What difference with AV?

In a recent online discussion somebody made the very valid point that people advocating different voting systems see elections as serving very different purposes. Those advocating a proportional system see the purpose of elections as producing a result which reflects the views of the electorate, whilst advocates of FPTP see elections as there solely to pick a winner. I'm not sure what proponents of AV see as the purpose of elections.

When I posted a few weeks back and mentioned estimates from a daily paper on the differences AV would have made in May I was challenged over how accurate these were. So over the past few weeks in spare moments, I've been having a look at things myself to come up with my own results. But first the challenges of working things out.

The problems of estimating vote re-allocations

Party-by-party some of the issues faced:

Lib-Dems - are perhaps the easiest party to figure based on the fact that there are polls among their supporters about who would they would have liked to be in coalition with. This would indicate second preferences of roughly one-third Conservative to two-thirds Labour. I would guess that these figures are roughly right, but what's less easy is how these balance out across the country. Are Lib-Dems voters in Scotland, for example, more left-leaning than those in the South-West. The MPs certainly seem to be. The other thing with Lib-Dems is that their voters are possibly the most likely to indicate second and third preferences, by nature of their political position.

Conservatives. Conservative voters are possibly (my own opinion) the least likely to express second preferences. There's also a bit of guesswork as to how many would go Lib-Dem to keep Labour out or how many would prefer just to switch to Labour. Election results would indicate that many voters do switch from one to the other and are actually very dsmissive of the third party. In Wales and Scotland the picture is even more complicated - who would unionist Tories prefer in contests between the Nationalists and Labour.

Labour. Instinctively, I feel Labour voters are more likely to switch Lib-Dem to keep the Tories out than the other way round. Of course, there are still some who might switch Tory, but I'm guessing most who are likely to do, did so at this election (Outside Scotland at least.). Many Labour voters are maybe aready tactically voting Lib-dem - under AV these votes would come back to Labour on the first vote, before switching back to Lib-Dems on second preference. (Interestingly, polling just prior to the election indicated that this time round the Tories were going to be the main beneficiaries of tactical voting - presumably either from UKIP or right leaning Lib-Dems wanting to get Labour out.)

Nationalists. Theoretically, both SNP and Plaid Cymru are Centre-Left party, making their second preferences likely to tend that way. Certainly in urban Scotland the SNP seem to be competing for votes with Labour in the poorer regions. However, in some rural areas their supporters perhaps have more in common with the Tories.

Greens. The Lib-Dems are theoretically the most Green of the big three and there did seem to be a correlation between Greens doing well and Lib-Dems doing badly. Norwich South disproved that with Green advance being almost exclusively at the expense of Labour. Their policies are more left than right, but many Green voters might otherwise be Conservative, who are probably a toucher greener under Cameron than Labour.

UKIP It appears a no-brainer that most UKIP voters would go Tory on second preference. But even here there is a counter-argument that if that was the case, in situations where Euro-sceptic Tories faced tough opposition, why haven't they already? Some argue evidence from doorsteps in South-West that many UKIP supporters would be Lib-Dem if not for policy on Europe. Other than Europe, UKIP policies are a bit of a hotch-potch of progressive and reactionary, so difficult to call.

BNP Although a far-right party, BNP voters almost exclusively come from white working class areas that would traditionally support Labour.

English Democrats _ i really don't have a clue about where they get their votes from.

The impact of AV itself. AV, as already mentioned, would see an unwinding of tactical voting in the first preference, but those votes would end up back where they started on the second preference. Similarly, I would expect to see a first preference boost for the Greens and maybe other small parties, before again those votes ended back where they started.

All of which makes it rather hard to calculate, except there are a huge number of seats where the winning candidate got over 50% or close enough to 50% with a large majority that no change would be possible. Of the other seats, I reckon that, if I've got it wrong, its more likely that less would change due to voters not expressing a second preference than that more would change.

My estimate of the results under AV:
(I've not attempted to look at Northern Ireland)

Labour - 275
Conservatives - 273
Lib-Dems - 74
SNP - 5
Plaid Cymru - 2
Greens - 1
Independant - 1

To summarise, I think the system would help Labour to hold onto a large number of (mainly Middle-England) seats that they ended up losing to the Tories by 5% or thereabouts (such as Bedford, Carlisle,Colne Valley, Hendon, Ipswich). They might also have held ARfon from Plaid and re-taken Dundee East from SNP and held the Lib_dems off in Bradford East. The Conservatives on the other hand are the big losers, the only gains I see for them might be Wells from the Lib-Dems and Walsall South from Labour, with possibly Mid-Dorset as well. The might also have failed to take Wyre Forest from the independant. For the Lib-dems, they might have held on to few seats they lost to the Tories (Falmouth, Oxford West, Camborne) as well as coming through to take the likes of Bristol NW, Watford and St albans. But they actually benefit more against Labour in urban areas, taking seats in the likes of Swansea W, Rochdale, Sheffield C, Oxford E, Oldham, Newport, Hull and Edinburgh South.

Of course, even if AV does come in for the next election, its likely to be accompanied by boundary changes, larger constituencies and who knows what impact the coalition will have on all the parties fortunes.