A preferential ballotThis excellent question was asked this morning by ERS Scotland, and it’s of course impossible to tell, because we all (well, a third of us) filled in a ballot paper with a single X on it, and because people might well have given a first preference to a party which didn’t get their X last week – i.e. anyone who voted tactically. But let’s assume that those clumsy Xs would have been sophisticated 1s and then get even further into handwaving and guesswork from there. Is Tory Iain McGill right to assume (in terms I disassociate myself from) that we’d simply have swapped a ‘kipper for a Nat?

Another assumption is that we’d use preferences for parties rather than the better version of preferences by candidate (for example, if I’d been able to preference amongst the SNP I’d have put Hudghton and Gethins ahead of Smith and Ahmed-Sheikh). For the purists, this assumption is equivalent to the assumption that every SNP voter gives a 1 to Hudghton, a 2 to Smith, a 3 to Ahmed-Sheikh and so on: voting the straight ticket in the order chosen by the parties. My final assumption (and this one is definitely wrong) is that all voters will rank all candidates. Let’s not get too far down the STV rabbithole by debating specific counting models, either, incidentally.

There were 1,343,483 valid votes cast. That would set the usual quota for election at 191,927. The SNP’s Ian Hudghton would comfortably claim election first, with 389,503 first preferences. Labour’s David Martin and the Tories’ Ian Duncan would also make quota directly. No-one else would make quota straight away, so all 197,576 unused SNP votes would transfer to Alyn Smith, while 156,292 Labour votes would go to Catherine Stihler, and 39,403 Tory votes to Belinda Don.

At this point Alyn Smith is also over quota by just 5,649 votes, so he is elected fourth, and those votes all transfer to Tasmina Ahmed-Sheikh.

So, four MEPs have been elected at this stage, two SNP, one Tory, and one Labour. Who’s next? Labour are closest to the next seat, 35,635 votes short, then UKIP, who are 51,393 votes short, then the Greens, who are 83,622 short. That means it’s time to start eliminating candidates with the lowest votes and time to get even deeper into guesswork about where their votes go.

The candidate with the lowest vote at this stage is actually the SNP’s third-placed Ahmed-Sheikh – her 5,649 preferences are lower than the 6,418 scored by No2EU, the union-led left eurosceptics. So she would be eliminated next: let’s assume that a third of SNP voters are voting the straight independence ticket, with the Greens next, a third voting soft social democrat with Labour next, and the final third splits equally between the Lib Dems, the Tories, and UKIP. There aren’t many of them so it doesn’t matter much.

Again, no-one’s much nearer to quota so No2EU get eliminated. Let’s assume half of them are primarily lefties and they go Green next, despite us not being anti-EU, and half of them primarily hate the EU so they go to UKIP. That’s 3,209 for each party, and again no-one’s on quota. The next two lowest are Britain First and the BNP. For the sake of simplicity, let’s assume that all the fash votes slide readily over to UKIP. That gives the ‘kippers another 23,851 votes, just 23,702 away from that fifth seat. Labour at this point are about 10,000 behind UKIP.

Now the real fun begins – no-one’s on quota, and the Tories’ surplus 40,030 votes (including transfers) are redistributed. Given the SNP are already out, that means their votes can only go to Labour, Lib Dem, Green, or UKIP. It’s hard to tell, but let’s assume half are rightwing and eurosceptic, people who’d put UKIP next, a third are coalitionistas prepared to put the Lib Dems next, and a sixth are non-racist unionists who’ll put Labour next.

This brings UKIP a tantalising 3,876 short of quota, with Labour 23,393 behind them, and it brings the Lib Dems to 109,951, just 3,446 behind the Greens. It’s not enough, though, and the Lib Dems are now the lowest-placed remaining party.

Their 109,951 votes (including transfers) are redistributed. Given the SNP and Tories are already out, that means their votes can only go to Labour, Green, or UKIP: bear in mind that the two remaining seats will be awarded only from that list too. I’ll assume (I know) that none of the Lib Dems have UKIP next, that a quarter of them are passionate unionists who’d put Labour next, and the remaining three quarters are lower-case g green-minded, transferring accordingly.

This takes Labour to 192335 and the Greens to 195860 (I have rounded in places so there’ll be an occasional vote astray). Both parties make quota, Labour’s Catherine Stihler narrowly for her party’s second seat, the Greens’ Maggie Chapman marginally more comfortably. UKIP remain those 3,876 votes short and thus return no Scottish MEP. The balance of Lib Dem transfers at the end isn’t particularly crucial – if more go to Labour first and then the Greens, then Labour would make quota more comfortably, with a surplus that could be expected to tend to preference Greens ahead of UKIP, although if not quite 10% of those Lib Dems had ranked UKIP above the Greens we’d be back to the same distribution as the actual result.

Of course, voters don’t vote mechanically like this with their preferences. Just look at any of the local election wards to see that. And you could make plenty of other assumptions that would see the final two seats go to Labour and UKIP, or just plausibly Labour and the Lib Dems (but no way could the SNP have won a third, contrary to Iain’s assumption). But wouldn’t it be nice to find out who we really want and really don’t want? Fair votes under STV aren’t just about proportionality, they’re about subtle and accurate expressions of our votes. I would like us to use nothing else.

(assumptions spreadsheet available on request)