating their activities more; or that voters are doing it for them;  or that the opposition parties are able to field candidates and run campaigns in wards were the Cons have been unchallenged for a generation. Or a bit of all those. But it spells danger for the Cons if it continues.

[thirdchill]

Jun 6, 2022 9:12:10 GMT jamesdoyle said:

It could be that the opposition are co-ordinating their activities more

There was definitely an element of this in May. Though more likely caused by parties not standing in seats (not necessarily deliberate) than deliberate cooperation campaigns. Cons seemed to do particularly badly when up against only one opponent, whichever party it might be. Particularly if it was a Con/LD, Con/Grn or Con/Ind contests, though they didn't do that well in Con/Lab contests either.

When multiple parties stood, Con held up (relatively) better than in the scenarios above.