by Anne Archist
This blog has covered the topic of the Alternative Vote referendum once already, thanks to Edd. Similar left-wing ‘no’ arguments have been put forward over at Third Estate by Reuben and Jacob, with Owen replying in favour of a ‘yes’ vote. It seems there are splits within blogs as well as between them on this topic.
Like Edd, I come from a position of being more on one side of the fence than the other (though unlike him, I tend towards AV); however, I also find it difficult to get particularly enthused about either option, and both campaigns have riled me up far more than I thought they would or could. I want to make a few key points about the referendum and the arguments made by each side in this post that I don’t think have been sufficiently considered, as well as to point out some aspects of the comparison that I haven’t seen clearly and impartially explained. Please forgive the structure of the post, which is likely to be somewhat akin to a FAQ.
1. Most predictions are shadow-boxing.
1a. Does AV hurt the far left?
1b. More proportional or more coalitions?
1c. How will they spin it?
2. AV requires a majority… of what?
3. That’s not funny.
1. Predictions have been flying around left, right and centre about the differences that AV would make to British politics. Unfortunately for the voter, almost every single one of these is utterly under-informed speculation; the problem is that by the very nature of AV, we would have to take into account the distribution of a whole series of preferences across the country, divided up by relatively arbitrary constituency boundaries, and the effect this had on the arithmetic of parliamentary seats, etc. Now, there may be some very good research out there that used a large and representative sample group, divided them up into realistic constituencies, and asked them about their preferences on an explicit understanding that the system used was AV. However, every prediction I’ve seen has been based on either no evidence at all, or else on data drawn from elections held under FPTP coupled with a bit of gut instinct about what we might reasonably expect people to do given some basic assumptions.
I’ve already covered the topic of multiple-peaked preference sets on this blog, which the more attentive of you will have noticed makes AV very difficult to predict at times – anyone who’s seen a lot of student union elections knows how difficult it is to pick a winner under relatively common conditions. It’s practically impossible, for instance, to work out how second preferences will be allocated from a novelty or single-issue candidate even if you assume that preferences are only single-peaked.
When you introduce the possibility that lots of people who are currently lib-dem voters could put green as their first preference, then lib-dem, then labour, for example, it becomes really difficult to tell whether AV will tend to swing parliament towards the left or right compared to FPTP. And even that makes the assumption that the parties are static across time, rather than changing their policies in reaction to the changing electoral system and distribution of votes… The point is that anyone making predictions about what AV would do is probably making it up as they go along, or just assuming things with no evidence. I’m now going to deal with three more kinds of predictions in a little more detail and address some of the claims made from other angles than just whether the evidence supports them.
1a. Given the difficulty in predicting anything, it’s a bit weird that people seem so certain AV will hurt the far left rather than help it. This argument relies on the premise that a socialist candidate is unlikely to get more than 50% of the votes; crucially however, a socialist candidate is also unlikely to get more votes than any other candidate! If AV ‘hurts’ the left, FPTP could be said to ‘hurt’ the left far more – at least under AV we should get a reflection of how many people actually want a socialist candidate to win – it strikes me as somewhat devious of some groups on the left to seek to conceal their level of electoral support (from even themselves).
I could write a whole post about how suspicious I am of some of the motives that far left leaderships could have in opposing AV, but hopefully it strikes others as odd that they would think their position better under FPTP… This argument also seems to represent a worrying trend towards parliamentarianism and perhaps reformism, in that it implies that the left should be actively trying to get far-left candidates into parliament rather than focusing on other strategies.
1b. Another set of predictions that’s bugged me are that AV will be “no more proportional” than FPTP and that it will lead to “more coalitions”. I’m not sure that the first prediction really makes any sense in itself – surely it depends on what you think proportionality entails? There is a wider question at stake in the AV debate, which isn’t decisive by any means, and isn’t easily assimilable to the questions of proportionality and representation and so forth. The wider question is whether you think parliamentary democracy should be about picking candidates most people can live with, or picking candidates who have the largest fanatical minority behind them. I don’t think there’s any easy answer to this question, and I don’t think it’s the same as asking whether you think “extremists” should be enfranchised or not, though it’s closely linked.
The second prediction seems to make the assumption that coalitions are inherently bad things, which seems like a remarkably odd thing to say. Many European countries are used to coalitions, and there are famous historical examples of left-wing coalitions (the popular front concept was applied in France and Spain in the 30s, for instance). Coalitions are not inherently weak or inherently right-wing, or inherently anything else you might think. A coalition government is just an executive whose offices are occupied by members of more than one party – it could be a Green-Socialist-Labour grouping, for instance. Incidentally it’s also a bit of a leap to assume that a coalition would form if there was no clear majority – there’s always the possibility of a minority government, with or without a voting pact with other parties. Finally, this criticism seems to put a higher priority on ‘strong government’ than real democratic representation, which is worryingly authoritarian, to say the least.
The two predictions together are really frustrating, though; they utterly contradict each other as far as I can see, yet people claim both at the same time. I’m open to being corrected by anyone more versed in political science than myself, but it seems to me that AV can be one or the other, but not both. If AV is no more proportional than FPTP then I fail to see how exactly it will lead to more coalition governments than FPTP. If it leads to more coalitions, surely it can do this only by being more proportional? How exactly do people argue both points in the same sentence without realising this inconsistency?
1c. And now, the predictions about what political impact different results to the referendum would have. These come in three varieties: “We should all vote yes, because we’re more likely to get PR that way”; “We should all vote no, because we’re more likely to get PR that way”; “We should all spoil our ballots/not vote, because we’re more likely to get PR that way”. Clearly these are all incompatible, like in 1b. That’s not the point, though, since I’ve not heard anyone use more than one of these claims at the same time. The problem here essentially comes back to the fact that we have no real evidence in any direction and no way of convincing anyone whose instinct is different to ours on this question.
Interestingly, AV and STV are in effect the same system. AV is basically just STV divided into single-member constituencies instead of on a national scale – we wouldn’t need to explain a whole new way of casting your vote if we decided to transition from AV to STV – you still just mark your preferences, the same as before. The similarity of the two systems is quite evident when you consider that many universities use AV to elect executive officers and STV to elect NUS delegates without voters even knowing that different systems are being used. From the perspective of the student voting, both voting systems are exactly the same; the difference is only really apparent to those counting the votes.
So, for the record, I think a yes vote is likely to get PR at an earlier date because it will create an informal ‘precedent’ for reform of the electoral system, because nobody really wants AV so there will be a greater impetus to change it whoever ends up in power next, and for the reason explained above – the similarity of the systems creating an ease of transition. I recognise, having said all this, that few people who think a “no” vote is the best route to PR are likely to change their mind.
2. Lots of AV supporters have been making the argument that in a democracy it’s surely best that we have candidates supported by more than half the votes, and that AV does this. Now, ignoring the rather dubious assumption that putting someone as your 4th preference means “supporting” them, this is frankly highly misleading at best, and downright false if you feel a bit more cynical (it depends how you define ‘votes’). AV, of course, does require candidates to meet a minimum level of votes, but that minimum is technically >50% of all valid votes. Votes vs valid votes… doesn’t make much of a difference, right? Wrong.
The reason this makes a big difference is that as soon as a ballot is spoilt it no longer counts as casting a valid vote; as soon as ballot returns no preference at a certain stage in the contest, it is in effect, spoilt. Combine these two facts and you’ll notice that so long as some voters don’t get any of their preferences, the winning candidate will not need >50% of the valid votes cast at the start of the election. I’ll give an example now to clear this up a bit.
Suppose that the contributors to this blog are running in an election for position of blog editor, and 100 people vote. Edd gets 25 of the first-preference votes, while I get 15; Pat and 1967 get 30 each. At the first round, I drop out of the contest and it transpires that none of my supporters have marked a second-preference (I feel ashamed to let down such faithful supporters!), meaning that their ballots are counted as ‘spoilt’ from the second round on. This means that whoever wins needs no more than 43 votes – >50% of valid votes at this stage, but only 43% of votes cast in the first round. This is much the same as counting only the votes of the two highest polling candidates under FPTP and claiming that this ensures all candidates have majority support!
Some supporters of AV have cottoned on to the fact that you don’t have to cast multiple preferences under AV, and used this as an argument to reassure those wary of the supposed complexity. “If you want to”, they say, “you can vote exactly as if it were FPTP – just put a 1 instead of an X”. This is entirely true but utterly undermines the idea that AV ensures majority support for a candidate, which is why I’ve included it in this discussion – if we all just vote for our first preference and leave it at that, as we are entitled to do, how can the voting system possibly ensure that the winner has ‘majority support’, whatever that means?
3. To leave you on a light note, I’d just like to point out one final thing that isn’t of any political significance at all. People keep making the joke that the referendum is biased since it will be conducted under FPTP. Yes, yes, very funny, but I hope you all realise this makes literally no difference whatsoever in an election with only two candidates? The result would be the same under almost any voting system.
(The exception that I can think of, if you’re wondering, is range voting, where you give a number to each candidate/option indicating a proportion of support from none to total. In theory, supporters of one side of the debate could tend to be far more adamant than the other side, thus meaning that the overall vote for one side outweighed the other even though more people put more of their support behind the losing candidate. Perhaps advocates of reform are far stauncher than those of conservatism on this issue, and so put 100% support behind the yes vote and 0% behind the no vote, whereas a slightly larger number of people split their vote 60%/40% in favour of the no vote, for instance; provided the ‘majority’ were not too large, the less popular option would win through strength of feeling).
Well, that’s all I can bring myself to write about the travesties that we call the referendum campaigns for now, though I might update this or post something new on it if something important pops into my head before the 5th May.
EDIT: I was hoping it wouldn’t sink to this level but I’ve heard people repeating the outright lie that AV gives some people more votes than others or (more vaguely) “an advantage” over others. Here’s a quick and messy little diagram for the purposes of showing how AV preserves one person, one vote: