|
|
Post by tiberius on Oct 24, 2019 8:00:43 GMT
i've got a *fun* Berkshire Torymander coming up.
|
|
|
|
Post by tiberius on Oct 24, 2019 8:27:36 GMT
 SPLIT UP SLOUGH! RIP UP READING!(err..sorry. got overexcited.) |
|
|
|
Post by bjornhattan on Oct 24, 2019 10:23:45 GMT
SPLIT UP SLOUGH! RIP UP READING!(err..sorry. got overexcited.) I think the orange seat might just be Labour (since it takes in some very safe Labour wards in Slough and much of its non-Slough electorate is in central Bracknell so could be pretty marginal). But great job! |
|
|
|
Post by manchesterman on Oct 24, 2019 19:50:52 GMT
The orange seat is insane  |
|
|
|
Post by bjornhattan on Jan 15, 2023 20:15:28 GMT
The constituency with the greatest proportion of students (according to the 2021 census) is Nottingham South, where just over a third of adults are in full time education. This made me wonder - could a majority student seat be drawn? Following the BCE's guidelines, almost certainly not. While there are 35 majority student wards in England, these are scattered all over the place and most do not even come close to bordering each other. There are some promising looking clusters of wards in Leeds, Sheffield, Nottingham, and one or two other cities, but in every case the heavily student wards simply aren't big enough to outweigh very much non-student majority neighbours.
But what if we allowed ourselves the freedom to use any boundaries? I used a combination of adult population figures (from the census) and the official electorate figures to estimate the electorate of each census output area - for example if a ward had 4,000 adults but an electorate of 3,200, I'd multiply all of the output areas' adult populations by 0.8 to give an approximate electorate figure. Obviously these aren't going to be perfect, but far better than just not splitting wards at all, using polling districts (which often have inaccurate boundaries), or using unadjusted population figures.
To identify possible locations for majority student seats, I then identified clusters of output areas with a large student population (at least 30%) - there were several hundred of these so I narrowed it down to those with an electorate of over 10,000 (as a starting point). This reduced my search to 27 different locations - many of these were clearly too small (or in the case of London, too diffuse) but four cities looked to have enough students concentrated in a fairly specific area and so were investigated fully: the three mentioned before as well as Manchester. In each city, I ordered all output areas in the area (not just those with more than 30% students) from most heavily student to least heavily student. I then established the maximum number of electors who could live in a majority student seat - essentially starting with the top of my list and working my way down until non students outnumbered students. In Sheffield, the student areas just didn't have the electorate required, with it only being possible for 67,000 or so electors to live in a seat with a student majority. In Leeds, it was even closer, but the total again fell just short - about 69,200. On the other hand, Nottingham and Manchester had enough output areas with a significant student presence to enable the creation of student constituencies.
But just because a seat is mathematically possible, doesn't mean it can actually be drawn to look at all reasonable! And that brings me to the crux of my post, and maps of the only two majority seats which I believe would currently be within the electoral quota: Nottingham Central and Beeston and Manchester Central and Irwell.   In Nottingham, things were relatively plain sailing. The student population here is quite concentrated into the corridor between Nottingham city centre and Beeston. This wasn't quite enough to hit the quota but fortunately areas like Hyson Green and Forest Fields have a few students so helped make up the numbers without causing the students to be outnumbered. Manchester was much harder, as you can tell by the uglier boundaries - the student population here is more diffuse, and there are also quite a few resolutely non student areas in parts of the city centre and Moss Side. So I was forced to cross into Salford, running a narrow finger all the way to Charlestown and also including a small section of Pendleton. I'm not massively sold on the name - I needed to acknowledge the cross-borough link somehow, but it's not as if the areas included make up any kind of coherent unit - it's purely a case of making up the numbers. In case you're wondering, the electorates are 69,765 and 70,261 respectively; the Nottingham seat has 48,182 students out of 96,180 adult residents (50.10%) and the Manchester seat has 46,675 students out of 93,001 adult residents (50.19%).
Obviously, in theory more seats could be drawn as majority student if we were willing to accept true American style gerrymandering. For instance, there are predominantly student areas in Bradford and Huddersfield which could enable such a constituency in West Yorkshire, and mathematically the numbers suggest up to 17 could theoretically be possible. But I think I've done enough damage for one day!
|
|
|
|
Post by Pete Whitehead on Jan 15, 2023 20:33:22 GMT
Brilliant. Is it possible do you think to draw any majority Black seats in London (or even one in the West Midlands)? Again obviously American style gerrymandering would be needed and it may not be possible even then |
|
|
|
Post by iainbhx on Jan 15, 2023 21:01:54 GMT
Brilliant. Is it possible do you think to draw any majority Black seats in London (or even one in the West Midlands)? Again obviously American style gerrymandering would be needed and it may not be possible even then I don't think that there are any Black majority MSOA's in England and Wales, so it would be incredibly hard to do out of LSOA's. |
|
|
|
Post by bjornhattan on Jan 15, 2023 21:03:29 GMT
Brilliant. Is it possible do you think to draw any majority Black seats in London (or even one in the West Midlands)? Again obviously American style gerrymandering would be needed and it may not be possible even then It is definitely impossible in the West Midlands. You can probably get to about 20,000 electors in parts of Aston and Ladywood, but that's the absolute limit.
In theory, it could be possible in South London, where up to 120,000 electors could live in a majority Black constituency (if you allow crossing the Thames to take in areas around the Royal Docks, Barking Riverside, and/or Thurrock, you could maybe even draw two such seats). Some of the output areas with a significant Black community to enable this are scattered randomly in the likes of Woolwich, Croydon, and Thamesmead, but there's a fairly solid band of output areas across central Lambeth, Southwark, and into Lewisham. I'll have a go now - but I'm not expecting it to be easy because the area is very well integrated without clear lines between communities.
|
|
|
|
Post by bjornhattan on Jan 15, 2023 21:29:39 GMT
Could a Birmingham Edgbaston and Aston constituency be drawn to have a majority of students? No - the issue with Birmingham is not the number of students (it has about 60,000 - which is more than was needed to create majority student seats in Nottingham or Manchester) but their distribution. Even more so than Manchester, the student population is very diffuse and most students live on streets where they do not form the majority (or even close to it). It'd probably be possible to draw a majority student seat in the West Midlands, but you'd have to adopt an American approach and have a strip of uninhabited land connecting Selly Oak to Earlsdon (and probably on to Coventry and/or the Warwick campus). |
|
|
|
Post by greenchristian on Jan 16, 2023 19:50:48 GMT
Could a Birmingham Edgbaston and Aston constituency be drawn to have a majority of students? No - the issue with Birmingham is not the number of students (it has about 60,000 - which is more than was needed to create majority student seats in Nottingham or Manchester) but their distribution. Even more so than Manchester, the student population is very diffuse and most students live on streets where they do not form the majority (or even close to it). It'd probably be possible to draw a majority student seat in the West Midlands, but you'd have to adopt an American approach and have a strip of uninhabited land connecting Selly Oak to Earlsdon (and probably on to Coventry and/or the Warwick campus). Warwick campus would be easier to get to cross-country and has the advantage of the population being nearly 100% student. I'd also prioritise adding Coventry city centre over Earlsdon. St Michael's ward is majority student, which isn't true of more than a handful of roads in Earlsdon.
|
|
|
|
Post by bjornhattan on Jan 16, 2023 23:22:51 GMT
Brilliant. Is it possible do you think to draw any majority Black seats in London (or even one in the West Midlands)? Again obviously American style gerrymandering would be needed and it may not be possible even then It looks rather more like Chicago than London, but it is just about possible - I present Greater Peckham. The boundaries are incredibly intricate - I've highlighted the areas within the constituency in yellow to clearly distinguish them from the rest of South London. There are two major issues that make drawing the seat very difficult: firstly the Black population tends to be found on particular estates which are clustered all over the place, and secondly there are various brownfield developments all over the place in South London which have mainly White and/or Asian residents.
The electorate of the seat is 70,857 - just over half of this comes from Southwark, with Lambeth contributing just under a third, Lewisham about one seventh, and a few hundred electors each from Greenwich and Wandsworth. I chose the name because Peckham is the only electoral ward where most residents of the ward end up in this proposed seat, despite the fact that twenty eight different wards contribute at least some electors.
In total, there are 111,945 residents in the seat: 56,043 are Black (50.06%). A few other census stats which may be of interest:
The full ethnic composition: 27.8% White (17.9% British, 8.4% Other, remainder mostly Irish); 7.3% Mixed; 7.8% Asian; 50.1% Black (30.6% African, 13.1% Carribbean, 6.4% "other Black"); 0.8% Arab; 6.2% Other. It would have a higher proportion of Black African, Mixed White/Black African, and "Other Black" residents than any other constituency in the country; for Carribbean population it would be surpassed only by Croydon North.
The proportion of adult residents with a degree would be much lower than for any current seats in this area, but at 40.8% would still be within the top 100 constituencies in England and Wales - comparable to Harrogate and Knaresborough, Macclesfield, or The Cotswolds. While not high compared with the country as a whole, apprenticeship holders would make up 3.2% of adults - higher than all but two current Inner London constituencies (Eltham and Lewisham East). - Christians would make up 52.4% of the population - a higher proportion than every current London constituency except for Hornchurch and Upminster.
- But perhaps the most extreme variable of all would be tenure. A staggering 68.2% of households would live in socially rented accommodation - more than 20% higher than any currently existing constituency. Home ownership rates would be just 16.7% with only 4.6% owning outright (the current lowest figures are 24.6% and 7.7% respectively). Even private renting would be quite rare - at 15.2%, the proportion of households with a private landlord would be among the lowest 150 constituencies in the country.
I suspect that it would be possible to draw a slightly neater constituency with a Black majority (or an even more untidy one which was perhaps 51-52% Black). But I doubt you could draw a significantly neater one...
|
|
|
|
Post by Pete Whitehead on Jan 16, 2023 23:28:42 GMT
Some seriously grim estates in that seat
|
|
|
|
Post by michaelarden on Jan 17, 2023 13:07:46 GMT
The constituency with the greatest proportion of students (according to the 2021 census) is Nottingham South, where just over a third of adults are in full time education. This made me wonder - could a majority student seat be drawn? Following the BCE's guidelines, almost certainly not. While there are 35 majority student wards in England, these are scattered all over the place and most do not even come close to bordering each other. There are some promising looking clusters of wards in Leeds, Sheffield, Nottingham, and one or two other cities, but in every case the heavily student wards simply aren't big enough to outweigh very much non-student majority neighbours.
But what if we allowed ourselves the freedom to use any boundaries? I used a combination of adult population figures (from the census) and the official electorate figures to estimate the electorate of each census output area - for example if a ward had 4,000 adults but an electorate of 3,200, I'd multiply all of the output areas' adult populations by 0.8 to give an approximate electorate figure. Obviously these aren't going to be perfect, but far better than just not splitting wards at all, using polling districts (which often have inaccurate boundaries), or using unadjusted population figures.
To identify possible locations for majority student seats, I then identified clusters of output areas with a large student population (at least 30%) - there were several hundred of these so I narrowed it down to those with an electorate of over 10,000 (as a starting point). This reduced my search to 27 different locations - many of these were clearly too small (or in the case of London, too diffuse) but four cities looked to have enough students concentrated in a fairly specific area and so were investigated fully: the three mentioned before as well as Manchester. In each city, I ordered all output areas in the area (not just those with more than 30% students) from most heavily student to least heavily student. I then established the maximum number of electors who could live in a majority student seat - essentially starting with the top of my list and working my way down until non students outnumbered students. In Sheffield, the student areas just didn't have the electorate required, with it only being possible for 67,000 or so electors to live in a seat with a student majority. In Leeds, it was even closer, but the total again fell just short - about 69,200. On the other hand, Nottingham and Manchester had enough output areas with a significant student presence to enable the creation of student constituencies.
But just because a seat is mathematically possible, doesn't mean it can actually be drawn to look at all reasonable! And that brings me to the crux of my post, and maps of the only two majority seats which I believe would currently be within the electoral quota: Nottingham Central and Beeston and Manchester Central and Irwell. In Nottingham, things were relatively plain sailing. The student population here is quite concentrated into the corridor between Nottingham city centre and Beeston. This wasn't quite enough to hit the quota but fortunately areas like Hyson Green and Forest Fields have a few students so helped make up the numbers without causing the students to be outnumbered. Manchester was much harder, as you can tell by the uglier boundaries - the student population here are more diffuse, and there are also quite a few resolutely non student areas in parts of the city centre and Moss Side. So I was forced to cross into Salford, running a narrow finger all the way to Charlestown and also including a small section of Pendleton. I'm not massively sold on the name - I needed to acknowledge the cross-borough link somehow, but it's not as if the areas included make up any kind of coherent unit - it's purely a case of making up the numbers. In case you're wondering, the electorates are 69,765 and 70,261 respectively; the Nottingham seat has 48,182 students out of 96,180 adult residents (50.10%) and the Manchester seat has 46,675 students out of 93,001 adult residents (50.19%).
Obviously, in theory more seats could be drawn as majority student if we were willing to accept true American style gerrymandering. For instance, there are predominantly student areas in Bradford and Huddersfield which could enable such a constituency in West Yorkshire, and mathematically the numbers suggest up to 17 could theoretically be possible. But I think I've done enough damage for one day!
Could a Birmingham Edgbaston and Aston constituency be drawn to have a majority of students? What about Edinburgh or Glasgow? |
|
|
|
Post by manchesterman on Jan 17, 2023 13:48:17 GMT
Brilliant. Is it possible do you think to draw any majority Black seats in London (or even one in the West Midlands)? Again obviously American style gerrymandering would be needed and it may not be possible even then It looks rather more like Chicago than London, but it is just about possible - I present Greater Peckham. The boundaries are incredibly intricate - I've highlighted the areas within the constituency in yellow to clearly distinguish them from the rest of South London. There are two major issues that make drawing the seat very difficult: firstly the Black population tends to be found on particular estates which are clustered all over the place, and secondly there are various brownfield developments all over the place in South London which have mainly White and/or Asian residents.
The electorate of the seat is 70,857 - just over half of this comes from Southwark, with Lambeth contributing just under a third, Lewisham about one seventh, and a few hundred electors each from Greenwich and Wandsworth. I chose the name because Peckham is the only electoral ward where most residents of the ward end up in this proposed seat, despite the fact that twenty eight different wards contribute at least some electors.
In total, there are 111,945 residents in the seat: 56,043 are Black (50.06%). A few other census stats which may be of interest:
The full ethnic composition: 27.8% White (17.9% British, 8.4% Other, remainder mostly Irish); 7.3% Mixed; 7.8% Asian; 50.1% Black (30.6% African, 13.1% Carribbean, 6.4% "other Black"); 0.8% Arab; 6.2% Other. It would have a higher proportion of Black African, Mixed White/Black African, and "Other Black" residents than any other constituency in the country; for Carribbean population it would be surpassed only by Croydon North.
The proportion of adult residents with a degree would be much lower than for any current seats in this area, but at 40.8% would still be within the top 100 constituencies in England and Wales - comparable to Harrogate and Knaresborough, Macclesfield, or The Cotswolds. While not high compared with the country as a whole, apprenticeship holders would make up 3.2% of adults - higher than all but two current Inner London constituencies (Eltham and Lewisham East). - Christians would make up 52.4% of the population - a higher proportion than every current London constituency except for Hornchurch and Upminster.
- But perhaps the most extreme variable of all would be tenure. A staggering 68.2% of households would live in socially rented accommodation - more than 20% higher than any currently existing constituency. Home ownership rates would be just 16.7% with only 4.6% owning outright (the current lowest figures are 24.6% and 7.7% respectively). Even private renting would be quite rare - at 15.2%, the proportion of households with a private landlord would be among the lowest 150 constituencies in the country.
I suspect that it would be possible to draw a slightly neater constituency with a Black majority (or an even more untidy one which was perhaps 51-52% Black). But I doubt you could draw a significantly neater one...
This is impressive for the sheer amount of labour (pardon the pun) which must have gone into working all this out! Chapeau! |
|
|
|
Post by minionofmidas on Jan 18, 2023 2:05:39 GMT
Assuming almost all of that mixed race population is part Black, you could probably have a far more compact majority Black Alone or in Combination district? The formerly case law relevant threshold in the US would be 50% Alone or in Combo of Voting Age Population.
|
|
|
|
Post by iainbhx on Jan 18, 2023 6:43:11 GMT
Assuming almost all of that mixed race population is part Black, you could probably have a far more compact majority Black Alone or in Combination district? The formerly case law relevant threshold in the US would be 50% Alone or in Combo of Voting Age Population. There are splits on that Mixed population, certainly in inner Birmingham you would be very unwise to assume that "almost all" is part Black. It seems to vary and not quite in the areas you'd think. |
|
|
|
Post by bjornhattan on Jan 18, 2023 14:01:29 GMT
Assuming almost all of that mixed race population is part Black, you could probably have a far more compact majority Black Alone or in Combination district? The formerly case law relevant threshold in the US would be 50% Alone or in Combo of Voting Age Population. Less than you'd think. My Greater Peckham seat would be approximately 54% Black alone or in combination (IIRC the Mixed race population there breaks down as 2.9% Mixed White/Black Caribbean, 1.4% Mixed White/Black African, 1.0% Mixed White/Asian, 2.0% Mixed Other) - this increase might enable a slightly neater seat but probably not by a huge margin. The numbers involved certainly wouldn't allow for a majority Black seat in the West Midlands or a second seat in London - a major reason for this is that the Mixed population is (unsurprisingly) extremely dispersed with few areas having very few Mixed residents but few having more than a few percent. |
|
