Redistribution by equalising population

[Davıd Boothroyd]

Following on from a thread in the local byelections board, I've had a look at what would be the effect of redistributing constituencies by usual resident population as measured in the Census, and whether it would change significantly if it was restricted to only the adult population (a sort of proxy for the total eligible electorate, although it also includes those unable to register due to nationality and other reasons). Here's the table and an indicative allocation of seats. I've based it on 650 seats and included the current numbers.

Region	All Usual Residents	Age 18+	% pop	% 18+	Seats pop	Seats 18+	Current seats
North East	2,596,886	2,070,844	4.11	4.16	26.72	27.06	29
North West	7,052,177	5,550,138	11.16	11.16	72.55	72.53	75
Yorkshire and The Humber	5,283,733	4,154,998	8.36	8.35	54.36	54.30	54
East Midlands	4,533,222	3,579,952	7.17	7.20	46.64	46.78	46
West Midlands	5,601,847	4,361,384	8.87	8.77	57.63	57.00	59
East of England	5,846,965	4,590,257	9.25	9.23	60.15	59.99	58
London	8,173,941	6,362,547	12.94	12.79	84.09	83.15	73
South East	8,634,750	6,775,054	13.67	13.62	88.83	88.54	84
South West	5,288,935	4,230,322	8.37	8.51	54.41	55.28	55
Wales	3,063,456	2,430,049	4.85	4.89	31.52	31.76	40
Scotland	5,295,403	4,252,806	8.38	8.55	54.48	55.58	59
Northern Ireland	1,810,863	1,380,100	2.87	2.77	18.63	18.04	18
Total	63,182,178	49,738,451			650	650	650

[johnsmith]

A massive increase in London seats, as if we are not dominated anywhere near enough by a metropolitan elite already. And marked increases in already Tory areas, with fewer seats in Labour-supporting ones.

Which is pretty much the whole purpose.

Maybe Labour needs to learn the lessons of this Tory gerrymandering and respond similarly when in power in ways that sound highly reasonable. Eg going back to basing constituency sizes on the actual sizes of the adult populations living there, and lowering the voting age to 16 whilst removing the vote from expats living abroad - "those who live here and pay their taxes here, vote here, those who don't, don't"

Of course, more 16-17 year olds vote left than right whilst more expats vote right than left, lol.

And we could seriously consider secure ways to introduce internet voting to "encourage democratic participation and make it easier for hard-working people in the modern age."

Of course, many more younger people - more likely to vote left than right - use the internet than older people, more likely to vote right than left, lol.

If the Tories only want to introduce those fair-sounding measures that help them, maybe we should introduce the fair-sounding ones that only help us? Lol.

[J.G.Harston]

Jan 12, 2016 11:17:37 GMT Davıd Boothroyd said:

...the adult population (a sort of proxy for the total eligible electorate, although it also includes those unable to register due to nationality and other reasons).

Because Sheffield has such huge numbers of Chinese students I felt it was required to drill further down the census figures to do my best to filter out non-eligible adults from my Sheffield figures, particularly as they were likely to be concentrated in few areas. But across a who;le region they are unlikely to skew the figures so the total 18+ figure is good enough for comparing whole regions.

Happy to see that it's another concrete piece of evidence that, at the broad level, electorate and/or adult totals is a sufficiently close proxy to population as to justifying using them to apportion the country by population, and to counter arguments that we should apportion the country by population by saying "we already do".

Of course, not to counter the fact that as you drill down into smaller areas such proxy figures become less accurate, just that, for example, Yorkshire "should" have 54 MPs by population, or 54 MPs by adult population, or 54 MPs by electorate.

[carlton43]

That is very telling David. Is London grossly under-represented because of fast growing population or because you are counting in a very large number of foreigners living there but who are not entitled to vote in HOC elections? Or both? Why has Wales been so grossly over-represented? Are the populations of the North-East and the West Midlands actually declining or were they over-represented?

On the 2015 GE Thread I have posted some thoughts and results relating to swing. I was wrong to state your position was absurdist over the absolute need for showing only Classic Conservative-Labour swing and I apologize. But I do contend that whilst Classic Swing should always be compiled and shown first, there is a case to be made for supplementary compilation of a swing between winner and non major runner-up that you might concede has added value.

[carlton43]

Jan 12, 2016 11:33:18 GMT johnsmith said:

A massive increase in London seats, as if we are not dominated anywhere near enough by a metropolitan elite already. And marked increases in already Tory areas, with fewer seats in Labour-supporting ones.

Which is pretty much the whole purpose.

Maybe Labour needs to learn the lessons of this Tory gerrymandering and respond similarly when in power in ways that sound highly reasonable. Eg going back to basing constituency sizes on the actual sizes of the adult populations living there, and lowering the voting age to 16 whilst removing the vote from expats living abroad - "those who live here and pay their taxes here, vote here, those who don't, don't"

Of course, more 16-17 year olds vote left than right whilst more expats vote right than left, lol.

And we could seriously consider secure ways to introduce internet voting to "encourage democratic participation and make it easier for hard-working people in the modern age."

Of course, many more younger people - more likely to vote left than right - use the internet than older people, more likely to vote right than left, lol.

If the Tories only want to introduce those fair-sounding measures that help them, maybe we should introduce the fair-sounding ones that only help us? Lol.

That is a terrible post John and I hope you reflect upon it.  London to be denied the extra seats entitled by population growth because you feel it already has too much power and influence!  Southern areas to be denied extra seats because they tend to vote Conservative! What a viciously anti-democratic and irrational position to hold. Surely you might note that the extra London seats would be more likely Labour than not? But even that is not good enough for you as they would be 'the wrong sort of Labour seats'!  

[johnsmith]

Jan 12, 2016 11:54:54 GMT carlton43 said:

Jan 12, 2016 11:33:18 GMT johnsmith said:

A massive increase in London seats, as if we are not dominated anywhere near enough by a metropolitan elite already. And marked increases in already Tory areas, with fewer seats in Labour-supporting ones.

Which is pretty much the whole purpose.

Maybe Labour needs to learn the lessons of this Tory gerrymandering and respond similarly when in power in ways that sound highly reasonable. Eg going back to basing constituency sizes on the actual sizes of the adult populations living there, and lowering the voting age to 16 whilst removing the vote from expats living abroad - "those who live here and pay their taxes here, vote here, those who don't, don't"

Of course, more 16-17 year olds vote left than right whilst more expats vote right than left, lol.

And we could seriously consider secure ways to introduce internet voting to "encourage democratic participation and make it easier for hard-working people in the modern age."

Of course, many more younger people - more likely to vote left than right - use the internet than older people, more likely to vote right than left, lol.

If the Tories only want to introduce those fair-sounding measures that help them, maybe we should introduce the fair-sounding ones that only help us? Lol.

That is a terrible post John and I hope you reflect upon it.  London to be denied the extra seats entitled by population growth because you feel it already has too much power and influence!  Southern areas to be denied extra seats because they tend to vote Conservative! What a viciously anti-democratic and irrational position to hold. Surely you might note that the extra London seats would be more likely Labour than not? But even that is not good enough for you as they would be 'the wrong sort of Labour seats'!  

Something of a misrepresentation, if I may say so. I in no way oppose basing numbers of seats on size of the adult population. I merely believe that it should be based on exactly that, and not just those on the electoral register. We must after all represent everyone, whether they are on the register or not. Some might not sign as a political protest in itself for whatever reason. They should still be counted.

[Davıd Boothroyd]

Jan 12, 2016 11:37:23 GMT carlton43 said:

That is very telling David. Is London grossly under-represented because of fast growing population or because you are counting in a very large number of foreigners living there but who are not entitled to vote in HOC elections? Or both?

Both; plus there is more electoral under-registration in London than anywhere else.

Why has Wales been so grossly over-represented?

That's not a bug, that's a feature. Wales has a separate electoral quota from England, which is set significantly lower. In effect Wales is deliberately over-represented because of its separate historical identity; the same thing used to apply to Scotland but no longer does since the creation of the Scottish Parliament. Welsh representation was not reduced at the same time because the National Assembly did not have primary legislative powers. It now does, and there really isn't a justification for the separate quota.

Are the populations of the North-East and the West Midlands actually declining or were they over-represented?

I think it's partly population decline, but also that electoral registration is relatively high.

[Pete Whitehead]

David's figures are based on resident population not just those on the register. This is the whole point. You see a pro-Tory gerrymander in what is actually intended as a pro-Labour one

[andrewteale]

Comparing the seats by population with the current seat numbers isn't quite a fair comparison, because the current seat numbers are based on December 2000 electorates and the population figures are based on the April 2011 census.  In particular, the population of the South East region has grown strongly since 2000.

Here's a table comparing David's population entitlements with entitlements based on the December 2010 parliamentary electorates. This strips out the effect of population changes from 2000 to 2010.

Region	Parly electors	Entitlement	By population	Difference	Loc gov electors	Entitlement	By population	Difference
South West	4042475	57.32	54.41	2.91	4107965	56.85	54.41	2.44
North West	5253019	74.48	72.55	1.93	5319218	73.61	72.55	1.06
North East	1971249	27.95	26.72	1.23	1984352	27.46	26.72	0.74
Scotland	3928979	55.71	54.48	1.23	3985161	55.15	54.48	0.67
East Midlands	3361089	47.65	46.64	1.01	3428913	47.45	46.64	0.81
Wales	2281596	32.35	31.52	0.83	2302300	31.86	31.52	0.34
West Midlands	4115668	58.35	57.63	0.72	4180455	57.85	57.63	0.22
Eastern	4280707	60.69	60.15	0.54	4385930	60.70	60.15	0.55
South East	6303428	89.37	88.83	0.54	6453819	89.32	88.83	0.49
Yorkshire	3848942	54.57	54.36	0.21	3901908	54.00	54.36	-0.36
Northern Ireland	1190635	16.88	18.63	-1.75	1202145	16.64	18.63	-1.99
London	5266904	74.68	84.09	-9.41	5715418	79.10	84.09	-4.99
TOTAL	45844691	650	Quota 70530		46967584	650	Quota 72258

So the two regions which are clearly disadvantaged by basing seats on parliamentary electorates rather than population are London and Northern Ireland, and the only region clearly advantaged is the South West.  Changing to local government electorate (which includes EU nationals) almost halves the disadvantage for London.

That suggests to me that the main drivers of the difference between electorate and population are (a) immigration from outside the EU and Commonwealth, which is concentrated in London, and (b) the population of London (and to a lesser extent Northern Ireland) having a relatively high proportion of children.  Other potential differences (such as population transience) appear to have a negligible impact when taken at regional level.

[Deleted]

Jan 12, 2016 11:17:37 GMT Davıd Boothroyd said:

Following on from a thread in the local byelections board, I've had a look at what would be the effect of redistributing constituencies by usual resident population as measured in the Census, and whether it would change significantly if it was restricted to only the adult population (a sort of proxy for the total eligible electorate, although it also includes those unable to register due to nationality and other reasons). Here's the table and an indicative allocation of seats. I've based it on 650 seats and included the current numbers.

Region	All Usual Residents	Age 18+	% pop	% 18+	Seats pop	Seats 18+	Current seats
North East	2,596,886	2,070,844	4.11	4.16	26.72	27.06	29
North West	7,052,177	5,550,138	11.16	11.16	72.55	72.53	75
Yorkshire and The Humber	5,283,733	4,154,998	8.36	8.35	54.36	54.30	54
East Midlands	4,533,222	3,579,952	7.17	7.20	46.64	46.78	46
West Midlands	5,601,847	4,361,384	8.87	8.77	57.63	57.00	59
East of England	5,846,965	4,590,257	9.25	9.23	60.15	59.99	58
London	8,173,941	6,362,547	12.94	12.79	84.09	83.15	73
South East	8,634,750	6,775,054	13.67	13.62	88.83	88.54	84
South West	5,288,935	4,230,322	8.37	8.51	54.41	55.28	55
Wales	3,063,456	2,430,049	4.85	4.89	31.52	31.76	40
Scotland	5,295,403	4,252,806	8.38	8.55	54.48	55.58	59
Northern Ireland	1,810,863	1,380,100	2.87	2.77	18.63	18.04	18
Total	63,182,178	49,738,451			650	650	650

This is an interesting exercise, and the difference for London is indeed very striking. Drilling down into the census figures should be illuminating, so I shall not comment further until I have done so.

On the by-election thread Davıd Boothroyd provided a link to various Electoral Commission reports on electoral registration. These are well worth studying, though my initial take on them is not entirely the same as his. Again, this is something I shall return to.

[carlton43]

Jan 12, 2016 12:05:50 GMT Davıd Boothroyd said:

Jan 12, 2016 11:37:23 GMT carlton43 said:

That is very telling David. Is London grossly under-represented because of fast growing population or because you are counting in a very large number of foreigners living there but who are not entitled to vote in HOC elections? Or both?

Both; plus there is more electoral under-registration in London than anywhere else.

Why has Wales been so grossly over-represented?

That's not a bug, that's a feature. Wales has a separate electoral quota from England, which is set significantly lower. In effect Wales is deliberately over-represented because of its separate historical identity; the same thing used to apply to Scotland but no longer does since the creation of the Scottish Parliament. Welsh representation was not reduced at the same time because the National Assembly did not have primary legislative powers. It now does, and there really isn't a justification for the separate quota.

Are the populations of the North-East and the West Midlands actually declining or were they over-represented?

I think it's partly population decline, but also that electoral registration is relatively high.

Thanks David. I thought that would be your take and I agree.  There is real problem in how to base figures as the Census is not purpose built and too infrequent. Registration is dependent on personal response and that has issues for some of you, but not for me. 

[East Anglian Lefty]

Jan 12, 2016 11:33:18 GMT johnsmith said:

A massive increase in London seats, as if we are not dominated anywhere near enough by a metropolitan elite already.

Except that London is rather far from homogenous. The areas with the greatest proportions of unregistered adults aren't going to be the likes of Hampstead and Battersea; they're going to be areas which feel just as excluded from the centres of political power as does the north-east.

[John Chanin]

I've had a look at census data on the basis of those eligible to vote, which is how I think it ought to be done.  England & Wales only as I can't be bothered to chase the Scottish & Irish data.  Based on UK born + others with UK passports + those commonwealth born with no passport(minimal).  Unfortunately this is total population, as constructing a query in NOMIS for those above 18 only is beyond my competence.  The table below shows current seats, percentage of population ineligible, zombie review figures, and revised figures on my basis. 

Region	Existing Seats	% Pop Ineligible	Zombie Review	This Basis
North-East	29	1.8%	26	25
North-West	75	2.5%	68	68
Yorkshire	54	3.0%	50	51
East Midlands	46	3.5%	44	43
West Midlands	59	3.0%	54	54
East	58	4.3%	56	56
London	73	12.1%	68	71
South-East	84	4.4%	83	82
South-West	55	3.0%	53	51
Wales	40	2.2%	30	30
Total	573	4.6%	532	532

Note that there isn't really a lot of difference, with London getting more seats on my basis, despite having a way higher percentage of ineligible voters in the population than anywhere else (as of course we would expect). 

[YL]

Some time ago (in this old thread) I calculated (for England only) ratios of theoretical entitlements working with population compared to registered electorate, based on the 2011 census for the former and the December 2010 electorate for the latter. For example, with 533 seats for England Cambridgeshire's theoretical entitlement based on population is 8.09 seats and its theoretical entitlement based on electorate is 7.80 seats, giving a ratio of 1.04. Over 1 suggests an area benefits from population based apportionment, compared to the rest of England.

"Counties" (ceremonial with some adjustments):

Sept 13, 2012 18:01:06 GMT YL said:

Greater London 1.13
Cambridgeshire 1.04
Bedfordshire 1.03
Berkshire 1.03
West Midlands 1.02
West Yorkshire 1.02
Bristol 1.02
Kent 1.01
Buckinghamshire 1.01
Nottinghamshire 1.00
South Yorkshire 1.00
Lancastrian Merseyside 1.00
Oxfordshire 1.00
Hertfordshire 1.00
Greater Manchester 0.99
Surrey 0.99
East Sussex 0.98
Northamptonshire 0.98
Shropshire 0.98
Wiltshire 0.98
Essex 0.98
Leicestershire & Rutland 0.98
Suffolk 0.97
Lincolnshire 0.97
Hampshire 0.97
Tyne & Wear 0.97
West Sussex 0.97
ex-Cleveland 0.97
Warwickshire 0.97
Herefordshire 0.96
Cheshire & Wirral 0.96
Norfolk 0.96
East Riding 0.96
Lancashire 0.95
Derbyshire 0.95
North Yorkshire 0.95
Northumberland 0.95
Staffordshire 0.95
Devon 0.94
Worcestershire 0.94
Gloucestershire 0.94
Dorset 0.94
Somerset 0.94
Cumbria 0.93
Durham 0.93
Cornwall 0.93
Wight 0.90

London boroughs:

Sept 13, 2012 18:08:19 GMT YL said:

Kensington and Chelsea 1.33
Westminster 1.27
Newham 1.26
Brent 1.24
Haringey 1.23
Hammersmith and Fulham 1.21
Hackney 1.20
Tower Hamlets 1.19
Waltham Forest 1.18
Ealing 1.18
Barking and Dagenham 1.17
Camden 1.16
Lewisham 1.16
Barnet 1.16
Greenwich 1.16
Southwark 1.16
Enfield 1.16
Lambeth 1.15
Hounslow 1.12
Islington 1.10
Merton 1.09
Croydon 1.08
Kingston upon Thames 1.08
Wandsworth 1.07
Richmond upon Thames 1.05
Redbridge 1.05
Harrow 1.05
Hillingdon 1.04
Sutton 1.03
Bexley 0.99
Bromley 0.97
Havering 0.95
City of London 0.90

Yorkshire Metropolitan boroughs:

Sept 13, 2012 18:08:19 GMT YL said:

Bradford 1.15
Sheffield 1.04
Kirklees 1.01
Calderdale 1.00
Leeds 1.00
Doncaster 1.00
Rotherham 0.97
Barnsley 0.95
Wakefield 0.94

[J.G.Harston]

Jan 12, 2016 12:50:38 GMT @goodoldcause said:

This is an interesting exercise, and the difference for London is indeed very striking. Drilling down into the census figures should be illuminating, so I shall not comment further until I have done so.

The first page of my spreadsheet here shows how I calculated the eligable adult population from the census figures for Sheffield.

[cibwr]

Jan 12, 2016 11:37:23 GMT carlton43 said:

That is very telling David. Is London grossly under-represented because of fast growing population or because you are counting in a very large number of foreigners living there but who are not entitled to vote in HOC elections? Or both? Why has Wales been so grossly over-represented? Are the populations of the North-East and the West Midlands actually declining or were they over-represented?

Wales has been over represented due to a couple of of factors,

1. it was not over represented initially, the minimum number of MPs being set at 35, which was roughly proportionate, any slight over representation was in terms of the size of the parliament negligible and could be accepted as slight compensation for the Welsh nation being part of the UK.

2. The number of seats increased as an incremental affect of the formula for calculating the number of seats for each county or preserved county in Wales.

However now the over representation, while still being negligible in terms of the UK parliament is now disproportionate to the size of the Welsh contingent to the UK parliament, and given that we now have a legislature of ours a reduction to 30 or 31 MPs would be reasonable.

[afleitch]

Worth noting is that back in the days of the initial and first reviews, the average electorate of Scottish and Welsh seats was the same as in England. Different rates of population growth has led to an imbalance (corrected in Scotland) which was later justified on the basis of the fact of 'non-devolution'. That's why Northern Irish seats were boosted after the suspension of Stormont.

[Arthur Figgis]

I am Darren Freeman.

[Davıd Boothroyd]

Supreme Court finds using total population is constitutional: www.supremecourt.gov/opinions/15pdf/14-940_ed9g.pdf

[john07]

Jan 16, 2016 19:52:47 GMT afleitch said:

Worth noting is that back in the days of the initial and first reviews, the average electorate of Scottish and Welsh seats was the same as in England. Different rates of population growth has led to an imbalance (corrected in Scotland) which was later justified on the basis of the fact of 'non-devolution'. That's why Northern Irish seats were boosted after the suspension of Stormont.

Well actually they were boosted because of horse trading as part of a deal with the UU to keep the Callaghan government afloat the late 1970s.

Stormont was first suspended in 1972 and then abolished in 1973 by the previous Conservative Government.