As you might recall, there was a referendum four years ago that would have significantly affected the results of the upcoming general election.
Entrenched power very successfully persuaded the British public that electoral form was dangerous and scary and might cause the wrong person to win, and they correspondingly voted against it.
I was, and am, very angry about this.
What I didn’t notice at the time was that the entire referendum was a massive con, that every option was a bad one, and AV would quite possibly have been even worse than the status quo.
Viewed as an alternative to FPTP in the abstract, AV is undoubtedly the better system. It’s not a great system but say we were to want to e.g. elect a mayor then AV is unambiguously preferable to FPTP (indeed, the Mayor of London is elected with a slightly weird and limited form of AV – it’s AV where you can only choose a first and second preference).
The problem is that electing a parliamentary body like the house of commons is not like electing a mayor, because you’re performing the same class of vote many times. This is what gives rise to things like Gerrymandering, but it also means that choosing apparently better electoral systems can give you globally worse results.
AV was sold as reducing tactical voting by allowing you to vote for your actual first choice preference first and not “waste” your vote. The problem is that in reality your first choice vote ends up being a red herring if it’s not for one of the two majority parties, because your candidate will drop out of the race anyway. Sure, you nicely signal your support, but ultimately the only thing about your vote that gets counted is whether you prefer Labour or Conservative.
The problem is that although AV raises the number of votes a minority party will get, it also raises the number of votes they need to get by an even greater margin.
Under FPTP, a minority party can and often will get a seat in a divided area because parties frequently win with a relatively small fraction of the vote, and because voters can “punish” the majority party they favour by defecting to a smaller party. Under AV, this doesn’t happen – they still vote for a major party, and the result is that rather than benefiting smaller parties as was claimed, AV reinforces the two party system.
Does this make AV the “worse” system? No. But I think it’s fair to say that it makes implementing AV at the constituency level a really bad idea. Fundamentally this is a problem with constituencies more than it is a problem with AV, but that doesn’t make implementing AV at the constituency level a good idea just because it’s “not it’s fault” that it would make things strictly worse.
I still think it’s clear that some form of electoral reform is required, ideally some flavour of proportional representation, and if the referendum came up again I would probably still vote yes simply to be able to signal that electoral reform was needed, but I think it’s important to realise what a massive scam that entire referendum was, and reflect on the complete lack of honesty and/or competence that that entails in the people proposing it to us.
Thanks for posting this. I was struggling to understand your criticisms from a sequence of Twitter messages.
I think you’re making a couple of assumptions which seem unsupported to me. The first is that the rising of the bar to 50% support will depress the chance for a minority party to win more than the ability to directly express preferences will inflate it. On what basis do you assume this? I can’t come across any real data on this, but some analysis of Australian suggestions at http://blogs.abc.net.au/antonygreen/2011/04/is-fptp-or-av-better-for-minor-parties-and-independents.html suggests that the Australian Green party has done somewhat better under AV than it would have under FPTP (barring unknown tactical voting).
The Australian system also allows parties to divert their preferences en masse, which exacerbates the flow of votes to large parties when small ones get eliminated; I don’t believe this would have been a feature of the UK system.
The second is the assumption that FPTP can deliver seats to minority parties on minority votes. There’s plenty of seats where a major party wins a seat on a minority of votes, but very few of a minority party winning a seat; the Greens in Brighton, UKIP in Rochester and ICHC in 2005 are the only ones in the current parliament, I think; UKIP in Clacton won on majority, as did ICHC in 2001. So I’m not sure how much space AV would have to *be* worse.
Your other point is interesting; you’re effectively relying on the local voting system some element of “noise” which is representative of the global preferences. I’ll admit this is an aspect of things I hadn’t considered! The obvious question here is how much you actually expect this noise to reflect global preferences.