By Graeme Archer

AV will end the scandal of safe seats for life; it will make candidates for elections work harder; it is a change supported by Benjamin Zephaniah (in urban areas) and Tony Robinson (in rural areas); it will end the expenses scandal; it will help minor parties to flourish; it is, by definition, fair, not least because it will lead to MPs who are supported by at least 50% of their electorate.

Other than the scantily-disguised racism of the differential ethnicity deployed in the celebrity pro-AV leaflets, not one of those statements is true. Yet they are the claims made for this change in the voting system, a change we are sleep-walking towards commiting; a change which, if the comments I see on ConHome are anything to go by, seems to have caught a certain demographic in a cognitively dissonant embrace. It's as though people are saying I know AV won't give me any of the things I want, but if I want it to enough, then it might. Well, it won't. It's time to stop pussy-footing around this issue and confront the bare bones of the electoral algorithm we are being asked to embrace.

The likely anti-Conservative outcome of an AV election wouldn't matter, of course, if the AV algorithm were capable of delivering a fair outcome. I would just have to put up with being governed by some form of socialism for the rest of my life. But the fairness claim for AV is its most egregiously false one, and it deserves to be taken apart. The lack of fairness is, fortuitously, independent of the empirical outcome of the application of AV. I'll come to that in a moment. Let's get some of the other myths about AV out of the way first.

AV will end safe seats. No, it won't. Around one third of our seats can be called 'safe' and would continue to be so under any reasonable electoral algorithm. People who say 'This area is so dominated by one party, my vote doesn't count for anything' should consider their solitary option: move somewhere else. Does that sound brutal? Guys, I live in Hackney. The Labour vote isn't so much weighed, as governed by one of those planetary constants, like gravity, only slightly harder to overcome. If altering the algorithm for counting could make things easier for the Hackney Tory Collective, I'd be happy. But it won't, and nor should it. And if you think that AV should make your vote 'count' for more (than it ought), ask yourself where AV's precious desire for fairness has left you? Why does being unpopular give you the right to have a vote count for more than the people who support the popular candidate?

AV will make candidates work harder… What is meant is that a candidate will have to appeal to people who don't want to vote for him. I've been thinking about this a lot, because I still long to fight a seat. The closest I got was the open primary at Bethnal Green & Bow before the election. (Now, open primaries – there's a way to tackle safe seats). Suppose the miracle had happened, and I'd been selected. I think that under AV the pressure would be on a candidate to say nothing controversial, for fear of offending that all-important group of people who might give you their second preferences. And this would lead to what, in an area like Bethnal Green? Do you think it would make it more likely, or less likely, that we'd get a candidate willing to speak clearly about what's going wrong in East London? The end result of this will be the increased homogenisation of candidates, which is ironic, given AV's claim to end the safeness of seats. The party label might change between elections, but the opinions mouthed by the candidates are likely to sound more and more alike. (Yes, this is a criticism which can be made now. The question is, do you think AV will make matters better, or worse?).

… by aiming for 50% of the vote. This is one of the most curious aspects of AV support. Of course, 50% support is by no means guaranteed by the algorithm, because you don't have to vote more than once, and many people won't; so when the BBC use this as a shorthand way to describe AV, they're lying (try not to be surprised). In any case, I'm struggling to think about the candidate in a marginal seat under FPTP who would be thinking 'Well I only need 42% of the votes to win, so I'll give up once it looks like I've reached that.' And why 50%? Why not require 65% of the support of your constituents? Or 50% + 1? Or 50% + x, where x is vanishingly small, but larger than zero? 'Fifty per cent' is a fetish, a made-up fetish, with no intrinsic worth, other than as a surrogate for the fact that we want the winning candidate to have more support than any of the others (we call this FPTP) – and it isn't guaranteed by AV in any case.

AV will end the expenses scandal. This is such drivel it's not worth dealing with; but it is apparently something that Nick Clegg believes. I guess he's been under a lot of stress, lately.

AV will help minor parties flourish. I'm similarly tempted to dismiss this as the obvious rubbish it is, except that I see a lot of comments by Centre-Right readers that it is indeed the case. I guess that what they mean is that they will be able to give their first choice to UKIP and their second to the Conservatives. Last week's nonsense argument about what AV would do for the BNP is another example that left me scratching my head: AV will help minor parties flourish, unless they're the sort of party the establishment disapproves of, is that it? Giving your first choice to a candidate who has practically no chance of winning isn't a way to help their party flourish: it is, rather, to make a virtue out of political impotence. There is only one minor party which will flourish under AV: the one which is currently led by the Deputy Prime Minister. He – or more likely Simon Hughes – will be in the cabinet for the rest of our lives. Not so much Simon Straight Choice Hughes as Simon No Choice At All Hughes.

AV is fair. All of the above is wrapped up into the fairness thing. It's not fair, we are told, that under FPTP the votes of some people don't count (we call these the 'votes of people who voted for the losing candidate'). AV apparently becomes fair not just because of the 50% thing (which is just gibberish, anyway), but because we can rank our preference of the candidates, if we so wish. If there are n candidates in the seat where you and I live, then we can both vote by supplying our ranked lists, with our favourite candidate marked 1st, and our least favourite candidate marked n'th. The AV algorithm takes our rankings, and all those of the other voters, and produces a 'fair' outcome.

Except it doesn't. There exists a theorem which shows that no system based on rankings can be guaranteed to produce an outcome which meets a reasonable definition of fairness. To speak in English rather than mathematically, the problem is that my last vote – for the candidate I want the least – counts as much as your first one. And that's not just a theory; it will happen with AV. The second choices of people will start to outnumber the votes of a larger group of people who picked another candidate as their first choice. In what benighted worldview can that possibly be described as 'fair'? No-one in the Yes2AV camp ever wants to address this issue; because I think, deep down, they recognise it as the fatal flaw. The closest excuse I've seen is that AV is somehow like a system of exhaustive run-offs. But it isn't an exhaustive run-off, or 'like' one. It's a huge and untenable assumption to make that all my ranks have equal weight, or that they wouldn't alter if the voting was carried out sequentially.

My last vote for my least-favourite candidate should not count as much as your first vote for your favourite one. If you agree with me, then Vote No to AV in May.

147 comments for: Graeme Archer: There Should Be No Alternative (Vote)

Leave a Reply

You must be logged in to post a comment.