2018 review - Yorkshire and the Humber

[J.G.Harston]

Dr Jordan's model from last time works quite well across eastern Sheffield. There's a natural lump sticking out of Richmond which removing makes an updated Heeley just right. Similarly, the north of Burngreave is Shirecliffe which was only ever added into Burngreave Ward to make up the numbers, rolls over the ward boundary, and is a natural split and companion with the areas to the north.

Western Sheffield is the right numbers for two seats, but other than the logical split along the western Rivelin Valley where to go from there on is a puzzle. I can see it looking like a difficult divorce with both parties wanting the house but not the children.

[YL]

Here's a map of the first try from last week (the "old wards" version):



(ward boundaries from election-maps.co.uk)

[Crimson King]

Feb 24, 2016 18:03:14 GMT YL said:

Metropolitan Boroughs


Bradford 4.37
Calderdale 1.91
Kirklees 3.95
Leeds 6.89
Wakefield 3.17

Grouping Bradford and Calderdale is just possible (the entitlement is 6.28, compared with a maximum of 6.30 for six seats). If that happened, Leeds could be treated on its own, but I suspect a more likely solution is to group Leeds and Bradford for 11 seats or Leeds, Bradford and Calderdale for 13. Including part of Wakefield with South Yorks would then allow Kirklees to be treated on its own.

I appreciate it is outwith the formal rules, but I wonder to what extent the West Yorkshire devolution settlement, which I understand is imminent, will effect thinking on this. There may be less disinclination (I put it that way round deliberately) to crossing authority boundaries in the context of a broader body

[J.G.Harston]

Feb 28, 2016 9:31:52 GMT YL said:

Here's a map of the first try from last week (the "old wards" version):

(ward boundaries from election-maps.co.uk)

Having Stannington joined to something south of the Rivelin seems so wrong, the last review that did that seems an abberation. It has nothing really in common with Ecclesall Bierlow. Similarly, the reach-around to pull Dore&Totley onto Heeley seems wrong, even though that also popped up once before.

I've had a play to see what I can do across the west. Using the old ward map shows little difference from the new ward map as most of the ward boundary changes end up away from the constituency boundaries.

old wards

new wards

[YL]

Feb 28, 2016 10:12:52 GMT Crimson King said:

Feb 24, 2016 18:03:14 GMT YL said:

Metropolitan Boroughs


Bradford 4.37
Calderdale 1.91
Kirklees 3.95
Leeds 6.89
Wakefield 3.17

Grouping Bradford and Calderdale is just possible (the entitlement is 6.28, compared with a maximum of 6.30 for six seats). If that happened, Leeds could be treated on its own, but I suspect a more likely solution is to group Leeds and Bradford for 11 seats or Leeds, Bradford and Calderdale for 13. Including part of Wakefield with South Yorks would then allow Kirklees to be treated on its own.

I appreciate it is outwith the formal rules, but I wonder to what extent the West Yorkshire devolution settlement, which I understand is imminent, will effect thinking on this. There may be less disinclination (I put it that way round deliberately) to crossing authority boundaries in the context of a broader body

It's hard to know, but I suspect it's more likely to be used as an excuse for crossing boundaries rather than a genuine reason.

I've been thinking a bit more about West Yorks. Some thoughts borough by borough:

Wakefield. If, as I suggest, one ward is used to fix Doncaster, then that leaves 20 wards and close to three quotas. It turns out that there's enough variation in the ward sizes to have one 6 and two 7s, and indeed there's more than one way of doing it, one of which keeps Normanton, Pontefract & Castleford unchanged.

Kirklees. OK, whose idea was it to divide a borough with close to 4 quotas into 23 wards? The current Dewsbury and Batley & Spen are within quota, which suggests all that's needed is to fix the other two seats, but that actually requires a very tight ward split, with only about 100 electors to play with. There is probably a way which involves an easier ward split but doesn't preserve the two OK constituencies. Borrowing a ward from Bradford didn't seem to work well; borrowing a ward from Calderdale causes problems there. If you don't want to split a ward a more complicated map involving multiple border crossings seems likely.

Calderdale. The current Calder Valley is within quota, but unlike in the zombie review simply adding Queensbury to Halifax doesn't work. I haven't tested all the Bradford wards to see whether there's one which does. What is done in Kirklees may have big knock on effects here.

Bradford. Bradford has 30 wards and a bit over four quotas. The ward sizes are quite variable; in particular Bingley and Bingley Rural are both large which makes it a challenge to find a legal constituency including both. There are, however, lots of combinations of seven Bradford wards which do work.

Leeds. Leeds is actually going to be able to escape some of the horrors proposed in the zombie review even without ward splitting, as IER appears to have reduced the electorates of many of the central wards to the extent that there are several ways of combining five of them for legal constituencies, including the current Leeds Central, so no Leeds Metropolitan & Ossett nonsense this time. However, ward splitting is still going to be needed if the west of the city is going to be covered without crossing borough boundaries. (I haven't looked to see whether there's a plausible solution which doesn't split wards but does cross borough boundaries.)

I do have a 10 seat map for Bradford, Calderdale and Kirklees without any ward splits, but it contains some truly horrible constituencies. The lowlight is possibly Bradford Central & Bingley, which does not include Shipley ward.

[Crimson King]

I have to see that one!

[YL]

Valleys of Colne & Calder (hi doktorb🏳️‍🌈🏳️‍⚧️ ) 73,998
Huddersfield 76,765
What on earth do you call this? Huddersfield East, Liversedge & Denby Dale 78,273
Batley & Dewsbury 75,378
Bradford South & Cleckheaton 75,606
Halifax 75,348
Bradford West & Brighouse 78,071
Keighley 76,636
Bradford Central & Bingley 77,734
Bradford East & Shipley 77,243

(NB: this is not a serious proposal.)

[Crimson King]

I could live with East and Shipley

[Davıd Boothroyd]

You could call it Heavy Woollen CC. That would confuse everyone who didn't know the area.

[Sibboleth]

Bradford Central & Bingley is the greatest troll proposal. Love it.

[YL]

Feb 28, 2016 13:57:48 GMT J.G.Harston said:

Feb 28, 2016 9:31:52 GMT YL said:

Here's a map of the first try from last week (the "old wards" version):

(ward boundaries from election-maps.co.uk)

Having Stannington joined to something south of the Rivelin seems so wrong, the last review that did that seems an abberation. It has nothing really in common with Ecclesall Bierlow. Similarly, the reach-around to pull Dore&Totley onto Heeley seems wrong, even though that also popped up once before.

I agree neither of those is ideal, though I think both could be defended: Stannington on the grounds that it's like that now, and Dore & Totley on the grounds of historical links and the fact that the distribution of population in the ward means it isn't quite as bad as it looks.

Where it came from was that if you try to get the 27 wards excluding Stocksbridge (because I thought that might make the rest of SY work more easily) into five seats with minimal splits you really want a 5 ward seat and four 5 and a bit ward seats. I could only find two 5 ward seats which worked, and preferred that one on minimal change grounds, as the BCE tend to like minimal change.

However, the other one is workable too. The 5 ward seat contains Stannington, Crookes, Hillsborough, Walkley and Central (71,482). I don't like the inclusion of Central much but it may be no worse than the things above. Then Fulwood, Ecclesall, Broomhill, Nether Edge, Dore & Totley and part of Beauchief & Greenhill can form a seat similar to the 1974-97 Hallam. Heeley loses part of B&G but gains Manor Castle, Attercliffe & Brightside is as in my previous suggestion, and the rest (the Ecclesfield wards and the current Brightside & Hillsborough minus, um, Brightside and Hillsborough) forms a final seat. That's probably neater (even with the Central issue) but it's definitely more of a change from the current map. It's also harder to come up with sensible names...

[Deleted]

Mar 1, 2016 20:22:31 GMT YL said:

Feb 28, 2016 13:57:48 GMT J.G.Harston said:

Having Stannington joined to something south of the Rivelin seems so wrong, the last review that did that seems an abberation. It has nothing really in common with Ecclesall Bierlow. Similarly, the reach-around to pull Dore&Totley onto Heeley seems wrong, even though that also popped up once before.

I agree neither of those is ideal, though I think both could be defended: Stannington on the grounds that it's like that now, and Dore & Totley on the grounds of historical links and the fact that the distribution of population in the ward means it isn't quite as bad as it looks.

Where it came from was that if you try to get the 27 wards excluding Stocksbridge (because I thought that might make the rest of SY work more easily) into five seats with minimal splits you really want a 5 ward seat and four 5 and a bit ward seats. I could only find two 5 ward seats which worked, and preferred that one on minimal change grounds, as the BCE tend to like minimal change.

However, the other one is workable too. The 5 ward seat contains Stannington, Crookes, Hillsborough, Walkley and Central (71,482). I don't like the inclusion of Central much but it may be no worse than the things above. Then Fulwood, Ecclesall, Broomhill, Nether Edge, Dore & Totley and part of Beauchief & Greenhill can form a seat similar to the 1974-97 Hallam. Heeley loses part of B&G but gains Manor Castle, Attercliffe & Brightside is as in my previous suggestion, and the rest (the Ecclesfield wards and the current Brightside & Hillsborough minus, um, Brightside and Hillsborough) forms a final seat. That's probably neater (even with the Central issue) but it's definitely more of a change from the current map. It's also harder to come up with sensible names...

Names are a sinch.

[YL]

It turns out that it isn't hard to get five seats in Barnsley and Rotherham, so Stocksbridge & Upper Don can come into a Sheffield seat. In fact, St. Helens ward in Barnsley is small enough that removing it takes my previous Barnsley East from just below the upper limit to just above the lower limit, but also big enough (unlike, say, Worsbrough) adding it to Barnsley West allows that to lose Stocksbridge.

Going for five seats within Sheffield suggests one 6 ward seat and four five and a bit ward seats. The 6 ward seat should probably be the current Heeley plus Manor Castle, which just squeezes in below the lower limit.

The Ecclesfield wards are going to have to come into a successor to the current Brightside & Hillsborough, and the obvious thing is for it to lose Hillsborough and part of one other ward (which can be the Brightside area as before). At this point, there's a danger that someone thinks that the easiest thing to do is to add Stocksbridge & Upper Don to Hallam and Hillsborough to Central, then splitting one current Hallam ward (probably Crookes) between the two. This is bad: the revised "Hallam" is an incoherent straggly thing with poor communication links and little in common between its northern and southern areas. A better idea, but involving more change, is what J.G.Harston is suggesting: a revived Sheffield Hillsborough (S&UD, Stannington, Hillsborough, Walkley, plus some bits to the south) and allowing Hallam to expand towards the centre to replace its lost territory to the north and get up to quota.



Doncaster North & Moorthorpe 75,064
Doncaster Central 72,729
Don Valley 75,835
Rother Valley 73,511
Rotherham 71,116
Wentworth & Wombwell 72,925
Barnsley East 71,180
Barnsley West & Penistone 71,101
Sheffield Burngreave & Ecclesfield 82,218-x
Sheffield Attercliffe & Brightside 66,987+x
Sheffield Ecclesall 82,241-y (NB part of core "Hallam" is in Crookes IMO)
Sheffield Hillsborough 71,618+y
Sheffield Heeley 78,366

x=Brightside/Wincobank, y=Crookesmoor/Commonside/Birkendale

In Sheffield, this is really very similar to J.G.Harston 's "old wards" version. (I'm waiting to see official electorates for the new wards before thinking about whether it's worth trying to use them instead.)

[East Anglian Lefty]

I've been having a look at trying to do West Yorkshire minus Wakefield without ward-splitting, and finding it hellishly difficult. Ward sizes in most of the boroughs, but particularly in Calderdale and Leeds, are extremely unhelpful, because you tend to end up with most of your seats in those boroughs parked at the upper end of the size distribution, which leaves you a ward short at the end. I'm sure there are solutions, but I'm already crossing borders willy-nilly as it is and the one solution I haven't yet tried is to split Keighley, which is just a gratuitously bad idea.

YL, you mentioned that it's possible to do Wakefield with two 7-warders and one 6-warder. Is it possible to do it with one 7-warder and 2 6-warders? I'm not keen on the prospect, because I'd like to leave Dewsbury unchanged if at all possible, but that might be one way out of the problem. Alternatively, if the Keighley seat strays just over the border into Skipton, it might be possible to sort that without inflicting knock-on changes in North Yorkshire.

[East Anglian Lefty]

I think I've cracked it, without needing to use North Yorkshire. The arrangement isn't lovely and won't win me friends in Bingley or Halifax (but that's alright, I don't know anybody in Halifax, and I don't know anybody who cares about parliamentary boundaries in Bingley), but it just about works:



Elmet & Rothwell unchanged
Batley & Morley 74,680
Leeds Central 77,012
Leeds East 76,213
Leeds North 74,883
Leeds North West & Bingley 78,131
Leeds West & Pudsey 78,232
Bradford East & Calverley 72,184
Bradford South 72,909
Bradford West & Shipley 75,008
Keighley 76,636
Halifax West 72,353
Halifax East 76,014
Spen Valley 71,297
Huddersfield 73,241
Colne Valley 77,130
Dewsbury unchanged

Lets the shouts of outrage commence...

[krollo]

Mar 2, 2016 11:50:11 GMT East Anglian Lefty said:

I don't know anybody who cares about parliamentary boundaries in Bingley

Well now you do...

I've always thought that the Bradford area, notwithstanding electorate numbers, would most naturally divide into two core Bradford seats, a sort of Aire Valley seat from the northern wards of Bradford proper up to Keighley, something like Wharfe Valley from Ilkley to Otley, and then linking Pudsey with some of Bradford's eastern wards, with the remainder going into Valleys of Worth and Calder or suchlike. It crosses council boundaries a lot, but it's a fair bit better than having the utterly different towns of Keighley and Ilkley sharing an MP (which is essentially a local joke). Ilkley looks much more towards to what is now Leeds NW than it ever will to Bradford; Keighley through to Shipley is almost a continuous urban area; again, much of what is optimistically called Bingley Rural has closer ties to Calderdale than to the Aire Valley. 

However, electorate sizes make this difficult, in northern Leeds especially, and even falls in areas like Headingley will make it difficult for anything like this to work due to the knock-on effects. West Yorkshire constituencies must largely be constructed by desperately trying to get Kirklees and Leeds to work, while the other areas get more or less what they are given. 

[East Anglian Lefty]

I didn't draw the map with any particular partisan intent (except insofar as trying to keep similar areas togethers naturally has that), but it struck me that it might actually be quite effective for Labour. So I ran it through Electoral Calculus. Their notionals are famously crap, but here's what they produce (unchanged seats excluded, all majorities Labour/Conservative unless otherwise stated):

Batley & Morley: Lab maj. 3377
Leeds Central: Lab maj. 13234
Leeds East: Lab maj. 15166
Leeds North: Lab maj. 8907
Leeds North West & Bingley: Con maj. 5045 over Lab, 6452 over LD
Leeds West & Pudsey: Lab maj. 6760
Bradford East & Calverley: Lab maj. 5498 over Con, 7283 over LD
Bradford South: Lab maj. 14915
Bradford West & Shipley: Lab maj. 4380
Keighley: Con maj. 6888
Halifax West: Lab maj. 1835
Halifax East: Con maj. 11583
Spen Valley: Lab maj. 1835
Huddersfield: Lab maj. 2096
Colne Valley: Con maj. 3033

I have to say I'm rather sceptical about the Halifax numbers, but that would break down as 12 seats for Labour (including Dewsbury) and 5 seats for the Conservatives (including Elmet & Rothwell). So theoretically, Lab +1, Con -1.5, LD -1.

[Deleted]

Bradford West & Shipley: Lab maj. 4380

oh bloody hell...  

[J.G.Harston]

Mar 2, 2016 8:22:53 GMT YL said:

In Sheffield, this is really very similar to J.G.Harston 's "old wards" version. (I'm waiting to see official electorates for the new wards before thinking about whether it's worth trying to use them instead.)

I'd be happy with that, it's the same as my combined JGH/DrJordan model with Brightside swapped for Burngreave. I'd prefer the Brightside/Burngreave swap as then the northern seat can keep the name 'Brightside' (and I want to rename South-East back to Attercliffe....), but if it comes down to that map with the only decision being whether to split Burngreave or Wincobank I'd be happy.

On the new wards, the numbers would be this (I'm assuming I can draw a line through your Wincobank to end up the same numbers as my Burngreave):
Attercliffe: 74,370 (-0.5%)  (Mosborough, Birley, Beighton, Woodhouse, Bramley Hill-from-Richmond, Darnall, Wincobank-from-Shiregreen)
Brightside: 76,580 (+2.4%) (West Ecc, East Ecc, Southey, Firth Park, Burngreave, Shiregreen-without-Wincobank)
Hillsborough: 75,750 (+1.3%) (Stocksbridge, Stannington, Hillsborough, Crookes, Walkley, North Broomhill)
Hallam: 74,964 (+0.3%) (Dore/Totley, Fulwood, Ecclesall, Nether Edge, CityCentre, South Broomhill)
Heeley: 77,463 (+3.6%) (Beauchief, Graves, Gleadless, Arbourthorne, ManorCastle, Richmond-without-Bramley Hill)
These numbers are based the ward electorates in the polling district review went (that was scheduled to go) to Full Council today.

[David Ashforth]

Mar 2, 2016 21:03:16 GMT J.G.Harston said:

...These numbers are based the ward electorates in the polling district review went (that was scheduled to go) to Full Council today.

If anyone is interested here is the Polling District Review document.
sheffielddemocracy.moderngov.co.uk/documents/s21667/Polling%20District%20and%20Polling%20Place%20Review.pdf