Sheffield Ward Review 2013

[J.G.Harston]

Moving over from the Elections thread (I hope I get this to work...)

Jan 3, 2013 18:05:21 GMT David Ashforth said:

The Local Government Boundary Commission for England is due to
commence the electoral review of Sheffield in the summer of 2013.
Sheffield was identified as one of the local authorities with an
electoral imbalance...
(...)
Alex Hinds
Review Assistant
Local Government Boundary Commission for England

Mar 5, 2013 19:47:34 GMT J.G.Harston said:

I've finished the polling district maps (at mdfs.net) (other than a bit of detail on Stannington and West Ecclesfield where I'll need to check individual street lists). I've also put together a summary city-wide map of electoral imbalances:



The size of the numbers indicates the strength of the disparity. Something that is immediately obvious is that almost all of the city is slightly under average - only mathematically possible because of the three wards that are so far above average, Darnall 10.9%, Burngreave 11.3%, Central 44.5%!!!

If the council remains with 28 wards (3*28=84 members) the majority of wards could remain, but I can't see any way to avoid wholesale change around the city centre.

* As I predicted in 2002 Darnall has increased enough to allow it to give the Handsworth salient back to Handsworth/Woodseats, but unfortunately, Woodseats is too small to absorb it. It would just swap the +10%/-6% to -6%/+10%. However, the Orgreave Lane/Rotherham Road/Retford Road triangle has about 1000 electorate, and is just the right size to transfer over to balance both at about 14,900.

* Shirecliffe and Busk Meadow could be transfered back out of Burngreave - it would drop Burngreave down to about 13,500 (down from +11.3% to -7%), pushing Southey up to 15200 (up from -3.8% to +5%) and Firth Park up to 15,600 (up from +0.5% to +8%).

Metropolitan authorities are no longer bound to use wards with exactly three members, they can now use any number up to three. I strongly discourage single-member wards as having a group of local members makes for better representation and mutual support - even if from different parties. I spent 8 of my 11 years on the council in a mixed ward.

I argued in 2002 that in the rural parished areas of Sheffield being forced to use 3-member wards forces you to create overly-artificial combinations of communities, and that the option of two-member wards would make for better representation. I put together a model to have six 2-member wards instead of four 3-member wards which would gather up groups of communities a lot better:

mdfs.net/User/JGH/Docs/Politics/WardReview/DraftPER/Rural.gif

This would be a good model to implement in the current review.

When drawing up the Fulwood/Broomhill boundary in 2002 we had great trouble as the line ran through the halls of residence. Moving the line by one foot changed the electorate by 1000! It would be better to revert to a model with a "horzontal" Broomhill ward that put all the halls in the same ward. I drew up a last-minute recommendation to do this, but it was too late for the 2002 review:

mdfs.net/User/JGH/Docs/Politics/WardReview/DraftPER/Crookes.gif

Again, this would be a good model to implement in this area, and it eliminates the odd Crookes/Fulwood boundary that snips off little bits here and there.

The current Central ward being +45% top heavy is the biggest problem. However, moving the Walkley boundary to the ring road, moving some into Broomhill and some into Nether Edge to balance their under-representation could get Central down to about +33% - which is 4 members' worth. It could then be quite easily split into two 2-member wards.

This could be the simplest option to implement:
* transfer Orgreave Lane triange to Woodseats
* transfer Shirecliffe to Southey
* transfer Busk Meadow to Firth Park
* transfer St. George's to Central
* transfer Broomhall to Broomhill
* transfer Woolsey Road to Nether Edge
* split Central into two 2-member wards
in addition, to tidy up and make for better representation:
* replace four 3-member northern rural wards with six 2-member wards
* rotate Crookes/Fulwood/Broomhill wards

People may not be aware, but in the 2002 review everybody agreed it made great sense to have a number of wards that would easily be arranged into parliamentary constituencies. Sheffield in 2002 had six constituencies, but was "entitled" to about 5.5, so a number of wards that would divide by 5.5 was chosen - 23 and 28 were considered, 28 was opted for as 23 wards would have had huuuge electorates of about 18,500 compared to the then average of about 13,500 with 29 wards.

Now, with a House of Commons of 650 members Sheffield is still entitled to 5.5 constituencies, so remaining at 28 wards is a sensible option. However, if the House of Commons is reduced to 600 members, Sheffield would be entitled to exactly 5 constituencies, so a multiple of 5 wards would make sense to avoid the dog's breakfast the Boundary Commission recommended a year ago.

In all the above I've worked on the assumption that reducing the HoC to 600 members is dead and buried - at least until the next ward review which should be about 2035. However, a reduction to 600 would require splitting 84 members up into 5 areas. However, with a handful of 2-member wards this becomes hugely much easier. When I have some time I'll throw a few ideas together.

Edit: I've just done calculations, and the electorate entirely inside the Inner Ring Road is exactly the right number for a single 2-member ward. That's so neat it would be perverse not to use it.

A review in 2013 would result in all-ups in 2016 for the whole council, regardless of how many wards are redrawn.

[marksenior]

Thanks for a very informative article .

[YL]

Thanks.

I must admit I'm slightly sceptical about 2-member wards in councils which elect by thirds. Comments in other threads have suggested that the LGBCE isn't keen on them, and I can imagine a certain type of person complaining that they aren't getting a say when everywhere else in the city is. (Of course the answer would be that they were getting a relatively larger say in the other years, but I'm not convinced that that type of person would really get that.) That said, your Stocksbridge/Bradfield/Ecclesfield map which uses them does look quite a bit better than what those areas have now.

As I understand it the first phase of the process is that the LGBCE will consult on council size, and only once they've done that do they start looking at boundaries. I don't know what the chances are of the council size changing, but those suggestions seem to give 85 rather than the current 84, as you've added one councillor in the area of the current Central ward but not removed one elsewhere.

Rotating Crookes, Broomhill and Fulwood seems sensible. Presumably the electorates on the map showing that are from the time of the last review?

[J.G.Harston]

Mar 7, 2013 19:43:36 GMT YL said:

I must admit I'm slightly sceptical about 2-member wards in councils which elect by thirds.

Yes, I don't like sub-3-member wards in urban areas. I made the point in 2002/2004 that forcing an urban-area solution in rural areas can give bad results. Before 1974 the municipal/borough areas that 3-member wards tended to be in were fully urban, the 1974 creation of metropolitan authorities without acknowledging that many of them were more than 50% open countryside where 3-member wards would be odd was one of the 1972/74 flaws.

Mar 7, 2013 19:43:36 GMT YL said:

As I understand it the first phase of the process is that the LGBCE will consult on council size, and only once they've done that do they start looking at boundaries. I don't know what the chances are of the council size changing,

In 2002/2004 we (Sheffield) went from 29*3 to 28*3, specifically to fit neatly into 5.5 constituencies that we (people submitting options) knew would result from the 2005/2010 Parliamentary review.

but those suggestions seem to give 85 rather than the current 84, as you've added one councillor in the area of the current Central ward but not removed one elsewhere.

Yes, I realised that about an hour later - but by then I was in bed It was only an initial scribble prompted by noticing that there was enough deficit around Central to slice little bits off to get it to 133% which would give 4 * 33.3%. What I would like to work to is a solution that preserves the "strong, distinctive, geographically obvious" boundaries on the ground formed by the three railways lines and avoid going back to the 1980 map which sliced almost all of Sheffield's communities straight down the middle. At the moment, Handsworth is really the only one left like that. In 1980-2004 Crookes (the area) was split in four wards!

Rotating Crookes, Broomhill and Fulwood seems sensible. Presumably the electorates on the map showing that are from the time of the last review?

Yes, they are from 2002/2004, as are the figures on the Parish area map, but they haven't changed by much, I'll update them at some point.

[J.G.Harston]

Not yet up to my usual standards, but an initial launching page to make resources and background data available to people who want to use it:

mdfs.net/maps/Sheffield/lgbr2013/

Anybody is free to use the resources there.

(I spent the weekend doing some major structural changes to my website. I hadn't expected more than one boundary review during the lifetime of my interest, so the directory structure was built on "the" boundary review, I've had to shuffle things around to create a new structure for more than one.)

[stepney]

I thought we'd had this discussion; it's no use working off the current electorates. The LGBCE will look to get electoral equality in 2019 (or whenever five years after the end of the review is), by which time Central at least should be much more than 44.5% above the average (correspondingly, all the other wards even if their electorates stay static will be dragged further away from the average).

[erlend]

I think it is 5 years from the beginning of the review. So 2018 or is it 2017.

[J.G.Harston]

May 14, 2013 10:59:49 GMT stepney said:

I thought we'd had this discussion; it's no use working off the current electorates. The LGBCE will look to get electoral equality in 2019 (or whenever five years after the end of the review is), by which time Central at least should be much more than 44.5% above the average (correspondingly, all the other wards even if their electorates stay static will be dragged further away from the average).

Yes, I'm still putting documents and data together, I'm just plonking everything I've got on my website so that people can use it.

I don't know if I'll have the time to put together anything other than a minimal proposal, I'll probably mainly draw up maps for other people's proposals and help with any sub-polling district house-by-house counts that are needed, like I did with the (abandoned) Parliamentary review two years ago.

Yesterday I bumped into a former party colleague who said: "oh, can you just put together a map of 25 wards", to which I tried to explain exactly how much work that is, it's not just a case of arbitarily drawing lines like some colonial administrator. I started the research for my proposal for the 2002 review four years before the review actually started. I'm too busy trying to earn enough to stay alive at the moment.

(Though, if any organisation in Sheffield wishes to actually employ me to draw up some proposals to their specification, I'm available.)

May 14, 2013 11:58:26 GMT erlend said:

I think it is 5 years from the beginning of the review. So 2018 or is it 2017.

In the 2002 review we used actual electorate data from the 2001 register (published in December 2000) and forecasts for 2006 (ie 2001 plus 5).

[stepney]

May 14, 2013 22:24:01 GMT J.G.Harston said:

I started the research for my proposal for the 2002 review four years before the review actually started. I'm too busy trying to earn enough to stay alive at the moment.

That, if I may say so, sounds like creating a bit too much work for yourself. I know our people here, where we've just had a ward review, didn't bother starting work until they'd got both the five-year projected electorate data and final confirmation from the LGBCE of the size of the council. I know things were different in 2002 in that there was no separate council size stage

It might be a bit easier this time if Central is projected to grow to the point you can simply have an 87-member council, split Central in two and tinker at the edges.

May 14, 2013 11:58:26 GMT erlend said:

I think it is 5 years from the beginning of the review. So 2018 or is it 2017.

It's five years following the end of the review. In practice, six years from the start. At the moment there is no page up for a Sheffield FER on the LGBCE website, so they've not started yet.

[Philip Davies]

May 13, 2013 22:39:34 GMT J.G.Harston said:

Not yet up to my usual standards, but an initial launching page to make resources and background data available to people who want to use it:

mdfs.net/maps/Sheffield/lgbr2013/

Anybody is free to use the resources there.

(I spent the weekend doing some major structural changes to my website. I hadn't expected more than one boundary review during the lifetime of my interest, so the directory structure was built on "the" boundary review, I've had to shuffle things around to create a new structure for more than one.)

This is an excellent set of resources. Especially the ward maps showing the polling districts. Are the locality names of the districts official ones or your own description?

In your "An initial minimal change model" summary you refer to Woodseats when I think your mean Woodhouse.

[erlend]

May 15, 2013 10:33:08 GMT stepney said:

It's five years following the end of the review

[/url]. In practice, six years from the start. At the moment there is no page up for a Sheffield FER on the LGBCE website, so they've not started yet.[/quote]OK. That is useful to know. I had always thought the opposite.

[J.G.Harston]

May 15, 2013 10:33:08 GMT stepney said:

May 14, 2013 22:24:01 GMT J.G.Harston said:

I started the research for my proposal for the 2002 review four years before the review actually started.

That, if I may say so, sounds like creating a bit too much work for yourself.

I didn't create the work, I just happened to start looking into it in about 1998. Over a couple of years on and off I put some background data together, mainly ensuring I had the maps and electorate data I'd need. I put my report together over several weeks in 2002, but actually sitting down and collecting together lumps of geography in appropriate ways and doing the calculations and working things out is hours of work.

As an example, you neccessarily have to start at the edges to avoid ending up with half a ward left over somewhere. Look at Stocksbridge & Upper Don ward - which is already too big geographically but would be too small numerically in an arbitary move to 25 wards. What extra bits of geography would you add to get the extra 3000 electors to create a 16,150 ward?

Ignoring the fact that Sheffield already has the largest wards in most of Yorkshire and is in the top ten of all councils, there is always an underlying current to decrease the number of members, increasing the size of wards. Some initial calculations I've done show that a drop to 27 wards would give workable sets of numbers. But it would neccessarily require the removal of one of the south-east wards, as they are all very undersized, I expect that would be met with shrieks of: how dare you suggest getting rid of a Labour ward!!!! Well, the alternative to dropping one ward would be to go up one ward - which would neccessarily have to be reducing the size of the centre-west wards - which are almost all too big - to add another.

May 15, 2013 10:33:08 GMT stepney said:

It might be a bit easier this time if Central is projected to grow to the point you can simply have an 87-member council, split Central in two and tinker at the edges.

I've done a first approximation of the 2018 electorate and Central becomes 28,000! More or less exactly two wards' worth, so could actually just be split into two 3-member wards.

May 14, 2013 11:58:26 GMT erlend said:

I think it is 5 years from the beginning of the review. So 2018 or is it 2017.

In 2002 we had to use the 2001 data and forecasts to 2006, so I would expect the 2013 review to use the 2012 data and forecasts to 2017.

May 14, 2013 11:58:26 GMT erlend said:

At the moment there is no page up for a Sheffield FER on the LGBCE website, so they've not started yet.

At the beginning of the thread I quoted somebody's reply from the LGBC confirming the review was due to start in 2013.

May 15, 2013 10:40:12 GMT Philip Davies said:

This is an excellent set of resources. Especially the ward maps showing the polling districts. Are the locality names of the districts official ones or your own description?

Do you mean the Polling District names? They are the official ones as listed in the electoral register.

[erlend]

The example give by the LGBCE cited by Stepney at www.lgbce.org.uk/__documents/lgbce/guidance-policy-and-publications/guidance/proposing-new-wards-guidance.pdf on p7 says

Local authority A has an electorate of 100,000 in 2012. In 2018 (five years
after the planned completion of the review), the electorate of Local authority A
is expected to be 103,000.

Without looking out the legislation it appears to be based on the completion date (not what I had believed).

[YL]

Will the projections be available during the council size stage? Looking at other councils at that stage at the moment, I can't see them.

If splitting Central in two and going for 29 wards with minimal change elsewhere is workable, then I think that should be the way to go. The current map is really quite good given the desire for electorate equality, the weird shape of Broomhill notwithstanding.

I don't think it's worth trying to second guess the parliamentary review process at this point. Yes we'd have ended up with much better proposals in the abortive review if there had been 25 or 30 wards, but we don't know what the rules (or indeed the city's entitlement) are going to be and anyway I think the next review may still be using the current wards.

[stepney]

May 15, 2013 15:12:30 GMT J.G.Harston said:

As an example, you neccessarily have to start at the edges to avoid ending up with half a ward left over somewhere. Look at Stocksbridge & Upper Don ward - which is already too big geographically but would be too small numerically in an arbitary move to 25 wards. What extra bits of geography would you add to get the extra 3000 electors to create a 16,150 ward?

I'm not really qualified to comment on Sheffield, but my immediate response would be "Grenoside". And if I were up there and doing it I would check first not the geography but whether it benefitted my party, do it if it did, and find the arguments for it afterwards. Putting Grenoside into Stocksbridge looks to me like a ward could then be formed (I've not checked the numbers) out of the discrete area of Chapeltown, High Green and Burncross, which looks like a much better argument could be made for it than the East/West Ecclesfield split.

Realistically (given they won't want to split the Ecclesfield CP into three different wards) I guess the more obvious area to take 3,000 electors from would be the top end of Stannington.

May 15, 2013 15:12:30 GMT J.G.Harston said:

I've done a first approximation of the 2018 electorate and Central becomes 28,000!

That is creating work for yourself. The projections the LGBCE use will be prepared by Sheffield City Council. Unless they muck up spectacularly and you're able to challenge their figures with solid evidence, you can't just use your own.

May 15, 2013 15:12:30 GMT J.G.Harston said:

In 2002 we had to use the 2001 data and forecasts to 2006, so I would expect the 2013 review to use the 2012 data and forecasts to 2017.

They've just completed a review here where they used projections to 2018. The Sheffield review, if it's starting this summer, will use forecasts to 2019.

[Philip Davies]

Assuming Sheffield's electorate continues to grow faster than the UK's as a whole then it might be entitled to something like 5.2 constituencies by 2017/8. That is assuming that the sixth periodic review is redone then with 600 seats. If so 26 wards might be the best approach so that you have 5 constituencies of 5 wards and one left to join up with either Barnsley or Rotherham.

However as mentioned above some two member wards might give more flexibility here.

[J.G.Harston]

May 15, 2013 17:29:32 GMT YL said:

Will the projections be available during the council size stage? Looking at other councils at that stage at the moment, I can't see them.

I'm expecting them soon after the official start of the review, as that will be when it will be officially specified which years' electorates must be used.

May 15, 2013 17:29:32 GMT YL said:

and anyway I think the next (Parliamentary) review may still be using the current wards.

A Parliamentary Review uses whatever wards are there at the start of the review. So, the 2005 review implmented in 2010 used the wards that had just come in in 2004. If the normal 15-ish-year timetable is used, the 2020 review will use whatever wards are there in 2020, which will be whatever "new"-ish wards have been set in place for 2016.

May 15, 2013 18:09:10 GMT stepney said:

May 15, 2013 15:12:30 GMT J.G.Harston said:

As an example ... look at Stocksbridge & Upper Don ward ... what extra bits of geography would you add to get the extra 3000 electors to create a 16,150 ward?

I'm not really qualified to comment on Sheffield, but my immediate response would be "Grenoside".

That ends up extending the "snake" looping around Wharncliffe Wood making for an ever weirder shape joing to an area with even less connections to Stocksbridge than Oughtibridge that had to go in for the numbers last time.

May 15, 2013 18:09:10 GMT stepney said:

Realistically ...I guess the more obvious area to take 3,000 electors from would be the top end of Stannington.

The problem is, there is no "top" of Stannington. Almost all of Stannington ward is in a compact area at the east of the ward, the rest is scattered villages. You'd end up having to take in all the countryside, and then slice bits off the urban area.

All of which shows that fiddlyness of finding the numbers that seems to just go over the head of certain "oh just draw us up a map" people

May 15, 2013 15:12:30 GMT J.G.Harston said:

I've done a first approximation of the 2018 electorate and Central becomes 28,000!

That is creating work for yourself. The projections the LGBCE use will be prepared by Sheffield City Council. Unless they muck up spectacularly and you're able to challenge their figures with solid evidence, you can't just use your own.

It's work I'd do anyway. I have form in doing future electorate projections better than the City Council's Most of my forecasts from 2002 projected to 2009 turned out quite near reality. For example, I forecast that Darnall didn't need to leak over the Sheffield Parkway into Handsworth, and lo, it has sufficient numbers not to. I recommended that East/West Ecclesfield should extend more into Parson Cross as they would be too small and Firth Park too big, and that has come to pass.

However, Central caught us all out. The City Council initially predicted about 8,000, I demonstrated how it would actually be about 13,000 - astonishingly, it's going to end up around 26,000!

I've finished my initial electoral forecast which I'll update when the City Council publish their forecast. It's sufficient to do some initial work with. This was done in idle moments while waiting for lots of Windows 7 installations to build

May 16, 2013 22:09:35 GMT Philip Davies said:

...then it might be entitled to something like 5.2 constituencies by 2017/8 ... with 600 seats. ... However as mentioned above some two member wards might give more flexibility here.

A 2-member Stocksbridge would very easily and neatly join with the Greater Penistone area to create a Penistone/Barnsley West-ish seat.

[YL]

Surely Stocksbridge & Upper Don (already 14,711) only needs a bit over 1000 voters to get it close to the quota for a 25-ward map. It looks like the rural parts of Bradfield parish would be enough, and they'd be a reasonable fit.

However, Stannington then needs to expand into Hillsborough (unless you cross the Rivelin, which isn't a good idea) and Hillsborough then needs quite a lot of electorate from somewhere. Either you cross the railway line at Wadsley Bridge, or you end up splitting Walkley, neither of which looks good.

I'll have more of a look if the LGBCE actually decides on 75 as a council size.

[J.G.Harston]

May 18, 2013 8:29:23 GMT YL said:

Surely Stocksbridge & Upper Don (already 14,711) only needs a bit over 1000 voters to get it close to the quota for a 25-ward map.

No, it would need about 1000 electors, which would be about 300 voters in that area.

May 18, 2013 8:29:23 GMT YL said:

However, Stannington then needs to expand into Hillsborough (unless you cross the Rivelin, which isn't a good idea) and Hillsborough then needs quite a lot of electorate from somewhere. Either you cross the railway line at Wadsley Bridge, or you end up splitting Walkley, neither of which looks good.

Yes, the north-west is severely constrained there. If the LGBCE adopts recommendations for a 75-member council I think the best solution in the north-west would have to have a 2-member ward in the mix.

I've finished putting together my preliminary resources at mdfs.net. I'll leave it now until the review is officially launched, all there data is there for anybody to use.
One of the sections is a set of wad maps with electorate counts in areas smaller than polling districts. If anybody needs this for any specific areas, drop me an email and I'll do the mapping and counting to draw them.

[J.G.Harston]

Sheffield Council has kicked off the consultation:

www.sheffield.gov.uk/your-city-council/elections/electoral-review.html