Boundary Review without Cross-County Seats

[East Anglian Lefty]

Arguably the most distinctive change in the new boundary review is that it will have seats which are made up of multiple counties or county-equivalents. This is also one of the elements that has attracted the most opposition (see for example the objections to a Devonwall seat in the previous failed reviews.) Without the disruption that causes, it's considerably more difficult to argue against the principle of more equal electorate sizes.

So what if we maintained the previous bar on crossing county boundaries, but kept a 5% maximum deviation? Well, the first thing to say is that that's not possible with a uniform national quota, because for instance Cumbria's notional entitlement is 5.31 constituencies, which is a 6% average deviation with 5 seats or a 12% average deviation with 6 seats.

So what if we instead substitute a single national quota with a quota for each county? This way you would have a greater deviation between counties, but generally you would still minimise the deviation between neighbouring seats and the variation from county to county would be random rather than systematic.

Assigning seats to English counties* (defined as the Fifth Periodic Review did - essentially the counties as they were in 1985, prior to the abolition of metropolitan counties, the abolition of Avon et al. and the subsequent creations of unitaries) produces the following set of quotas:

Avon: 11.64
Bedfordshire: 6.37
Berkshire: 8.65
Buckinghamshire: 8.00
Cambridgeshire: 8.06
Cheshire: 11.27
Cornwall: 5.97
County Durham: 6.43
Cumbria: 5.31
Derbyshire: 10.78
Devon: 12.52
Dorset: 8.00
East Sussex: 8.40
Essex: 18.38
Gloucestershire: 6.59
Greater Manchester: 27.26
Hampshire: 18.44
Hereford & Worcestershire: 8.03
Hertfordshire: 11.47
Humberside: 9.32
Kent: 18.05
Lancashire: 15.18
Leicestershire & Rutland: 10.69
Lincolnshire: 7.52
Merseyside: 14.31
Norfolk: 9.21
North Yorkshire: 8.46
Northamptonshire: 7.22
Northumberland: 3.40
Nottinghamshire: 11.22
Oxfordshire: 6.81
Shropshire: 5.12
Somerset: 5.80
South Yorkshire: 13.73
Staffordshire: 11.35
Suffolk: 7.60
Surrey: 11.72
Teesside: 5.67
Tyne & Wear: 11.11
Warwickshire: 5.89
West Midlands: 26.41
West Sussex: 8.81
West Yorkshire: 22.54
Wiltshire: 7.27

I've then rounded those quotas to the nearest whole number (I considered using harmonic means, as was the case in previous boundary reviews, but this tended to increase the number of seats which obviously was the opposite of the direction of travel the Fifth Review sought.

I've then used these county quotas to draw seats for those counties, noting the extent to which they've differed from those we're getting, in an attempt to understand whether this would tend to produce worse or better seats.

In the rest of this thread, I'm going to post maps of what I came up with. Comments and criticisms of the methodology are obviously more than welcome.

*Neither Northern Ireland nor Scotland used county-equivalents in drawing seats for the Fifth Review, whilst the Welsh preserved counties are rather small and using them tends to militate against electoral equality, so I've ignored them for the purposes of the thought experiment. I've also excluded London, because if you treat Greater London as a county equivalent then it's not an interesting exercise, and if you instead treat the boroughs as county-equivalents then you a) aren't following the practice of the Fifth Review and b) it's an awful mess for similar reasons to Wales.

[East Anglian Lefty]

Starting with the North East, as it's the smallest region, Northumberland is entitled to three seats (a reduction from four at present); Tyne & Wear to eleven (a reduction from twelve at present); County Durham (including Darlington) is entitled to six (a reduction from seven at present); and the former Teesside is entitled to six (as at present). In both Northumberland and Durham, the extra seat is fairly narrowly missed, so the surviving seats are large, whilst in Teesside the six seats are small (as at present, but with better electoral equality between them.)

Northumberland

An entitlement to 3.39 constituencies means that the county gets three seats, with the permissible range of electorates being 78981-87295. As the seat in the middle, Wansbeck is the one that gets chopped up. Moving the boundary of Blyth Valley north to the River Wansbeck isn't especially unreasonable; putting Morpeth in with Hexham isn't great but there are at least links south along the A1; putting Ashington in with Berwick is obviously absurd but with the rules I set it was the natural combination.



Berwick & Ashington 81480
Hexham & Morpeth 82010
South East Northumberland 85924

Tyne & Wear

An entitlement to 11.22 constituencies means eleven seats, with a range of electorates slightly above what our rules permit but not by much (70425-77839). Without a seat crossing into Northumberland it's necessary to have a seat crossing the Tyne, and I couldn't find a nice way to do this in the centre of Gateshead so I've done it further west. The bigger issue is that that seat needs three Gateshead wards, and so far as I can tell there isn't a direct road link between Whickham South & Sunniside ward and either Chopwell & Rowlands Gill or Winlaton & High Spen, which means the third ward can't be Blaydon (otherwise the obvious candidate.) So the cross-river seat is fairly horrible, but I'm not unhappy with the other ten.



Sunderland Central 72688 (unchanged)
Houghton & Sunderland South 76883 (matches BCE final recommendations)
Washington 71775 (matches BCE final recommendations)
South Shields 77038
Jarrow & Gateshead East 76701
Gateshead West & Whickham 72136
Newcastle West & Ryton 75668
Newcastle Central 73082 (suggest a better name for the previous seat and I'll happily call this Newcastle West)
Newcastle East 71044
North Tyneside 71054
Tynemouth 77382 (unchanged)

County Durham

6.43 seats means six seats, all of which need to be on the large size as the permissible range of electorates is 74738-82605. As always with mapping Durham, the challenge is what to do with the City and once you solve that everything else draws itself. Here I've ended up giving it quite a tight haircut to the west but adding Spennymoor to make up the numbers; with nicer ward boundaries and more convenient ward electorates I'd probably have made the opposite choice.



Darlington 78498 (co-extensive with the local authority)
Bishop Auckland 75162
East Durham 81049
City of Durham 79199
North West Durham 77966
North Durham 80157

Teesside

Whereas everywhere else in the region has larger than average seats, Teesside has 5.67 quotas so it's six are on the small size, with a permissible range of 65883-72818. For the most part it's a least change map (only two wards moved after realignment to new ward boundaries); the major flaw is the very ugly boundary in Middlesbrough which could be improved if you're willing to move another couple of wards.



Hartlepool 71228 (unchanged)
Stockton North 71804
Stockton South 71977
Middlesbrough 68094
Middlesbrough South & East Cleveland 66774
Redcar 66256 (unchanged)

[Khunanup]

Nov 8, 2023 21:27:26 GMT East Anglian Lefty said:

Arguably the most distinctive change in the new boundary review is that it will have seats which are made up of multiple counties or county-equivalents. This is also one of the elements that has attracted the most opposition (see for example the objections to a Devonwall seat in the previous failed reviews.) Without the disruption that causes, it's considerably more difficult to argue against the principle of more equal electorate sizes.

So what if we maintained the previous bar on crossing county boundaries, but kept a 5% maximum deviation? Well, the first thing to say is that that's not possible with a uniform national quota, because for instance Cumbria's notional entitlement is 5.31 constituencies, which is a 6% average deviation with 5 seats or a 12% average deviation with 6 seats.

So what if we instead substitute a single national quota with a quota for each county? This way you would have a greater deviation between counties, but generally you would still minimise the deviation between neighbouring seats and the variation from county to county would be random rather than systematic.

Assigning seats to English counties* (defined as the Fifth Periodic Review did - essentially the counties as they were in 1985, prior to the abolition of metropolitan counties, the abolition of Avon et al. and the subsequent creations of unitaries) produces the following set of quotas:

Avon: 11.64
Bedfordshire: 6.37
Berkshire: 8.65
Buckinghamshire: 8.00
Cambridgeshire: 8.06
Cheshire: 11.27
Cornwall: 5.97
County Durham: 6.43
Cumbria: 5.31
Derbyshire: 10.78
Devon: 12.52
Dorset: 8.00
East Sussex: 8.40
Essex: 18.38
Gloucestershire: 6.59
Greater Manchester: 27.26
Hampshire: 18.44
Hereford & Worcestershire: 8.03
Hertfordshire: 11.47
Humberside: 9.32
Kent: 18.05
Lancashire: 15.18
Leicestershire & Rutland: 10.69
Lincolnshire: 7.52
Merseyside: 14.31
Norfolk: 9.21
North Yorkshire: 8.46
Northamptonshire: 7.22
Northumberland: 3.40
Nottinghamshire: 11.22
Oxfordshire: 6.81
Shropshire: 5.12
Somerset: 5.80
South Yorkshire: 13.73
Staffordshire: 11.35
Suffolk: 7.60
Surrey: 11.72
Teesside: 5.67
Tyne & Wear: 11.11
Warwickshire: 5.89
West Midlands: 26.41
West Sussex: 8.81
West Yorkshire: 22.54
Wiltshire: 7.27

I've then rounded those quotas to the nearest whole number (I considered using harmonic means, as was the case in previous boundary reviews, but this tended to increase the number of seats which obviously was the opposite of the direction of travel the Fifth Review sought.

I've then used these county quotas to draw seats for those counties, noting the extent to which they've differed from those we're getting, in an attempt to understand whether this would tend to produce worse or better seats.

In the rest of this thread, I'm going to post maps of what I came up with. Comments and criticisms of the methodology are obviously more than welcome.

*Neither Northern Ireland nor Scotland used county-equivalents in drawing seats for the Fifth Review, whilst the Welsh preserved counties are rather small and using them tends to militate against electoral equality, so I've ignored them for the purposes of the thought experiment. I've also excluded London, because if you treat Greater London as a county equivalent then it's not an interesting exercise, and if you instead treat the boroughs as county-equivalents then you a) aren't following the practice of the Fifth Review and b) it's an awful mess for similar reasons to Wales.

You've forgotten Isle of Wight.

[YL]

This is fairly close to my own preferred approach so I'll be interested to see the results.

If legislating for this approach the counties would have to be defined. Personally I would not go back to the 1974 counties but would base them on the ceremonial counties with mergers to deal with the small ones: Rutland and the City of London most obviously, but I'd also consider pairing Herefordshire with Worcestershire, Northumberland with Tyne & Wear and maybe Bristol with Gloucestershire. I'd also suggest fixing the Mersey as the boundary below Warrington, i.e. putting Halton north of the river with Merseyside and the Wirral with Cheshire. Overall I think that would give better boundaries, avoiding seats crossing the Mersey estuary and any monstrosities trying to curve round the head of the Humber.

[YL]

Nov 8, 2023 23:34:09 GMT Khunanup said:

You've forgotten Isle of Wight.

It would have presumably either remained a single seat, as it did in the Fifth Review, or ended up being split into two in exactly the same way as it actually has been.

[East Anglian Lefty]

Nov 8, 2023 23:34:09 GMT Khunanup said:

You've forgotten Isle of Wight.

Forgot to mention it, but I did take it into consideration - I just didn't bother mapping it for the same "it's not very interesting" reason.

Nov 9, 2023 7:36:10 GMT YL said:

This is fairly close to my own preferred approach so I'll be interested to see the results.

If legislating for this approach the counties would have to be defined. Personally I would not go back to the 1974 counties but would base them on the ceremonial counties with mergers to deal with the small ones: Rutland and the City of London most obviously, but I'd also consider pairing Herefordshire with Worcestershire, Northumberland with Tyne & Wear and maybe Bristol with Gloucestershire. I'd also suggest fixing the Mersey as the boundary below Warrington, i.e. putting Halton north of the river with Merseyside and the Wirral with Cheshire. Overall I think that would give better boundaries, avoiding seats crossing the Mersey estuary and any monstrosities trying to curve round the head of the Humber.

In this case I did treat Herefordshire and Worcestershire as a unit, but in practice the numbers work perfectly for the former to have two and the latter to have six.

As treating Merseyside as a unit does force an awful Mersey Tunnel seat, I also mapped it putting Wirral in with Cheshire (but not moving Halton the other way.) That produces a much better arrangement of seats, but it changes the allocation from fourteen for Merseyside and eleven for Cheshire to fifteen for Cheshire + Wirral and eleven for Lancastrian Merseyside, so a net increase of one seat overall.

The other issue with ceremonial counties is the boundary between Durham and North Yorkshire, which split the Borough of Stockton in two. I don't think that's necessarily problematic, but worth flagging up.

[East Anglian Lefty]

Moving on to the North West:

Cumbria

Cumbria is entitled to 5.31 constituencies, so that's five large seats with a range from 74046 to 81841. In theory that means Allerdale could stand alone, but in practice the geography rules that out. So instead you get one seat made up of Copeland district and Workington, another with all but one ward of Carlisle, an expanded Barrow, a Penrith & Solway seat and one which tries to stick as close as possible to the historic borders of Westmorland.



Carlisle 75868 (matches BCE final recommendations)
Penrith & Solway 78924
Whitehaven & Workington 78432
Westmorland & Lonsdale 76742
Barrow & Furness 79751

Lancashire

Lancashire's entitlement is 15.18, meaning fifteen seats, which is one fewer than it presently has. The acceptable range is 70556-77983. In practice, most seats can be preserved with reasonably minor changes, but those in the middle of the county are intruded into from multiple directions and Wyre & Preston North is the abolished seat. A few of these seats are definite improvements over the ones we're getting, a few are arguably slightly worse.



Morecambe & Lunesdale 71296
Lancaster & Wyre 72954
Blackpool North & Fleetwood 71640
Blackpool South 76071 (matches BCE final recommendations)
Fylde 70833
Preston 72946 (matches BCE final recommendations)
Mid Lancashire 75150
South Ribble 75953
West Lancashire 73652 (unchanged)
Chorley 74568 (unchanged after realignment)
Blackburn 76323
Rossendale & Darwen 77369
Hyndburn 73684
Burnley & Brierfield 75436
Clitheroe & Colne 76168

Greater Manchester

GM has an entitlement to 27.26 seats, meaning 27 constituencies with a permissible range of 70385-77794. This is no change from the present entitlement. Ten of those seats can be left unchanged, but there's more disruption around Manchester itself, which effectively gains a seat at the expense of the eastern boroughs. I personally think this is superior to what we got (and the slightly different electorate range means I didn't need to split any wards.)



Wigan 75607 (unchanged)
Makerfield 77400 (unchanged)
Leigh 77416 (unchanged)
Bolton West 73149 (unchanged)
Bolton South East & Walkden 76199
Bolton North East 76808
Bury North 75652
Bury South 76935
Salford 71657
Worsley & Eccles 76306
Stretford & Urmston 73212 (unchanged)
Altrincham & Sale West 73934 (unchanged)
Wythenshawe & Sale East 76971 (unchanged)
Manchester Withington 71614 (matches BCE final recommendations)
Manchester Rusholme 70692 (matches BCE final recommendations)
Gorton & Denton 74306 (matches BCE final recommendations)
Stockport 74769 (matches BCE final recommendations)
Cheadle 73775 (unchanged)
Hazel Grove 72941 (matches BCE final recommendations)
Stalybridge & Hyde 73028 (unchanged)
Ashton under Lyne 72278 (matches BCE final recommendations)
Manchester Central 71923
Blackley & Middleton 75653
Heywood & Royton 72853
Rochdale 71697 (matches BCE final recommendations)
Oldham East & Saddleworth 72997 (unchanged)
Oldham West 73746

I'm going to split this region into two, since as I've already said I considered two different ways of dealing with the Wirral, placing it with either Cheshire or the rest of Merseyside, so that means four more maps and this post is long enough as it is.

[johnloony]

Nov 8, 2023 21:27:26 GMT East Anglian Lefty said:

:Avon: 11.64
Bedfordshire: 6.37
Berkshire: 8.65
Buckinghamshire: 8.00
Cambridgeshire: 8.06
Cheshire: 11.27
Cornwall: 5.97
County Durham: 6.43
Cumbria: 5.31
Derbyshire: 10.78
Devon: 12.52
Dorset: 8.00
East Sussex: 8.40
Essex: 18.38
Gloucestershire: 6.59
Greater Manchester: 27.26
Hampshire: 18.44
Hereford & Worcestershire: 8.03
Hertfordshire: 11.47
Humberside: 9.32
Kent: 18.05
Lancashire: 15.18
Leicestershire & Rutland: 10.69
Lincolnshire: 7.52
Merseyside: 14.31
Norfolk: 9.21
North Yorkshire: 8.46
Northamptonshire: 7.22
Northumberland: 3.40
Nottinghamshire: 11.22
Oxfordshire: 6.81
Shropshire: 5.12
Somerset: 5.80
South Yorkshire: 13.73
Staffordshire: 11.35
Suffolk: 7.60
Surrey: 11.72
Teesside: 5.67
Tyne & Wear: 11.11
Warwickshire: 5.89
West Midlands: 26.41
West Sussex: 8.81
West Yorkshire: 22.54
Wiltshire: 7.27.

One method of
(1) keeping the strict 5% limit
and
(2) keeping all county boundaries as constituency boundaries

would be to

(1) have a whole number of constituencies in each county, plus a leftover bit
(2) combine the leftover bits to make the remaining number of constituencies
while also making sure that
(3) no two adjacent leftover bits are in the same constituency.

Frig zample:

Avon has 11 constituencies, plus a chunk which has 0.64 of a quota of electorate
Bedfordshire has 6 constituencies, plus a chunk of 0.37
The 0.64 plus the 0.37 combine to create a constituency called “Avon and Bedfordshire Leftovers”.
The 0.37 of a quota of electorate which is leftover after you have drawn the boundaries of the 6 other Bedfordshire constituencies doesn’t even have to be in one single lump. It could be multiple exclaves and pockets all over the place. A bit like when they cook a pork pie and it ends up with pockets of jelly in between the meaty bit and the crusty bit.

Just get all the fractional numbers from the above list (the two decimal places after the decimal point) and combine them in combinations each of which adds up to something between 0.95 and 1.05.

[nyx]

I think the most reasonable way to handle the Wirral would be to keep the rule of using 1985 counties (i.e. Wirral is in Merseyside), but Wirral's geographically disjointed status from the rest of Merseyside would make it deserve island status like Anglesey does.

i.e. 3 seats for Wirral and 11 for the rest of Merseyside

[Khunanup]

Nov 10, 2023 2:09:40 GMT nyx said:

I think the most reasonable way to handle the Wirral would be to keep the rule of using 1985 counties (i.e. Wirral is in Merseyside), but Wirral's geographically disjointed status from the rest of Merseyside would make it deserve island status like Anglesey does.

i.e. 3 seats for Wirral and 11 for the rest of Merseyside

Look at a map of The Wirral and tell me why divorcing the Met Borough from the rest of the peninsula makes any sense for boundary purposes (and therefore leaving it with three enormous seats) and more to the point, try to figure out why South Wirral (ie the old Ellesmere Port and Neston borough) was never included in the Met Borough in the first place.

[islington]

Nov 8, 2023 21:27:26 GMT East Anglian Lefty said:

Arguably the most distinctive change in the new boundary review is that it will have seats which are made up of multiple counties or county-equivalents. This is also one of the elements that has attracted the most opposition (see for example the objections to a Devonwall seat in the previous failed reviews.) Without the disruption that causes, it's considerably more difficult to argue against the principle of more equal electorate sizes.

So what if we maintained the previous bar on crossing county boundaries, but kept a 5% maximum deviation? Well, the first thing to say is that that's not possible with a uniform national quota, because for instance Cumbria's notional entitlement is 5.31 constituencies, which is a 6% average deviation with 5 seats or a 12% average deviation with 6 seats.

So what if we instead substitute a single national quota with a quota for each county? This way you would have a greater deviation between counties, but generally you would still minimise the deviation between neighbouring seats and the variation from county to county would be random rather than systematic.

Assigning seats to English counties* (defined as the Fifth Periodic Review did - essentially the counties as they were in 1985, prior to the abolition of metropolitan counties, the abolition of Avon et al. and the subsequent creations of unitaries) produces the following set of quotas:

Avon: 11.64
Bedfordshire: 6.37
Berkshire: 8.65
Buckinghamshire: 8.00
Cambridgeshire: 8.06
Cheshire: 11.27
Cornwall: 5.97
County Durham: 6.43
Cumbria: 5.31
Derbyshire: 10.78
Devon: 12.52
Dorset: 8.00
East Sussex: 8.40
Essex: 18.38
Gloucestershire: 6.59
Greater Manchester: 27.26
Hampshire: 18.44
Hereford & Worcestershire: 8.03
Hertfordshire: 11.47
Humberside: 9.32
Kent: 18.05
Lancashire: 15.18
Leicestershire & Rutland: 10.69
Lincolnshire: 7.52
Merseyside: 14.31
Norfolk: 9.21
North Yorkshire: 8.46
Northamptonshire: 7.22
Northumberland: 3.40
Nottinghamshire: 11.22
Oxfordshire: 6.81
Shropshire: 5.12
Somerset: 5.80
South Yorkshire: 13.73
Staffordshire: 11.35
Suffolk: 7.60
Surrey: 11.72
Teesside: 5.67
Tyne & Wear: 11.11
Warwickshire: 5.89
West Midlands: 26.41
West Sussex: 8.81
West Yorkshire: 22.54
Wiltshire: 7.27

I've then rounded those quotas to the nearest whole number (I considered using harmonic means, as was the case in previous boundary reviews, but this tended to increase the number of seats which obviously was the opposite of the direction of travel the Fifth Review sought.

I've then used these county quotas to draw seats for those counties, noting the extent to which they've differed from those we're getting, in an attempt to understand whether this would tend to produce worse or better seats.

In the rest of this thread, I'm going to post maps of what I came up with. Comments and criticisms of the methodology are obviously more than welcome.

*Neither Northern Ireland nor Scotland used county-equivalents in drawing seats for the Fifth Review, whilst the Welsh preserved counties are rather small and using them tends to militate against electoral equality, so I've ignored them for the purposes of the thought experiment. I've also excluded London, because if you treat Greater London as a county equivalent then it's not an interesting exercise, and if you instead treat the boroughs as county-equivalents then you a) aren't following the practice of the Fifth Review and b) it's an awful mess for similar reasons to Wales.

Let me accept the invitation extended by the words in bold.

Although this is an attractive proposal in some ways, and I share the evident aversion to cross-county seats, I can see two problems. One is relatively easy to overcome; the other is more fundamental.

The easy problem to solve is that the chosen method of distribution does not guarantee the correct number of seats. The counties listed, each rounded up or down to the nearest whole number, sum to 465 seats (if my addition is correct). London to the nearest whole number is 76 and the IoW is actually slightly nearer two than one, so on these particular numbers, and adding in Scotland (57), Wales (32) and NI (18) we get 650, which happens to be correct. But it's easy to see how only a small shift in numbers in one or more of the marginal counties might have given us 649, or 651, or even a wider departure from 650.

This is readily rectified. The simplest way is to say that even if 650 is the target, we accept that we may narrowly miss it and we are prepared to accept an outcome departing from 650 by a seat or two. The second method, if we decide that we want to hit 650 exactly, is to apportion the seats by a priority value system (preferably the Sainte-Lague (aka Webster) formula) because such a method guarantees that we can hit 650 or whatever other preset number we happen to choose.

My more fundamental concern, however, is that apportioning seats as proposed requires counties to be rigidly defined at the outset. This isn't a problem where the county is clear-cut; but it will be an issue in cases where counties are less well defined because of successive waves of piecemeal, ill-thought-out local government changes. 

For instance, this scheme recreates the unloved county of Avon. This means that if N Somerset and B&NES, together, cannot be formed into a whole number of seats within 5% of the Avon quota, then the boundary has to be crossed into either Bristol or S Glos even though it might work better in terms of numbers on the ground, and would almost certainly be more acceptable to local opinion, to treat these two UAs with Somerset. And similar comments can be made elsewhere: currently, I'd suggest, we can make a pragmatic choice about whether Halton and Warrington UAs should be treated with Cheshire, Merseyside, or Gtr Manchester; but under this scheme, those UAs are locked into Cheshire even if it is obvious that an alternative approach would have worked better.

This is already a problem with the existing regional system. For instance, we are forced to treat N Lincs and NE Lincs with other parts of Y&H when it would probably be more natural to include them with Lincolnshire; and MK has to go with Bucks even if it might work better with Beds or Nhants. But problems of this kind will be far more widespread if every county boundary, no matter how doubtful or ambiguous, becomes an uncrossable line.

And I agree with the point made upthread that we really, really don't want a system that forces the creation of a 'Mersey Tunnel' seat.

So while this is an interesting thought experiment, for my part I'd prefer to stick to the current system.

[nyx]

Nov 10, 2023 2:53:27 GMT Khunanup said:

Nov 10, 2023 2:09:40 GMT nyx said:

I think the most reasonable way to handle the Wirral would be to keep the rule of using 1985 counties (i.e. Wirral is in Merseyside), but Wirral's geographically disjointed status from the rest of Merseyside would make it deserve island status like Anglesey does.

i.e. 3 seats for Wirral and 11 for the rest of Merseyside

Look at a map of The Wirral and tell me why divorcing the Met Borough from the rest of the peninsula makes any sense for boundary purposes (and therefore leaving it with three enormous seats) and more to the point, try to figure out why South Wirral (ie the old Ellesmere Port and Neston borough) was never included in the Met Borough in the first place.

I had assumed that the whole point of this exercise was to reduce the number of seats crossing multiple authority areas. The Wirral is hardly the only place where it's easy to draw a natural cross boundary seat; plenty of the results of the actual review (e.g. Hitchin) are natural enough.

The rules behind this thread already mean the existence of some larger seats (see Northumberland).

[doktorb🏳️‍🌈🏳️‍⚧️]

"Mid Lancashire 75150"

Decent seat. Terrible name. "Ribble Valley".

[Khunanup]

Nov 10, 2023 12:28:04 GMT nyx said:

Nov 10, 2023 2:53:27 GMT Khunanup said:

Look at a map of The Wirral and tell me why divorcing the Met Borough from the rest of the peninsula makes any sense for boundary purposes (and therefore leaving it with three enormous seats) and more to the point, try to figure out why South Wirral (ie the old Ellesmere Port and Neston borough) was never included in the Met Borough in the first place.

I had assumed that the whole point of this exercise was to reduce the number of seats crossing multiple authority areas. The Wirral is hardly the only place where it's easy to draw a natural cross boundary seat; plenty of the results of the actual review (e.g. Hitchin) are natural enough.

The rules behind this thread already mean the existence of some larger seats (see Northumberland).

The exercise has nothing to do with not crossing local authority areas, it's to stick with the c. '85 counties and not cross those boundaries (thus the thread originator mentioning that if he does that rigidly you're going to end up with a cross Mersey seat).

The problem with The Wirral is that the met boundary as drawn is completely bonkers for a geographically confined and unique area.

[East Anglian Lefty]

North-West part 2:

As mentioned, there are obvious reasons why sticking to the old county boundaries in the Liverpool area produces awful seats. Nevertheless, for the purposes of the thought experiment, Merseyside on its own has an entitlement to 14.31 constituencies, so if you treat it alone then that's 14 seats, one fewer than it currently has. The range is 71246-78746.

This works really incredibly badly.

The two St Helens seats can be left unchanged, but Knowsley is over-sized, meaning you need to split either Kirkby or Huyton. Sefton looks at first glance like it could remain unchanged, but in fact the higher upper limit means that both Sefton Central and Southport are too small. The former can be dealt with by a ward swap (albeit not a good one) but the latter requires a very precise ward chop (or to say that being within 5% of the national quota is fine if you're not within 5% of the local quota.)

Compared to that, Liverpool gets off relatively easy, though the Wavertree seat is comprehensively sliced and diced.

It goes without saying that the Mersey Tunnel seat is bad. I only used one Wirral ward in it, as it leaves 21 remaining for the other three seats, and I eventually decided that Seacombe was the least worst option, as its boundaries with neighbouring wards are stronger than Birkenhead & Tranmere's. Unfortunately, numbers for the other three seats are very tight and you also need to stick Rock Ferry in with Wirral South to make the numbers work.

Overall, I think it is fair to say that the rules I'm working with work incredibly poorly here.



Southport 71037*
Sefton Central 71512*
Bootle 73767
Liverpool Walton 74929
Liverpool West Derby 74409
Mersey Tunnel 71355
Wallasey 77790
Birkenhead 78464
Wirral South 78248
Liverpool Mossley Hill 76327
Halewood & Woolton 77204
Knowsley 71724
St Helens South & Whiston 77099 (unchanged after realignment)
St Helens North 76082 (unchanged)

*Before ward splits - will have to be done at a sub-polling district level.

Cheshire

Cheshire (including Warrington and Halton) has an entitlement to 11.27 councillors, which equates to 11 seats and a permissible range of 71458-78980. That's the same as at present.

In general, this can be done without too much change (and indeed with a few seats unchanged.) The major weakness here is that Eddisbury stretches as far north as Frodsham, which isn't ideal but is hardly a disaster.



Warrington North 72350 (unchanged)
Warrington South 72286
Halton 76249 (unchanged after realignment)
Weaver Valley 75057
Tatton 77800
Macclesfield 75881 (unchanged)
Congleton 72770
Crewe & Nantwich 76236
Eddisbury 77665
City of Chester 74747
Ellesmere Port & Neston 76373



As I mentioned, the alternative option is to consider the Wirral as part of Cheshire for these purposes. Now admittedly this goes against the strictures I set myself, but it's not ridiculous on the face of it, given that the Mersey is a much more obvious barrier than whatever line the Wirral/CWaC boundary follows. It's easy to headcanon this if we imagine that for whatever reason when the 2011 bill passed they had been more willing to accept amendments to minimise local disruption and it's an excuse to draw more maps, which let's be honest is the main purpose of this.

'Lancastrian Merseyside'

Minus the Wirral, we're talking about an entitlement to 10.97 constituencies, so that's eleven seats pretty much exactly at the national average, which is the same amount those areas have at present. The range is 69546-76866.

And this works much more neatly. Sefton is entirely unchanged, every seat has a logical successor and the only downside is that Huyton gets split. Which isn't ideal but is a lot better than a Mersey Tunnel seat.



Southport 71037 (unchanged)
Sefton Central 70085 (unchanged)
Bootle 75194 (unchanged)
Liverpool Walton 74929
Liverpool West Derby 74409
Liverpool Riverside 70797
Liverpool Wavertree 74990
Garston & Huyton South 76311
Kirkby & Huyton North 71745
St Helens South & Whiston 69688
St Helens North 76082 (unchanged)

Cheshire & Wirral

This area has an entitlement to 14.61 constituencies, which rounds up to 15. The allowable range is 67899-75047. It's the same number of seats as at present, but in practice it means shifting about half a seat from the Wirral to Cheshire.

Unfortunately, Wirral's wards are still extremely unhelpful. Whilst most of Cheshire is quite easy to draw like this (and it's possible to eliminate one of the crossings of the CWaC/Cheshire East boundary), I found I needed to have two seats crossing the Wirral/CWaC border and still split Ellesmere Port in half. This could be eliminated with a ward split, but I'm trying as much as possible to avoid those (and if you did you'd have to split Bebington anyway).



Birkenhead 74456
Wallasey 69149
Wirral West 71273
Bebington & Ellesmere West 71202
Ellesmere Port East & Runcorn 74210
City of Chester 70992
Eddisbury 69470
Crewe & Nantwich 69054
Congleton 70920
Macclesfield 72174
Tatton 69167
Winsford & Northwich 72976
Halton 72415
Warrington South 72286
Warrington North 72350 (unchanged)

[YL]

The Lancastrian Merseyside map there is the same as the Merseyside one in the first option, which it obviously shouldn't be.

[East Anglian Lefty]

Nov 10, 2023 19:50:16 GMT YL said:

The Lancastrian Merseyside map there is the same as the Merseyside one in the first option, which it obviously shouldn't be.

Should be fixed now.

[John Chanin]

Perhaps we could solve the Wirral problem, which is a perennial thorn in boundary discussions, by ceding it to Wales, and let them deal with it.

[doktorb🏳️‍🌈🏳️‍⚧️]

Nov 10, 2023 20:06:54 GMT John Chanin said:

Perhaps we could solve the Wirral problem, which is a perennial thorn in boundary discussions, by ceding it to Wales, and let them deal with it.

""De Cilgwri a Gogledd Sir y Fflint", exactly. 

[East Anglian Lefty]

I'm going to skip Yorkshire & Humber for now and do the East Midlands instead.

Working north to south and east to west, we begin with Derbyshire. Here there's no point doing a map - the entitlement is 10.78, equating to eleven seats, and the easiest way to do that is exactly what we got from the BCE under the new rules.

Nottinghamshire

Notts has an entitlement to 11.22 seats, meaning it keeps the eleven it currently has. The target range is 71132-78620. That slightly different range allows Mansfield to remain unchanged, though on the flip side my map puts the bits of Hucknall west of the A611 in a different seat from the rest of the town.



Mansfield 77409 (unchanged)
Bassetlaw 77311
Sherwood 71252
Newark 74940
Rushcliffe 76171 (matches BCE)
Gedling 73482 (unchanged after realignment)
Nottingham East 78574
Nottingham South 72829
Nottingham North & Kimberley 74515 (matches BCE)
Broxtowe 72461 (matches BCE)
Ashfield 74694

Lincolnshire

Thus far we've only dealt with areas losing seats or holding steady, but Lincolnshire is the first gainer under these rules. With 7.52 quotas, it just qualifies for an eighth seat, with a target range of 65539-72437. Lincoln and South Holland are just slightly too small to stand alone, whilst East Lindsey is slightly too large. Making the necessary shuffles forces the creation of a seat that is essentially just Grantham, Sleaford and the minimum amount of countryside necessary to join the two.



Stamford 65826
South Holland 71363
Boston & Skegness 71680
Louth & Horncastle 71740
Gainsborough 66341
Lincoln 66856
North Hykeham 68678
Grantham & Sleaford 69420

Leicestershire & Rutland

The entitlement Leicestershire and Rutland is 10.69 counties, meaning eleven seats and a permissible range of 67770-74903. Under the current rules it would have been possible for Leicestershire and Rutland to stand alone, but the national quota limited your range for manouevre, whereas with a county quota you've got a bit more freedom. I'm quite pleased with how this one turned out, especially for the seats on the Leicester outskirts.



Rutland & Melton 69413
Harborough 69095
Charnwood 74852
Loughborough 68358
North West Leicestershire 73657
Hinckley & Bosworth 71505
Mid Leicestershire 68127
Oadby & Wigston 69376
Leicester South 72687
Leicester East 74785
Leicester West 72848 (matches BCE)

Northamptonshire

I'm considerably less happy with Northamptonshire, because the old county council boundaries really make it difficult to draw sensible seats. 7.22 quotas equates to 7 seats with a range of 71915-79485. You could draw a perfectly sensible least-change map using the old district wards with very few changes outside Northampton, or there are many options if you're willing to split a ward or two.

If you're not, then this is the best I could manage. I think there are perhaps three and a half good seats here?



Kettering 72777 (unchanged)
Corby 76706
South East Northants 78695
Wellingborough 76697
Northampton North 72175
Northampton South 75050
Daventry 77799