Post by therealriga on May 24, 2016 17:35:28 GMT
I've been doing this on and off for months during slack periods at work. Calculating hypothetical results at general elections from 1983 to 2015 under a single transferable vote system. 1983 was the starting point because I had all the results and maps from then on (don't know where to find constituency maps for 1979 and earlier?) Of course, the probability of Thatcher introducing STV in her first term is probably the same as George Galloway, Dennis Skinner and Corbyn all announcing their joint defection to the Conservatives.
District magnitude was 3 to 5 members (with Western Isles, Orkney&Shetland and Isle of Wight as exceptional single member constituencies under AV.) Respecting local authority boundaries by not crossing county boundaries and keeping cities together as much as possible was the next rule. The bias, unless other geographic factors dictated it, was for smaller constituencies. One of these exceptions was in Hampshire where the 1983 boundaries meant that three 5-seaters was the best option. The thinking behind smaller district magnitudes was that it would arise as part of a compromise, with the larger party insisting on smaller district magnitude, this giving them a seat bonus at the expense of proportionality.
On the first drawing, having constituencies of a similar magnitude was the aim. For 1997 and 2010, continuity of existing constituencies was the aim unless changes warranted it. For example, Liverpool had two 3-seaters 1983-1997 and a 5-seater thereafter. Bedfordshire went the other way, a single constituency 1983-1997 and then split.
In terms of boundaries, I simply chained Westminster constituencies together. While it's possible that that could be done as a rush job for the first elections, subsequent elections would have had different boundaries to mine.
A number of other caveats apply. People would vote differently under a different voting system. Incumbency would play a role which it didn't in these set of results. Governments would be different and voting would be affected that way. The party system itself might end up differently as STV would make smaller parties more viable. A continuing SDP? 2 Conservative parties in the 1990s? So these are more a set of self-contained results.
Speaking of the SDP, I assumed they'd have entered a transfer pact with the liberals in the 80s. The number of candidates chosen by either party in the Lib-SDP alliance skewed some local results in 1983 and 1987 in 3 and 5-seaters. For example in the 3-seater of Waltham Forest the lib-sdp vote was around 25% in 1983 and 1987, Enough for a seat. In 1983, with candidates in 2 constituencies there, the SDP won it. In 1987, they only had one candidate, resulting in a loss to the liberals which almost certainly wouldn't have happened in real life.
With all that said, these were the results:
Other: SNP 5, PC 2, UUP 6, DUP 4, SDLP 4, UPUP 1, Alliance NI 1, SF 1
Enough there for a Lab-Lib-SDP government, but would the SDP back Labour? Con-SDP is another possibility.
Con 290 (+3)
Lab 198 (+26)
Lib 90 (-6)
SDP 47 (-24)
Other: SNP 6, PC 2, UUP 7, SDLP 4, DUP 3, Alliance NI 2, SF 1
As before, would depend on the attitude of SDP towards Labour, with Lab-Lib-SDP and Con-SDP coalitions having majorities of around 20.
Con 291 (+1)
Lab 233 (+35)
LD 92 (*-34)
SDP 1 (*)
Other: SNP 16, PC 2, UUP 6, SDLP 4, DUP 3, Alliance NI 2, SF 1
Rosie Barnes or Cartwright holds a single SDP seat, the Conservatives win a seat in Northern Ireland and the SNP have a big increase. Since a Lab-LD coalition would have the barest of majorities, Con-LD looks the likeliest option, unless Lab-LD can do a deal with the SNP.
Lab 336 (+103)
Con 195 (-96)
LD 88 (-4)
Other: SNP 17, PC 3, UUP 9, SDLP 5, DUP 2, SF 2, Speaker 1, Martin Bell 1
The only election when a party wins an overall majority.
Lab 320 (-16)
Con 200 (+5)
LD 98 (+10)
Other: SNP 15, PC 7, DUP 6, UUP 5, SDLP 4, SF 3, Health concern 1
No majority, but easily enough for a Labour minority with nationalist support.
Lab 253 (-67)
Con 210 (+10)
LD 151 (+53)
Other: SNP 9, PC 3, DUP 6, UUP 5, SF 4, SDLP 3, Respect 1, Independent (Blaenau Gwent) 1
Down to the Lib Dems....
Con 245 (+35)
Lab 202 (-51)
LD 172 (+21)
Other: SNP 10, PC 2, DUP 5, UUP 4, SF 4, SDLP 3, Ind Unionist 2, Independent (Bob Spink) 1
Spink's win was tight, with UKIP backing I had him edging out Labour and the third Conservative, in the 4-seater South-East Essex.
Again, down to the Lib Dems, who did even better than in 1983 in this election.
Con 281 (+36)
Lab 231 (+29)
UKIP 52 (+51 since Spink rejoins in 2014.)
SNP 36 (+26)
LD 25 (-147)
Other: PC 2, Green 2, DUP 7, SF 4, UUP 3, SDLP 2, Ind Unionist 1, Alliance NI 1
As in real life, the LibDem get massacred. Con-UKIP is the only coalition possible there, but if Farage played hard ball, another election could result.