As in 2015 and 2017, I’ve done a little thought exercise about what would have potentially happened in the recent General Election had the single transferable vote method been used to elect MPs.
The usual caveats apply, especially as the number of tactical voters this time is supposed to have been much higher than normal – mainly there is no guarantee that people would have voted for the parties they actually voted for, and there is no inclusion of the effect of popular candidates. Also, I’ve combined the votes for the Brexit Party and UKIP. Finally, I’ve used the convention that seats must return 3, 4 or 5 MPs – like they do in Ireland.
Headline figures then – the Conservatives would still have won a majority of MPs, but with a majority of 16 rather than the 80 they currently have. The number of Labour MPs would still have declined significantly (from 293 to 230), and the Liberal Democrats resume their position as the third party, with 32 MPs as opposed to the SNP’s 31. The Greens and the Brexit Party would have won one MP each. (On a slightly poignant note, assuming the debatable fact that I would have been the most popular Lib Dem in Wiltshire North, I would have been elected on this system.)
This time, because the boundaries were set on electorates which are over a decade old, I decided to adjust the number of MPs per multi-member seat according to the electorate figures for this general election.
This adjusted the number of MPs for some of my multi-members seats, with most Welsh seats losing 1 MP. Because of the extent of these changes, I redrew the Welsh boundaries I used for my multi-member seats slightly, so that Mid Wales and Gwynedd are now combined, and Clwyd is now one 5-member seat. Doing this properly would have meant splitting the new 6-member seats into two 3-member seats, which I didn’t do. The nett effect is that the Conservatives gain one seat each from SNP and Plaid Cymru.
The loss of the “red wall” is noticeable in this analysis too. So it’s still pretty bad news for Labour all round, and proof that STV can moderate but not eliminate the effects of targeted voting.
As usual, the spreadsheets are available for your own analysis should you so wish.