Crimson King
Lib Dem
Be nice to each other and sing in tune
Posts: 9,859 
Member is Online
|
Post by Crimson King on Nov 30, 2014 22:45:55 GMT
Now that's an interesting idea
|
|
|
|
Post by carlton43 on Nov 30, 2014 23:55:28 GMT
If you want there to be a significant "backbench" contingent of MPs who are not on, or aspiring to be on, the government payroll vote, then Yes. One of the reasons why I support having about 2,000 MPs. All elected by FPTP in small single member constituencies. What would be your fixed limit to members of the government down to the lowest level of unpaid 'bag carrier'? I think we need a fixed limit or it will forever expand. Also a fixed upper limit to number of departments of state. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 1, 2014 0:56:07 GMT
Now that's an interesting idea We can retain (a very close approximation to) 100% proportionality while having 650 single member constituencies with no top up lists. Easy. Draw one vote as the winner out of the hat. Whoever that person voted for has won. If the vote is not valid a 2nd one should then be drawn. It can be done on a party only basis and then the party chooses who represents the seat. Problem solved. It would make election night rather fun too. |
|
|
|
Post by johnloony on Dec 1, 2014 1:57:52 GMT
Now that's an interesting idea We can retain (a very close approximation to) 100% proportionality while having 650 single member constituencies with no top up lists. Easy. Draw one vote as the winner out of the hat. Whoever that person voted for has won. If the vote is not valid a 2nd one should then be drawn. It can be done on a party only basis and then the party chooses who represents the seat. Problem solved. It would make election night rather fun too. You say "easy" but in order for it to work, there would have to be a proper and rigorous method of doing the verb "draw" which ensures that it is a genuinely random choice, and also properly define "the hat". |
|
J.G.Harston
Lib Dem
Leave-voting Brexit-supporting Liberal Democrat
Posts: 14,804
|
Post by J.G.Harston on Dec 1, 2014 3:17:33 GMT
We can retain (a very close approximation to) 100% proportionality while having 650 single member constituencies with no top up lists. Easy. Draw one vote as the winner out of the hat. Whoever that person voted for has won. If the vote is not valid a 2nd one should then be drawn. There's an Asimov short story based on that system. Multivac picked The Voter at random (weighted by various factors), and whoever The Voter voted for was President. |
|
|
|
Post by greenchristian on Dec 4, 2014 23:21:53 GMT
We can retain (a very close approximation to) 100% proportionality while having 650 single member constituencies with no top up lists. Easy. Draw one vote as the winner out of the hat. Whoever that person voted for has won. If the vote is not valid a 2nd one should then be drawn. There's an Asimov short story based on that system. Multivac picked The Voter at random (weighted by various factors), and whoever The Voter voted for was President. That story is actually based on a true story from the 1952 US Presidential election. CBS news used UNIVAC 1 - one of the earliest computers - to analyse the election returns. Conventional wisdom said that Adlai Stevenson would win. However, UNIVAC predicted an Eisenhower landslide from the first 1% of votes. The TV station were too embarrassed to admit that result at first. It wasn't until quite late on election night that they confessed what the computer had originally predicted. Asimov simply extrapolated the success to the ultimate degree - the computer (Multivac) finds the perfectly average voter and then asks him lots of questions about various issues, using that to project what the result would have been if they had actually held a national election. Multivac doesn't actually ask him who he would vote for. |
|
john07
Labour & Co-operative
Posts: 15,818
|
Post by john07 on Dec 5, 2014 15:42:26 GMT
The above looks rather like the 'as goes Maine goes the nation' adage because of its reputation as a bellwether state. If that were always true then the lection could be confined to Maine and everyone else could stay at home!
In 1936 the Republicans carried Maine and Vermont but no other state giving Alf Landon only seven electoral votes.
It was summarized "as goes Maine so goes Vermont'.
For the same election the Literary Digest ran (one of) the first opinion polls with a sample of 10 million that predicted Landon would beat Roosevelt by 55% to 42%. The actual election was 61% to 37% the other way.
|
|
|
|
Post by Davıd Boothroyd on Dec 5, 2014 15:54:41 GMT
The point about Maine was that until 1957, Maine held its general election in September - two months ahead of the rest of the country, so Maine's results were known well in advance of everywhere voting. And up to 1936, Maine had been a reasonably good bellwether predicting the result of a Presidential election.
|
|
|
|
Post by psephos on Dec 5, 2014 18:37:57 GMT
There's an Asimov short story based on that system. Multivac picked The Voter at random (weighted by various factors), and whoever The Voter voted for was President. That story is actually based on a true story from the 1952 US Presidential election. CBS news used UNIVAC 1 - one of the earliest computers - to analyse the election returns. Conventional wisdom said that Adlai Stevenson would win. However, UNIVAC predicted an Eisenhower landslide from the first 1% of votes. The TV station were too embarrassed to admit that result at first. It wasn't until quite late on election night that they confessed what the computer had originally predicted. Asimov simply extrapolated the success to the ultimate degree - the computer (Multivac) finds the perfectly average voter and then asks him lots of questions about various issues, using that to project what the result would have been if they had actually held a national election. Multivac doesn't actually ask him who he would vote for. How true is this? Wikipedia (which is never wrong) says "Throughout the entire campaign, Eisenhower led in all opinion polls, and by wide margins in most of them." and that is borne out: en.wikipedia.org/wiki/Historical_polling_for_U.S._Presidential_elections#United_States_presidential_election.2C_1952But one of the online pieces on Univac instead says "Pre-election polls had predicted anything from a Democratic landslide to a tight race with the Demo candidate, Illinois Gov. Adlai Stevenson, slightly ahead of the Republican, five-star Gen. Dwight D. Eisenhower, Supreme Commander of Allied Forces in Europe in World War II." www.wired.com/2010/11/1104cbs-tv-univac-election/ and that seems to be the received wisdom on Univac. Given this was Ike, the great universally respected war hero who in the end took everything but the deep south, I can't believe for one minute the polls all said Stevenson would win and Univac confounded them. There must be more than meets the eye. |
|
|
|
Post by greenchristian on Dec 5, 2014 23:09:29 GMT
That story is actually based on a true story from the 1952 US Presidential election. CBS news used UNIVAC 1 - one of the earliest computers - to analyse the election returns. Conventional wisdom said that Adlai Stevenson would win. However, UNIVAC predicted an Eisenhower landslide from the first 1% of votes. The TV station were too embarrassed to admit that result at first. It wasn't until quite late on election night that they confessed what the computer had originally predicted. Asimov simply extrapolated the success to the ultimate degree - the computer (Multivac) finds the perfectly average voter and then asks him lots of questions about various issues, using that to project what the result would have been if they had actually held a national election. Multivac doesn't actually ask him who he would vote for. How true is this? Wikipedia (which is never wrong) says "Throughout the entire campaign, Eisenhower led in all opinion polls, and by wide margins in most of them." and that is borne out: en.wikipedia.org/wiki/Historical_polling_for_U.S._Presidential_elections#United_States_presidential_election.2C_1952But one of the online pieces on Univac instead says "Pre-election polls had predicted anything from a Democratic landslide to a tight race with the Demo candidate, Illinois Gov. Adlai Stevenson, slightly ahead of the Republican, five-star Gen. Dwight D. Eisenhower, Supreme Commander of Allied Forces in Europe in World War II." www.wired.com/2010/11/1104cbs-tv-univac-election/ and that seems to be the received wisdom on Univac. Given this was Ike, the great universally respected war hero who in the end took everything but the deep south, I can't believe for one minute the polls all said Stevenson would win and Univac confounded them. There must be more than meets the eye. My understanding is that it's as true as that newspaper with the Dewey beats Truman headline. Polling back in the 50s was far less accurate than it is today, due to the absence of statistical weighting, and the fact that some polls were taken by telephone in the days when large segments of the population didn't have one. |
|
john07
Labour & Co-operative
Posts: 15,818
|
Post by john07 on Dec 6, 2014 13:47:28 GMT
How true is this? Wikipedia (which is never wrong) says "Throughout the entire campaign, Eisenhower led in all opinion polls, and by wide margins in most of them." and that is borne out: en.wikipedia.org/wiki/Historical_polling_for_U.S._Presidential_elections#United_States_presidential_election.2C_1952But one of the online pieces on Univac instead says "Pre-election polls had predicted anything from a Democratic landslide to a tight race with the Demo candidate, Illinois Gov. Adlai Stevenson, slightly ahead of the Republican, five-star Gen. Dwight D. Eisenhower, Supreme Commander of Allied Forces in Europe in World War II." www.wired.com/2010/11/1104cbs-tv-univac-election/ and that seems to be the received wisdom on Univac. Given this was Ike, the great universally respected war hero who in the end took everything but the deep south, I can't believe for one minute the polls all said Stevenson would win and Univac confounded them. There must be more than meets the eye. My understanding is that it's as true as that newspaper with the Dewey beats Truman headline. Polling back in the 50s was far less accurate than it is today, due to the absence of statistical weighting, and the fact that some polls were taken by telephone in the days when large segments of the population didn't have one. That was the reason why the Literary Digest poll in 1936 went so spectacularly wrong. It had a huge sample but was based on subscribers to the magazine and Bell Telephone customers. That subsample of the electorate would have been strongly weighted to the Republicans. |
|
|
|
Post by Pete Whitehead on May 16, 2015 14:08:15 GMT
2010 notional results (assuming everybody cast their vote the same way, which they wouldn't have done obvs) | | Total seats | Con | Lab | LD | Nat | | | | | | | | | Highlands & Islands | 5 | 1 | 1 | 2 | 1 | | Aberdeen & Angus | 9 | 2 | 2 | 2 | 3 | | Forth | 7 | 1 | 4 | 1 | 1 | | Strathclyde | 6 | 1 | 3 | 1 | 1 | | Glasgow | 6 | 0 | 4 | 1 | 1 | | Lanarkshire | 7 | 0 | 5 | 1 | 1 | | Ayrshire & Galloway | 6 | 2 | 3 | 0 | 1 | | Lothian | 5 | 1 | 2 | 1 | 1 | | Edinburgh | 5 | 1 | 2 | 2 | 0 | | | | | | | | | Scotland | 56 | 9 | 26 | 11 | 10 | | | | | | | | | Clwyd-Gwynedd | 7 | 2 | 3 | 1 | 1 | | Dyfed-Powys | 5 | 2 | 1 | 1 | 1 | | Glamorgan | 9 | 1 | 5 | 2 | 1 | | Cardiff | 5 | 2 | 2 | 1 | 0 | | Gwent | 6 | 2 | 3 | 1 | 0 | | | | | | | | | Wales | 32 | 9 | 14 | 6 | 3 | | | | | | | | | Northumberland | 8 | 2 | 4 | 2 | | | Tyne & Wear | 7 | 2 | 4 | 1 | | | Durham | 7 | 1 | 4 | 2 | | | Cleveland | 6 | 2 | 3 | 1 | | | | | | | | | | Cumbria | 6 | 3 | 2 | 1 | | | North Lancashire | 8 | 4 | 3 | 1 | | | East Lancashire | 6 | 3 | 2 | 1 | | | West Lancashire | 5 | 2 | 2 | 1 | | | Liverpool | 6 | 0 | 5 | 1 | | | Warrington-St Helens | 5 | 1 | 3 | 1 | | | Bolton-Bury-Wigan | 8 | 3 | 4 | 1 | | | Oldham-Rochdale | 7 | 2 | 3 | 2 | | | Manchester | 10 | 2 | 5 | 3 | | | East Cheshire | 7 | 3 | 2 | 2 | | | West Cheshire | 7 | 3 | 3 | 1 | | | | | | | | | | North Yorkshire | 9 | 5 | 2 | 2 | | | East Yorkshire | 6 | 2 | 2 | 2 | | | Leeds | 8 | 3 | 3 | 2 | | | Bradford-Halifax | 7 | 3 | 3 | 1 | | | Wakefield-Huddersfield | 8 | 3 | 4 | 1 | | | Barnsley-Doncaster-Rotherham | 8 | 2 | 5 | 1 | | | Sheffield | 5 | 1 | 2 | 2 | | | | | | | | | | Lindsey | 6 | 3 | 2 | 1 | | | Lincoln, Holland & Kesteven | 5 | 3 | 1 | 1 | | | North Nottinghamshire | 5 | 2 | 2 | 1 | | | Nottingham | 6 | 2 | 3 | 1 | | | North Derbyshire | 5 | 2 | 2 | 1 | | | Derby | 6 | 3 | 2 | 1 | | | Leicester | 5 | 2 | 2 | 1 | | | Leicestershire & Rutland | 6 | 3 | 1 | 2 | | | Northamptonshire | 7 | 4 | 2 | 1 | | | | | | | | | | Warwickshire | 6 | 3 | 2 | 1 | | | Coventry & Solihull | 5 | 2 | 2 | 1 | | | Birmingham | 10 | 3 | 5 | 2 | | | Wolverhamton & Walsall | 5 | 2 | 2 | 1 | | | Sandwell & Dudley | 7 | 3 | 3 | 1 | | | South Staffordshire | 7 | 4 | 2 | 1 | | | North Staffordshire | 5 | 2 | 2 | 1 | | | Shropshire | 5 | 3 | 1 | 1 | | | Hereford & Worcester | 8 | 5 | 1 | 2 | | | | | | | | | | Norfolk | 9 | 4 | 2 | 3 | | | Suffolk | 8 | 4 | 2 | 2 | | | Cambridgshire | 8 | 4 | 1 | 3 | | | North Essex | 8 | 5 | 1 | 2 | | | South Essex | 10 | 6 | 2 | 2 | | | East Hertfordshire | 6 | 4 | 1 | 1 | | | West Hertfordshire | 6 | 3 | 1 | 2 | | | Bedfordshire | 6 | 3 | 2 | 1 | | | | | | | | | | Buckinghamshire | 8 | 5 | 1 | 2 | | | Berkshire | 9 | 5 | 2 | 2 | | | Oxfordshire | 7 | 4 | 1 | 2 | | | East Surrey | 5 | 4 | 0 | 1 | | | West Surrey | 6 | 4 | 0 | 2 | | | East Kent | 8 | 4 | 2 | 2 | | | West Kent | 10 | 6 | 2 | 2 | | | East Sussex | 8 | 4 | 2 | 2 | | | West Sussex | 9 | 5 | 1 | 3 | | | North Hampshire | 6 | 4 | 0 | 2 | | | West Hampshire | 7 | 4 | 1 | 2 | | | South Hampshire | 7 | 4 | 1 | 2 | | | | | | | | | | Gloucestershire | 7 | 4 | 1 | 2 | | | Bristol | 7 | 2 | 2 | 3 | | | Wiltshire | 7 | 4 | 1 | 2 | | | Dorset | 8 | 4 | 1 | 3 | | | Somerset | 10 | 4 | 1 | 5 | | | North Devon | 7 | 3 | 1 | 3 | | | South Devon | 5 | 2 | 1 | 2 | | | Cornwall | 6 | 3 | 0 | 3 | | | | | | | | | | West Essex | 9 | 4 | 4 | 1 | | | London East Inner | 7 | 1 | 5 | 1 | | | London Central | 9 | 3 | 4 | 2 | | | East Middlesex | 8 | 3 | 3 | 2 | | | Central Middlesex | 8 | 3 | 4 | 1 | | | West Middlesex | 6 | 3 | 2 | 1 | | | London South | 9 | 4 | 2 | 3 | | | London South Inner | 8 | 2 | 4 | 2 | | | London South East | 10 | 5 | 3 | 2 | | | | | | | | | | England | 545 | 243 | 171 | 131 | | | | | | | | | | GB | 633 | 261 | 211 | 148 | 13 |
| | Total Seats | DUP | UUP | .SF... | SDLP | ALL | | | | | | | | | | Antrim | 7 | 3 | 1 | 1 | 1 | 1 | | Down & Armagh | 5 | 2 | 1 | 1 | 1 | 0 | | Londonderry, Fermanagh & Tyrone | 5 | 1 | 0 | 3 | 1 | 0 | | | | | | | | | | Northern Ireland | 17 | 6 | 2 | 5 | 3 | 1 |
Same system based on 2015 election Con 274 +13 Lab 225 +14 UKIP 72 +72 SNP 33 +23 LD 26 -122 PC 3 nc NI 17 Surprisingly no seats for the Greens - they just missed out on the last seat in East Sussex. Labour won all 6 seats in Liverpool. Lib Dems win no seats in the East and West Midlands |
|
iain
Lib Dem
Posts: 11,451
|
Post by iain on May 16, 2015 18:19:07 GMT
Where would the LD seats be?
|
|
|
|
Post by Pete Whitehead on May 16, 2015 20:08:11 GMT
3 in Scotland (2 in Highlands & Islands and 1 in Aberdeenshire & Angus)
1 in Mid Wales
2 in the NW (1 in East Cheshire, 1 in Cumbria)
2 in Yorkshire (1 in North Yorkshire, 1 in Sheffield)
4 in East of England (1 each in Cambridgeshire, Norfolk, North Essex, West Hertfordshire)
4 in the South East (1 each in Oxfordshire, East Sussex, West Sussex, North Hampshire)
8 in the South West (2 in Somerset, 1 each in Bristol, Gloucestershire, Wiltshire, Dorset, North Devon, Cornwall)
2 in London (1 in London South and 1 in North Surrey)
|
|
|
|
Post by manchesterman on May 17, 2015 3:48:28 GMT
Fantastic work Pete.
I'm a big fan of d'Hondt myself.
Those results show a far greater degree of proportionality (fairness in relation to votes cast) whilst retaining a local MP presence, albeit not a constituency MP in the strictest sense
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 19, 2015 18:16:19 GMT
Suggestion (for Scotland): * Separate region of Borders (D&G with Scottish Borders given historic links) * Lothian including Edinburgh * Aberdeenshire/Grampian - Aberdeen and Aberdeenshire * Tayside - Perthshire, Angus, Dundee and Stirling * Strathclyde and Ayrshire - group Ayrshire with Strathclyde Or just use a modified version of the Holyrood regions that don't cross council boundaries: Highlands & Islands (Highland, Argyll & Bute, Moray, Na h-Eileanan an Iar, Orkney, Shetland) North East Scotland (Dundee, Angus, Aberdeen, Aberdeenshire) Mid-Scotland & Fife (Stirling, Clackmannanshire, Perth & Kinross, Fife) West Scotland (East Dunbartonshire, West Dunbartonshire, East Renfrewshire, Renfrewshire, Inverclyde, North Ayrshire) Glasgow (Glasgow) Central Scotland (North Lanarkshire, South Lanarkshire, Falkirk) Lothian (Edinburgh, Midlothian, West Lothian) South Scotland (East Lothian, Scottish Borders, Dumfries & Galloway, East Ayrshire, South Ayrshire) |
|
|
|
Post by Pete Whitehead on May 20, 2015 8:59:57 GMT
My proposed constituencies don't cross council boundaries in Scotland or elsewhere (with the sole exception of Richmond upon Thames). They do therefore cross constituency boundaries making calculations of the results imprecise. There are probably numerous different permutations of Scottish boundaries that can be done due to the frequency with which counties/regions have chopped and changed in the last 40 years. I have chosen to limit constituency sizes so that they have not fewer than 5 seats and not more than 10. I'm not sure if some of the alternatives proposed would meet those criteria but of course people may propose whatever they like and use whatever criteria they like. Part of the rationale for drawing my boundaries the way I did was that they also gave some recognition to historic county boundaries |
|
|
|
Post by helenb on May 26, 2015 17:32:44 GMT
Introductory post from a lurker here. I've played around with this kind of idea many times in the past. Personally I prefer STV in 3/4/5 member seats, but I can see mileage in allowing 4/5/6/7 members per seat. I did some analysis on 2010 and 2015 (using d'Hondt within 3/4/5 member seats based on current constituencies) published at challengingjourneys.wordpress.com/2015/05/19/just-look-what-you-could-have-won/, but I didn't get round to doing a map. In summary: 2010 - CON 264, LAB 205, LD 150, SNP 9, PC 2, Green 1 2015 - CON 298, LAB 236, UKIP 39, SNP 35, LD 18, PC 2, Green 2 Trying to balance out MPs against local authority boundaries is pretty difficult because of the tricky fractions (South Yorkshire 13.51 MPs, Tyne & Wear 11.48, Berkshire 8.36) and I've not yet found a good way to try to work out an optimum number of MPs which minimises these ugly fractions. So doing the same exercise trying to create multi-member constituencies where each MP has roughly the same number of electors involves lots of crossing of boundaries. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 27, 2015 10:09:05 GMT
Helen - a female member *and* a boundary geek? Welcome!
|
|
|
|
Post by Pete Whitehead on May 27, 2015 10:32:14 GMT
Nice piece of work - thanks for sharing
|
|
