AV, or IRV (Instant Runoff Voting), is arguably the worst form of preferential voting in existence. Take this example of IRV totals for four candidates ordered by preference:
30% a; b; c; d
35% c; b; d; a
20% d; b; a; c
15% b; d; c; a
Given the voting breakdown, most people would agree that candidate b should win, because no candidate got a majority of first preference votes and b got a sweeping majority of second preference votes. However, under IRV, b is eliminated immediately, because b received the fewest first preference votes.
tr;dr: IRV is terrible, and has a history of being adopted in states and cities in the US only to be abolished not long after, because of the bad electoral outcomes it produces. If you're going to pick a preferential voting system, it ought to be one that at the very least preserves the Condorcet winner; IRV is not that system.
1. The system that doesn't take alternative preferences into account (FPTP);
2. The system that does (AV),
which would you pick? AV isn't the perfect preferential voting system, but it's our only hope of a transition to any form of preferential voting right now, which opens the door for refinement later instead of closing the debate for another 20 years.
The choice today is a FPTP choice: the alternatives are A and B, not some mythical third choice that you just invented. Voters in the UK should vote for the best choice of those available, which is clearly IRV.
Even your statement is simply silly. You're assuming that a voting system ought not eliminate a candidate when he was the least favorite of the voters, and you proceed from there to bash IRV. Big assumption much?
In the example given, "a" seems the least favorite, unless you ignore all preferences except the "first choice" preference. However, voting systems that take into account second and third choices (or more) cannot be perfect, since groups do not really have preferences in the same way that individuals do: a group can legitimately prefer every candidate to every other candidate, as Condorcet voting shows.
Unfortunately we haven't been offered a suggestion box, just a choice between AV and FPTP - personally, AV seems like an improvement on FPTP, even if it's not the best possible system overall.
Why not switch to proportional representation instead, using D'Hondt or Saint-Laguë? Or is there such strong support for the current single-winner system in the UK that noone wants to change the base idea?
Neither of the main political parties (Labour, Conservatives) want to support STV because it will greatly diminish their power in Parliament, therefore there will be no possibility of it being on the agenda either in this Parliament or at any other time in the next twenty years.
As it stands, the "No" campaign merely for AV has, for a number of reasons, managed to convince more than enough people that this is a hatefully undemocratic change to our electoral system - regardless of how tiny an effect it will have on anything - that I can't see STV having a chance for many decades yet.
"It's undemocratic", the rallying cry of anti-progressives in these sorts of situations. The Conservative Party of Canada was calling coalitions undemocratic. (But suspending parliament when coalitions against them are being discussed... apparently not undemocratic.) Conservative supporters were calling voting strategically (voting for whoever had the likeliest chance of lowering the odds of a Conservative majority) "undemocratic".
The only reason we got the referendum at all is because the Conservatives didn't win an outright majority of seats, and so weren't able to form a government after the last election. They had to form a coalition with the Liberal Democrats, and the LD's terms for lending their support to a party that they have very little in common with was they they insisted on the AV referendum. The Conservatives grudgingly accepted this, but would never have agreed to a referendum on full PR.
Basically, the entire thing is, sadly, driven by self-interest. The Tories don't want AV because they will probably lose some seats under it; but they DEFINITELY don't want PR, because they (and Labour) definitely would lose seats to the Lib Dems.
... and, lest anyone think that losing seats to the LibDems just means it's the LibDems' self-interest driving their wish to change, look at the comparison of the vote and seat percentages. At the last election, for every LibDem MP elected there were 5 Conservative MPs elected, but for every 2 LibDem voters there were only 3 Conservative voters. Is that really the parliament we asked for?
We need AV not because it's the best forever but because it's our only chance of reform for a generation. I'm voting yes.
I don't disagree. But in the spirit of "the perfect is the enemy of the good" if AV/IRV is an improvement over FPTP (which it seems to be) - then is there any downside (in terms of fairness of outcode) to a 'yes' vote in this referendum?
This is a personal view but I think there is. I think that AV could make the process of voting different/confusing enough for many voters to be turned off of voting or, perhaps, screwing up their ballots. If I agree with only one element of the "No" campaign, it's that enough of the electorate is stupid/ignorant enough to find the change difficult changing the outcome of future elections.
This probably sounds terrible, but if you are too "stupid/ignorant" to write your preferences in order 1-5, are you really competent enough to be voting?
I really don't expect there are many people (or any, I mean, really?) who are that stupid - that's just doesn't seem a valid reason to be against it.
If you just write "1" instead of an "X", it's exactly as easy as the current system.
Could you do it? Can you really think of anyone who couldn't do it? The entire tory party elected Cameron that way, it's not complicated at all. It does seem to be an effective scare tactic, though ;)
To quote Dara O Briain from twitter, "Listen, vote NO, if you're happy to trade lack of representation for a clear result. That's a perfectly fine choice to make. I respect that. But if you vote NO, because it's sounds complicated or you don't like Clegg, or the Irish are in debt, you're an idiot.
In Australia every polling booth has party volunteers handing out How To Vote flyers showing how they recommend you distribute your preferences. So if you like the Greens but don't really know where those two obscure independents fit in to the picture, you'd just follow their guidance on whether to put them at the top or the bottom of the list.
I suspect most Australians probably just follow their preferred candidates preferences.
Allocation of preferences is the mechanism that gives the smaller parties a say in the political process. The Greens (say) might know they have no chance of getting more than a couple of percent of the votes in a giving electorate, but they can do deals with the major parties based on that. e.g. "If renewable energy becomes a campaign promise, we'll direct our supporters to put you as their second preference."
I suspect most Australians probably just follow their preferred candidates preferences
Do they really? I never have, and I've never understood the mindset that would. Maybe I will vote for you, Mr Candidate, but I'm not gonna take your advice on the numbers I stick in everybody else's boxes... especially since those numbers probably indicate whatever deals you've made rather than some actual order of preferability based on competence or ideology. For this reason I've always thought that the whole preference-deal thing was a complete waste of time, but hey, maybe there's a lot of people out there genuinely following the how-to-vote card.
It hardly matters anyway. There's very few electorates with more than two viable candidates, so all that really matters is whether you put "Liberal" before "Labor" or vice versa.
The link says most people just go along party lines.
Even though in many electorates your vote will end up going to Liberal or Labor, the path it takes to get there is super important. That's kinda the point. If you think the environment is a really important issue, go vote Green because typically they'll have worked out a deal with one of majors to get some of their issues addressed.
You like smoking pot and there's a nutty local candidate who'll never be voted in in a million years and who none of the major parties will have bothered to do deals with? He's still worth a vote so that once the election is over the other candidates can see there's some people who consider it an important issue.
The entire beauty of preferential voting is it gives some voice to the other candidates in shaping policies, regardless of who gets in.
Fair enough. The other advantage is that it really does let us know when fringe candidates are fringe. If the nutty independent gets 3% of the vote then he can't say "I woulda got elected if it weren't for this damn two-party system"... we just look at him and say that, yep, he really is on the fringe.
Personally my strategy is to always vote number 1 for someone I'm pretty sure is going to get less than 4% of the vote. Why? Electoral funding rules. The government gives the parties money for their election campaigns based on how many votes they get. Except for folks who get less than 4% of the vote -- they get nothing. So if you vote 1 for a major party you're voting for taxpayer money to be given to political parties (yeech) but if you vote for some minor <4% nutjob then no money gets given on your behalf (hooray!)
Or so I've heard. I've never actually verified that this is true.
I'm also one of the few people who actually fills out every box below the line on the upper-house ballot sheets. A few years ago I think there were two hundred and something candidates, so I just wound up numbering consecutively in pretty spirals and so forth. Then I folded my ballot paper into a hat and wore it across the room to the box. They glared at me.
Perhaps this comes from living in the north, but the people I associated with are less sophisticated than you give them credit for. My wife has next to zero understanding about any of this and is not voting because she has no understanding of the issues involved. Ditto for my parents. And my in-laws.
Assuming that everyone's an intellectual and will research and consider the options and then make an informed choice is both idealistic and unrealistic. Many people have better things to do with their lives than thinking about politics and they're out there doing them. Making the voting process even more complicated is hardly going to compel them to get involved.
Well, that's a bit depressing. :/ But half my family live in the north, and I don't think you give them enough credit :)
Also, you seem to be talking about the decision making process of assessing whether to vote for AV or FPTP, not the not the complexity of voting using an AV system.
I'm not convinced that the AV voting method is too difficult to understand, though. Ranking things in order of preference is not hard for anyone to do. The finer points of the difference in electoral outcomes given different voting statistics might be more sophisticated, but that could be applied to any of the voting systems.
Voting itself though is no more complicated saying "I'll have a pint of Stella. If they don't have Stella, I'll have Fosters."
The No campaign has made the AV system out to be more complicated than it is as a scare tactic to put people off voting for it. That seems to have worked, which is deeply unfortunate, but indicative of political campaigns across the world. Fear and doubt are excellent motivators, and seem to be far more effective than a clear, reasoned and 'intellectual' argument.
Call me an unrealistic idealist, but I would much rather people research and consider the options before making an informed choice. Making a case for FPTP because it encourages more people who are uninformed and unconsidered to vote isn't going to win me over, though :)
However, as I say, the AV voting mechanism really isn't complicated at all - my wife is an early years teacher. Voting under an AV system is so easy that - literally - a five-year-old can understand it. I don't see anyone being put off by it.
Moreover, I think more people will be enthused by i) their ability to vote for the candidate that they believe in, ii) the greater accountability of their elected MPs to engage with their constituents and more accurately reflect their views and iii) their ability to express their opinions in a more meaningful way through the ballot box, that turnouts would be greater, not smaller.
Low turnout and voter apathy has a multitude of causes, but the ones that I hear most often are "lack of accountability" and "my vote doesn't count anyway."
Surely people who are so stupid/ignorant that this system would stop them voting, are too stupid/ignorant to be able to vote effectively in any election.
You may argue that it sounds a bit elitist, but personally, I want an electorate capable of making informed choices in the voting booth.
I agree with your last point entirely, but unfortunately it's a politically unacceptable one. Passing some sort of civics exam before having the right to vote would be fine by me.
It's the same justification given for not letting minors vote.
If it's acceptable to de-juris disenfranchise the under-18s on the spurious grounds that they are incapable of voting effectively, surely we can de-facto disenfranchise any adults incapable of operating an AV ballot paper. (which only requires the ability to count up to 1)
An important factor in this electoral system is that you don't have rank all the candidates, just those that you wish to receive your vote. This makes your sample election a trifle unrealistic.
Since everyone voted for everyone in your sample election, everyone would be reasonably happy with the result, regardless of who wins.
Condorcet just points out that group preferences are not necessarily consistent. This is a property of group preferences, not merely of some voting systems; the voting systems just allow the inconsistency to be noticed.
In the example I provided there is (intuitively, not with relation to any criterion) a consistent preference among the voters for one candidate over another, but the less-preferred candidate is the Condorcet winner.
So, the simplified version of your example would be each person having two votes that cannot be cast for same person. Top vote getter wins (regardless if it is over 50% or not).
That seems simple enough and easy to explain. Not sure if it would work.
30% a; b; c; d
35% c; b; d; a
20% d; b; a; c
15% b; d; c; a
Given the voting breakdown, most people would agree that candidate b should win, because no candidate got a majority of first preference votes and b got a sweeping majority of second preference votes. However, under IRV, b is eliminated immediately, because b received the fewest first preference votes.
tr;dr: IRV is terrible, and has a history of being adopted in states and cities in the US only to be abolished not long after, because of the bad electoral outcomes it produces. If you're going to pick a preferential voting system, it ought to be one that at the very least preserves the Condorcet winner; IRV is not that system.