• Hi Guest!

    The costs of running this forum are covered by Sea Lion Press. If you'd like to help support the company and the forum, visit patreon.com/sealionpress

Imperial Parliament STV Constituencies

Brainbin

Kingpin of the Cultural Cartel
This is just an idea that I had which may or may not ever come into use for a timeline I may or may not someday write. But for those who don't know, what we now call STV (the Single Transferable Vote) was developed in the mid-nineteenth century in the United Kingdom, and by the end of the century it became very popular with advocates of electoral reform, and was strongly associated with the British Empire in particular (leading to the term "British Proportional Representation" to describe it). Under the name Hare-Clark system it was introduced to Tasmania in 1896. (Even today, when many British politicians and reform advocates refer to "proportional representation" - they are talking about STV, and STV is largely used in Anglosphere countries.) I had the idea to combine it with another popular late-nineteenth-century fancy - the Imperial Federation. So my concept is that the "lower house" of the Imperial Parliament (which I call the "Assembly") elects MIPs under representation by population to multi-member STV constituencies. But how to shape these constituencies? What should their size be? Some guidelines:

- Unless the top-level division (say a province in Canada, a state in Australia, etc.) a constituency serves is insufficiently populous so as to support three members, this should be the minimum, so as to allow a minority voice to represent its constituents. (There's a rich irony here that we'll get to later.)
- No constituency should return more than one per cent of the total Members of the Imperial Parliament. This makes for a nice guideline as to the size of the Assembly itself. It needs to be a multiple of 100 to allow constituencies to elegantly comply with this rule. My suggestion is 700 seats total - slightly higher than the UK Parliament in the early 20th century so of comparable size but not unyieldy, and it means seven is the maximum. Therefore, all constituencies should return no fewer than three members (though some will return two or only one - elected by AV - by necessity) and no more than seven.
- Because of this flexibility, constituencies can more naturally adhere to logical borders. Counties, cities, districts, shires, wards, etc. Nowhere is this more clear than in my proof-of-concept for this project, the County of London.

The population of the Imperial Federation which is the basis for apportionment is derived from the first censuses of the 20th century. I've included all of the pre-WWI White Dominions - yes, South Africa too - though (in classic "white man's burden" logic), those in charge of such things decided that the electorate should not include "insufficiently civilized aboriginal peoples". In Canada, Australia, New Zealand, and Newfoundland, these are a minority. In South Africa they are a large majority, greatly reducing South Africa's representation in the Parliament relative to its total population. But this is the price you pay when you won't enfranchise your people. (Of course at least in theory the plan is for these people to be "civilized" and gain representation in the Imperial Parliament. In theory.)

Therefore, with 700 seats, the following populations can expect the following representation in the Assembly:

England (incl. Wales) - 32,527,843 (432)
England proper - 30,512,831 (405)​
Wales - 2,015,012 (27)​
Scotland - 4,472,103 (59)
Ireland (united island) - 4,458,775 (59)
United Kingdom total - 41,458,721 (550)

Dominion of Canada - 5,371,315 (72)
Dominion of Newfoundland - 217,037 (3)
Commonwealth of Australia - 3,773,248 (50)
Dominion of New Zealand - 815,862 (10)

Cape of Good Hope - 579,541 (8)
Natal - 97,109 (1)
Orange Free State - 142,679 (2)
Transvaal - 297,977 (4)
Union of South Africa total - 1,116,806 (15)

Imperial Federation - 52,752,989 (700)

As you can see, England is sufficiently populous that a majority of MIPs would represent English constituencies. Even several English counties have more people (and thus more MIPs) than most of the Dominions. The County of London, with a population of 4,536,641, has more MIPs than even Scotland or Ireland (who, delightfully, have the exact same number of MIPs). The County of London was reorganized into 28 metropolitan boroughs in 1900, and these should form the guiding boundaries for any STV constituencies. None should cross the boundary lines between the county and its surrounding counties (Middlesex, Essex, Kent, and Surrey) because this is not logical. (I'm also toying with having England's "top-level division" be counties, though a few are too small to support 3 MIPs.)


County of London Imperial Parliament Constituencies.png

London divides its 60 MIPs amongst 11 constituencies. Although these adhere to the boundary lines of the boroughs, they also, wherever possible, evoke the 1868-85 Parliamentary constituency boundaries (and their names) so as to create a sense of continuity. The 11 constituencies are:

I. The Cities of London and Westminster. Population 209,874. 3 MIPs.
II. Chelsea, consisting of the Royal Borough of Kensington and the Boroughs of Hammersmith, Fulham, and Chelsea. Population 500,013. 7 MIPs. Given the special status of Kensington as a Royal Borough, perhaps the constituency might be named "Kensington" instead.
III. Marylebone, consisting of the Boroughs of St Pancras, Hampstead, St Marylebone, and Paddington. Population 550,729. 7 MIPs.
IV. Finsbury, consisting of the Boroughs of Holborn, Finsbury, Islington, and Stoke Newington. Population 547,041. 7 MIPs. Given the geographical and population dominance of Islington within the constituency, perhaps it might be named "Islington" instead of retaining the historical name of "Finsbury".
V. Hackney, consisting of the boroughs of Shoreditch, Bethnal Green, and Hackney. Population 482,298. 7 MIPs.
VI. Tower Hamlets, consisting of the boroughs of Stepney and Poplar. Population 467,386. 6 MIPs.
VII. Southwark, consisting of the boroughs of Bermondsey and Southwark. Population 365,770. 5 MIPs.
VIII. Greenwich, consisting of the boroughs of Deptford, Woolwich, and Greenwich. Population 323,435. 4 MIPs.
IX. Lewisham, consisting of the boroughs of Camberwell and Lewisham. Population 386,718. 5 MIPs.
X. Lambeth. Population 301,873. 4 MIPs. The only borough to be a constituency all on its own, to avoid disturbing the approximate shape of the 1868-85 boundaries without violating the boundaries of the boroughs or the county itself.
XI. Wandsworth, consisting of the boroughs of Wandsworth and Battersea. Population 400,926. 5 MIPs.

Obviously, in the early going the three dominant parties are going to be the Conservatives/Unionists, the Liberals, and Labour (particularly in the East). Whether the STV electoral system means that MIPs from additional, nascent parties - as well as local interest parties, even "London regionalists" - are to be elected - is something worth discussing (and/or writing a TL about), especially for anyone who knows anything about early-20th century British politics. I thought I would start with the County of London because it's easiest to find lower-level boundary census data there. It was much harder for Lancashire, for example.

Anyway, that's what I have of my little side project for now. I'd love to read your thoughts and other comments about it!
 
Last edited:
That's absolutely brilliant.

Have you considered yet what the solution might be for greater enfranchisement later on? Larger Parliament or...?
 
I suspect any Imperial Federation formed after 1900 would have to account for a disproportionate amount of MIPs for the Dominions, but that doesn't detract from how amazing this is. I don't know if this is going to be part of a wider project of mapping out what an Imperial Federation might look like, or if this is a well-done one-off, but I still love it.
 
Wow! What a great response. Thanks for liking and commenting, everyone! I posted this with no idea how well it would go over, and I'm glad to see it went over so well! In a perfect world I would love to map every constituency, but drawing constituencies based on precedent is difficult. This is one of the reasons why I started with London, because it turned out to be ridiculously easy. I do have a spreadsheet working where I can allocate a number of MIPs to each English county (which at this time refers to the administrative counties). Many would consist of only one constituency, which will of course resemble the "good old days" of the Unreformed Parliament, so it will at least be familiar.

For example, Cornwall (population 322,334) would return 4 MIPs. There's no point splitting this further because it's within the "Goldilocks" zone between 3 and 7 MIPs. Devon (population 661,314), on the other hand, would return 9 MIPs, which requires at least two constituencies within it. These could be drawn along the old lines of the two 1832-68 seats, North Devon and South Devon, with minor boundary adjustments to account for changes in the urban and rural districts within Devon after the 1894 review.

This is excellent. I've seen attempts at something similar in the past, but never this detailed or using such contemporary influences such as the older constituency boundaries--usually they just try to work with what currently exists in the Commonwealth countries. Looking forward to more!
It really is funny - seeing the 1868-85 map of Metropolitan London constituencies really brought my idea of what the County of London would look like into much sharper focus. I actually did allocate for each individual borough, but of course most of them would only have returned one or two MIPs, which would have been pointless. That's when I decided to combine boroughs, but had no idea of how. I did briefly consider following the lead of the OTL 1965 reforms, but that made no sense! So I'm hugely relieved to have found the 1868-85 lines.

Have you considered yet what the solution might be for greater enfranchisement later on? Larger Parliament or...?
I don't see Parliament greatly increasing in size. 700 is very large already - for OTL comparison, the European Parliament only seats 751 MEPs and even the Lok Sabha in India only allows for 552 MPs. Currently approximately 75,000 people are served by each MIP - a relatively small number. Since the maximum size of constituencies is tied to the total number of MIPs I can't see the larger, more established parties favouring an increase and thus a reduction in the threshold required to elect an MIP. (To be elected to a seven-seat constituency a candidate needs to receive at least 12.5% +1 of the counted votes.)

Now, as far as enfranchisement goes, my theory is that each "realm" (England, Scotland, and Ireland are each a "realm" of their own - the UK is basically a legal fiction in this Imperial Federation scenario) chooses which of its people to enfranchise. Of course, in New Zealand women have the vote and thus I imagine Kiwi politicians would demand double the MIPs to account for that. Since we're on the verge of women's suffrage anyway I figured that might be the catalyst needed to give women the vote throughout the Federation - at least on the Imperial level. As far as South Africa is concerned, I imagine the franchise will be extended gradually - to Coloureds and Asians first, and then to Blacks. It's far more dependent on how South African politics develop over the next few decades. It will be an enshrined goal of the Imperial Federation to "civilize" the peoples of the Empire, for better and (mostly) for worse. It's definitely something a TL would cover in considerable detail.

I suspect any Imperial Federation formed after 1900 would have to account for a disproportionate amount of MIPs for the Dominions, but that doesn't detract from how amazing this is. I don't know if this is going to be part of a wider project of mapping out what an Imperial Federation might look like, or if this is a well-done one-off, but I still love it.

To get the elephant out of the way, obviously India (the Raj as a whole, really) is never joining this Imperial Federation. It's unworkable. (And I would never want to draw that.)

That said, yes, eventually the UK is going to have a diminished presence within the Imperial Federation. Based on 1901 (1904 for South Africa) census figures it accounts for a whopping 78.5% (!) of the voting population - one reason many within the UK are probably willing to sign on to a Federation ITTL, look how dominant the UK is - but that will decline rapidly over the coming decades even if other Dominions aren't admitted.

What the UK will hope to do is export its cultural and sociopolitical values (and its emigrants) to the other realms - and to be fair, in the early 20th century it did a very good job of this. That's something I would want to explore in a TL, of course.

Thanks again for your comments, everyone! I'll try to get the allocation of MIPs for all the English counties posted soon.
 
Last edited:
I think this is great, the best example of this idea I’ve seen so far and it uses STV in a plausible and workable way. I’m looking forward to seeing more.

Just one question - you suggest NZ wouldn’t have enfranchised its Maori population, I don’t think that would be the case in your TL as it wasn’t in ours, indeed a Maori MIP would be perfectly possible and could become an influential spokesman for all indigenous people in the Empire.
 
I think this is great, the best example of this idea I’ve seen so far and it uses STV in a plausible and workable way. I’m looking forward to seeing more.
Thank you!

SteveBP said:
Just one question - you suggest NZ wouldn’t have enfranchised its Maori population, I don’t think that would be the case in your TL as it wasn’t in ours, indeed a Maori MIP would be perfectly possible and could become an influential spokesman for all indigenous people in the Empire.
Thank you for the correction, I have duly restored the Maori population to New Zealand's total. I've also done some digging into New Zealand's population at the 1901 census and - conveniently for me - the two main islands had virtually identical populations. Therefore, it's very easy to split New Zealand's ten seats into two constituencies: North Island, with 5 MIPs, and South Island, also with 5 MIPs. Now whether either constituency can elect a Maori MIP is another matter. The Maori population in 1901 was only 43,000 (against 773,000 whites). That gives them an electoral strength of 5.27%. The victorious candidate in a five-member constituency requires 16.67% + 1 of all votes counted to emerge victorious. One of the four Maori MPs from the Maori electorates might be willing to make the jump to Imperial politics, but would they be able to capture the votes they need to win election? That would remain to be seen. I'd be interested in your thoughts.

An Imperial Federation with an allied Dominion of India isn't something I've seen before I have to say.
An allied Empire of India, actually. My idea is that the Raj becomes fully independent of the British Empire but remains in close association - including personal union. (That way the Sovereign is a Dual Emperor - which trumps any of his fellows.) Similar to the OTL situation post-Statute of Westminster, actually. Preferential trade links and a military alliance would still exist but otherwise India would have its own agenda. This is very loosely sketched out and probably happens about a half-century after my planned POD, so I can't go into too much detail.

As promised, here is a list of how seats in England would be apportioned by county (the top-level division in England) per the 1901 census:

Bedfordshire - 171,707 - 2 MIPs
Berkshire - 252,571 - 3 MIPs
Buckinghamshire - 197,046 - 3 MIPs
Cambridgeshire proper - 120,264 - 2 MIPs
Isle of Ely (Cambridgeshire) - 64,495 - 1 MIP
Cheshire - 827,191 - 11 MIPs
Cornwall - 322,334 - 4 MIPs
Cumberland - 266,933 - 4 MIPs
Derbyshire - 610,522 - 8 MIPs
Devon - 662,196 - 9 MIPs
Dorset - 202,063 - 3 MIPs
County Durham - 1,187,474 - 16 MIPs
Essex - 1,083,998 - 14 MIPs
Gloucestershire - 708,439 - 9 MIPs
Hampshire proper - 717,164 - 10 MIPs
Isle of Wight (Hampshire) - 82,418 - 1 MIP
Herefordshire - 114,125 - 1 MIP
Hertfordshire - 258,423 - 3 MIPs
Huntingdonshire - 54,125 - 1 MIP
Kent - 961,139 - 13 MIPs
Lancashire - 4,387,043 - 58 MIPs
Leicestershire - 437,490 - 6 MIPs
Lincolnshire, Holland - 77,610 - 1 MIP
Lincolnshire, Kesteven - 103,962 - 1 MIP
Lincolnshire, Lindsey - 318,450 - 4 MIPs
County of London - 4,536,541 - 60 MIPs
Middlesex - 792,314 - 11 MIPs
Norfolk - 476,553 - 6 MIPs
Northamptonshire (proper) - 294,506 - 4 MIPs
Soke of Peterborough (Northamptonshire) - 41,122 - 1 MIP
Northumberland - 603,119 - 8 MIPs
Nottinghamshire - 514,459 - 7 MIPs
Oxfordshire - 186,460 - 2 MIPs
Rutland - 19,709 - 0 MIPs (yes, that's right, Rutland isn't populous enough to return even one MIP).
Shropshire - 239,783 - 3 MIPs
Somerset - 434,950 - 6 MIPs
Staffordshire - 1,236,919 - 16 MIPs
Suffolk, East - 255,800 - 3 MIPs
Suffolk, West - 117,553 - 2 MIPs
Surrey - 653,549 - 9 MIPs
Sussex, East - 450,702 - 6 MIPs
Sussex, West - 151,553 - 2 MIPs
Warwickshire - 939,904 - 12 MIPs
Westmorland - 64,409 - 1 MIP
Wiltshire - 271,394 - 4 MIPs
Worcestershire - 453,734 - 6 MIPs
Yorkshire, East Riding - 385,007 - 5 MIPs
Yorkshire, North Riding - 377,338 - 5 MIPs
Yorkshire, West Riding - 2,750,493 - 37 MIPs
City of York - 77,914 - 1 MIP

I've omitted Monmouthshire, which I'm going to treat as part of Wales despite its thorny and ambiguous status at this time (the official 1901 census, from which I derive these population figures, considers Monmouthshire part of England).

I'm also unclear on the status of the City of York, which doesn't appear to have been an administrative county of its own but is listed separately in the census from all three ridings of Yorkshire (each of which formed an administrative county), presumably due to its independent history.

Many counties do fall within the "Goldilocks range" of 3-7 MIPs, as you can see. Of those that are too small, several used to form part of a county which, when recombined into a united whole, is the right size (Cambridgeshire, Lincolnshire, Northamptonshire, and Suffolk). However, Hampshire and Sussex, were they reunited, would be too large to form a single constituency but can't divide along these "natural" new boundaries as one of the component parts is insufficiently populous, although perhaps I could let that slide.

And a couple of counties are just plain too small and I find it hard to justify letting them form their own constituency - Herefordshire, Huntingdonshire, Westmorland, and of course Rutland. It should surprise no one to learn that all four of these counties have subsequently ceased to exist IOTL - although both Herefordshire and (shockingly) Rutland have since been re-created. I'd have to put my finger on the scale to give Rutland an MIP at all, which means another county loses a seat (Middlesex, according to my spreadsheet - a rounding error meant that I already had to cut Herefordshire down from two to one to make sure England "only" has 405 MIPs.)

Scotland and Ireland should be next. Each will return only 59 MIPs, so having counties be the top-level division in their case would probably be impractical, but I'll run them through to see how it looks. In Ireland's case it makes far more sense for the top-level division to be the four provinces, so I'll see how that goes as well. As for Scotland? I'm open to suggestions.

As for maps, more will follow when I get good boundary lines for c. 1900. I've got some promising leads. Someone on Wikipedia actually made a beautiful map of every urban and rural district and borough in England... in 1931. Still, it's a good place to start!
 
As for maps, more will follow when I get good boundary lines for c. 1900. I've got some promising leads. Someone on Wikipedia actually made a beautiful map of every urban and rural district and borough in England... in 1931. Still, it's a good place to start!

It's unfortunately inaccurate in a lot of places. I've noticed a lot of borders which didn't come into force until 1935, and Alvaston and Boulton UDC was a going concern until 1932.

I'd suggest attaching Rutland to Kesteven in this situation.

Really annoyingly, Vision of Britain doesn't include the 1901 census for some reason.
 
When you get to Canada, are you going to do it by province and/or its Senate allocation of seats redux? (I mention this since if the latter is the case, then in Québec's case you'd have to use the actual Senate districts - Québec is the only province which per the Constitution actually has clearly-defined ridings for the Senate, theoretically to ensure equality between English-Canadians and French-Canadians.)
 
It's unfortunately inaccurate in a lot of places. I've noticed a lot of borders which didn't come into force until 1935, and Alvaston and Boulton UDC was a going concern until 1932.
Thanks for letting me know! I will say the Vision of Britain site you referenced has some very promising maps which might help me to draw some of the other counties. Well, that and another website I found - histpop.org - which appears to have archived the entire 1901 census. With luck I can go through the horrifyingly daunting index section to determine the population of each and every rural and urban district in the United Kingdom - or at least England and Wales - c. 1901. I think I should try starting with a less populous county first.

Alex Richards said:
I'd suggest attaching Rutland to Kesteven in this situation.
The more I think about it, I really think it seems more English to allow for exceptions based on tradition and just let the smaller counties have their single-member constituencies. But I'm willing to allow for further deliberation on this.

I'm really enjoying this project, Brainbin.
Thank you very much! :)

When you get to Canada, are you going to do it by province and/or its Senate allocation of seats redux? (I mention this since if the latter is the case, then in Québec's case you'd have to use the actual Senate districts - Québec is the only province which per the Constitution actually has clearly-defined ridings for the Senate, theoretically to ensure equality between English-Canadians and French-Canadians.)
Canada's top-level division will be the provinces. (The territories - as is the case during this time on the federal level - will not be represented in the Imperial Parliament.) I've already crunched the numbers on this and the only two provinces which will return more than 7 MIPs based on the 1901 census are Ontario and Quebec. (Ontario will return 28 MIPs, and Quebec will return 21 MIPs.) Using Quebec's Senate designations is actually a really clever idea... if only there weren't more of them (24) than the number of MIPs Quebec will be sending to the Assembly.

And now a special treat for you all, the Province of Connaught:

Province of Connaught Imperial Parliament Constituencies.png


The Province of Connaught (population 649,635) returns 9 MIPs to the Assembly, and these are divided between three constituencies, each of which elects 3 MIPs:

Galway (population 192,146)
Mayo (population 202,627)
Northeast Connaught - consisting of Counties Leitrim, Roscommon, and Sligo (population 254,862) - I'm not married to the name, BTW.

Surprisingly enough, despite being the least-populated province of Ireland, Connaught was also the easiest to divide neatly. Go figure!
 
I assume the three counties of NE Connaught just aren't large enough to justify their own single member seats?

(A campaign to split the constituency would be a fun side thing going on really).

You could always go the Scottish route and just call it 'Leitrim, Roscommon and Sligo', or Leitrim — Roscommon — Sligo for the Canadian version.
 
Thank you!

Thank you for the correction, I have duly restored the Maori population to New Zealand's total. I've also done some digging into New Zealand's population at the 1901 census and - conveniently for me - the two main islands had virtually identical populations. Therefore, it's very easy to split New Zealand's ten seats into two constituencies: North Island, with 5 MIPs, and South Island, also with 5 MIPs. Now whether either constituency can elect a Maori MIP is another matter. The Maori population in 1901 was only 43,000 (against 773,000 whites). That gives them an electoral strength of 5.27%. The victorious candidate in a five-member constituency requires 16.67% + 1 of all votes counted to emerge victorious. One of the four Maori MPs from the Maori electorates might be willing to make the jump to Imperial politics, but would they be able to capture the votes they need to win election? That would remain to be seen. I'd be interested in your thoughts.

It's a stretch especially given that Maori had disproportionatly few seats compared to others when electorates were drawn up, but maybe there could be a Maori seat created... Another option might be a Maori politician with a Pākeha name and ancestry slipping through the net. He might not be that sympathetic to the global indiginous cause though...
 
This looks great, I'm not familiar with the 1901 boundaries, I am surprised that County Durham has over a million people, I can only guess it includes that Sunderland/Gateshead/ Middlebourgh as part of it, which would explain my Northumbria is quite small
 
This looks great, I'm not familiar with the 1901 boundaries, I am surprised that County Durham has over a million people, I can only guess it includes that Sunderland/Gateshead/ Middlebourgh as part of it, which would explain my Northumbria is quite small

Yep, Durham at the time went from the Tyne to the Tees.
 
Hope you don't mind, but I think I've got a 4:4 Derbyshire North/South split that's pretty workable:

Derbyshire North: Bakewell Urban, Baslow and Bubnell, Blackwall, Bolsover, Bonsall, Brampton and Walton, Buxton, Chesterfield Urban, Clay Cross, Dronfield, Fairfield, Glossop, Matlock, Matlock Bath and Scarthin Nick, Newbold and Dunston, New Mills, North Darley, South Darley, Whittington, Wirksworth, Bakwell Rural, Blackwell, Chapel-en-le-Frith, Chesterfield Rural, Clowne, Glossop Dale, Hayfield, Norton, Sudbury: Population: 292,822

Derbyshire South: Alfreton, Alvaston and Boulton, Ashbourne Urban, Belper Uban, Derby, Heage, Heanor, Ilkeston, Long Eaton, Ripley, Swadlincote, Ashbourne Rural, Belper Rural, Hartshorne and Seals, Repton, Shardlow, Basford (Notts administration): Population: 317,700
 
This is very interesting.

An allied Empire of India, actually. My idea is that the Raj becomes fully independent of the British Empire but remains in close association - including personal union. (That way the Sovereign is a Dual Emperor - which trumps any of his fellows.) Similar to the OTL situation post-Statute of Westminster, actually. Preferential trade links and a military alliance would still exist but otherwise India would have its own agenda. This is very loosely sketched out and probably happens about a half-century after my planned POD, so I can't go into too much detail.

That strikes me as very plausible, far more than those ideas of malapportioning Indian seats or of the white ruling class just letting India have a majority of seats. But just a thought on that idea - how does decolonization go? Judging by how India’s own nationalist movement has more or less achieved its 1920s goals, that would probably embolden other nationalist movements. It strikes me that any India would have strong sympathies towards the independence movements, strong enough that sharing a head of state may be seen as tacit acceptance of the colonial order. So, even a less messy Indian independence is, in my view, bound to end up with friction between India and its former colonial masters, which may very well be enough for India to become a republic.
 
Back
Top