Richard Willis is a Conservative Councillor on Reading Borough Councillor, and over on his blog he has a thorough post looking at the pros and cons of both first past the post and the alternative vote. He is particularly unimpressed with the misleading nature of the debate so far, and helpfully shoots down quite a few of the canards propounded by the No campaign:
I have long thought that AV was a good system for use at Parliamentary elections and I have heard nothing from the “No” campaign to persuade me otherwise. Indeed I have been annoyed by the misleading statements made by some in the “No” campaign. Just this morning I received a “No” pack from Conservative Campaign HQ which includes the misleading statement “Under AV, other people’s votes can be counted multiple times while yours is only counted once”. This is simply untrue! A Conservative in Liverpool or most of Scotland may find that their preferred candidate is eliminated in the early rounds of counting and that their second or third preference vote is redistributed to their chosen candidate.
The post concludes:
Taking all factors into account on 5th May I will be voting “Yes” to change our electoral system to one which ensures that a winning MP has the suport of at least half of those voting and which allows me to rank candidates in order of my preference.
It’s good to see that the yes to fairer votes campaign is drawing support from all parts of the political spectrum, against what looks like an increasingly narrow no campaign, dominated by conservative Conservative MPs and conservative Labour MPs and peers.
Follow @libdemvoice.org on Bluesky
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
"The old savage England": D.H. Lawrence visits the Stiperstones
Church Green video
Thinking of standing for election or re-election next year? Helpful questions to consider
Endometriosis and acetylation
Henry
Daniel Walker
David Raw
27 Comments
A liberal case against the Alternative Vote
by David Allen Green
http://www.newstatesman.com/blogs/david-allen-green/2011/02/vote-system-party-liberals
Very good last point. The only argument of the NO campaign is conservatism. Whenver they try to make an argument it dissolves in the face of our counters.
David Allen Green makes two points. Firstly “AV is not in fact a good form of proportional representation”, which is untrue only because it is not a system of proportional representation at all.
The second is the tiresome one about “the votes cast by those who favour the most popular candidate are not of equal value to the votes cast for less popular candidates.” We are hearing this all the time, and it is untrue. David Allen Green says “The charge that AV means repeated bites at the cherry for some voters but not others is impossible to rebut”. AV simply has the same result as re-running the election under FPTP but with progressively fewer candidates not standing. Everyone’s vote counts the same each time, it cannot give a result that would not be given by FPTP when run with fewer candidates so it cannot give the result “someone who is everyone’s second choice but no-one’s first can win” as has been claimed in places.
It is said “AV means the election is decided by people who vote BNP”. What this actually means is when the BNP candidate is eliminated, his/her votes go to their next preference. Well, yes, but that’s where they would have gone anyway had the BNP not stood a candidate. So you might as well say under FPTP every seat where there is not a BNP candidate is decided by BNP voters.
AV repeatedly reruns the election under FPTP, the only reason everyone’s vote is not physically counted again is that it assumes people who voted for A will not switch their vote to B because C is not standing. So, when C is eliminated, it is only necessary to go through and count C’s next preferences. There is no need to go through the votes for A and the votes for B physically, because there they are already, sitting in piles counted for A and B. The “second bite of the cherry argument” applies only if there are people who would go to the polling station and see “Ah there is a Labour, a Conservative, a Liberal Democrat and a Green standing – I’ll vote Conservative”, but if they went to the polling station and found no Green standing they would say “Ah, then I’ll vote Labour instead of Conservative”.
It is no wonder this country is in a mess when not only are there are people so innumerate as not to be able to understand the simple algorithm that is AV, there are people who suppose themselves seriously politicians fit to run the country who actually publicly sign up to a campaign which is based on proudly asserting one’s innumeracy.
I would feel better about the “No To AV” people if they argued using the REAL argument for opposing AV instead of the bogus innumerate one. The REAL argument against AV is that not having it forces people to vote for the two biggest parties for fear to do otherwise would “split the vote”. That is a genuine argument, I don’t agree to it, but I accept it is genuine – “it is good to have a system where people pay a potentially big penalty for voting for anything but the two biggest parties, and are therefore forced to vote for a party that isn’t really their first choice, because that keeps the two party system and the good government Britain has enjoyed all these years is due to that”.
So, anyone who campaigns against AV but is not honest in using the above argument is either innumerate or knowingly lying on the basis most of the rest of the country is innumerate and so will not realise that it is a lie.
In our local campaigning for the coming elections I have found yes and no’s from voters of all parties. Obviously more of the Lib Dems supporters have said they are Yes voters than No but there have been a significant number from Labour and Conservative voters too.
This is heartening to see as this issue is far more important than party political boundaries.
If the country votes YES for change, the next step must be a search for a system to replace AV.
Our electoral system should be fair to all parties.
A desperately needed characteristic of any new voting system should be that every vote should make a difference, wherever you vote. Does anyone want to argue against that?
AMS and AV+ will be considered, as well as STV, (the marmite system) but have a look at Direct Party and Representative Voting.
http://www.dprvoting.org or google dpr voting
“If the country votes YES for change, the next step must be a search for a system to replace AV”
Given the historic trend for conservatism in referenda I would be careful about shouting that too loud!
“A desperately needed characteristic of any new voting system should be that every vote should make a difference, wherever you vote. Does anyone want to argue against that?”
Sure, i’m not interested in this torturous definition of “fairness”, what matters to me is a constituency link and a decisive manifesto mandate for policy reform from a party with the ability to implement it. The corollary to this is that the benchmark for success or failure can readily be assessed, and punishment readily assigned at the next election in the event of its failure.
Of course Willis is right to debunk the nonsense about some votes being counted several times.
But it’s a bit ironic that he concludes with a fallacy of his own about AV being a system “which ensures that a winning MP has the suport of at least half of those voting”!
I think both campaigns have been poor on facts. Both sides of the argument can be made without resorting to untruths.
Clegg’s speech the other day was patently wrong on a number of counts which were covered in the thread at the time. The No camp have also used the same tact on cost etc. People actually deserve the facts not the spin being offered from both sides.
Rebember how David Cameron was going to be hands off and not campaign for the No vote ?
He lied.
He’s going to be campaigning hard personally against it from now unitl May.
And just so everyone is clear, almost all Lib Dem voters will vote Yes, almost all Conservative voters will vote No, so it’s blindingly obvious where the swing votes are and why pretending the Conservatives are the Lib Dems friends on this is going to backfire spectacularly.
http://www.dailymail.co.uk/news/article-1358914/Alternative-vote-Yes-campaign-activists-breaking-rules-cold-calling.html
I find it unfortunate that neither campaign seems to be doing the right thing, and just telling the truth and keeping within established rules…
To Jedibeeftrix
The idea that ‘every vote should make a difference, wherever you vote’ is not just about fairness, but about democracy. If you ask people to vote, you should not have a system that ignores their votes.
If every vote made a difference, more people would vote.
You may or may not get a government from one party with a clear mandate, but that should be down to the electorate, not just a fix that is the direct consequence of the electoral system.
If you have looked at DPR Voting you will know that it is a system that works with the existing single member constituencies but unlike FPTP doesn’t depend on frequent boundary changes.
I agree that AV is a step towards a fairer system. Whilst AV is not proportional it could later become proportional with the addition of a “+”, extra regional seats being added to enable proportionality.We should use this campaign to get the AV and then work on the need for proportionality.
Those wanting proportionality (whether AV+ or STV) will have a lot of work to do to gain support from Labour and Conservative MPs before this becomes a realistic option.
Couldn’t you say exactly the same of FPTP, though?
@ stephen J – IMO the benefits don’t outweigh the costs – a vote is made within, and has a mandate within, the UK of GB and NI, if you don’t like the area you live within; move.
@ Chris – you could equally have FPTP “+” just as was suggested by lord wotsisname at the end of the jenkins report.
Having posted my opposition to AV on other lib-dem-voice blogs, I’m changing my blog name from plain Chris to avoid confusion with the Chris who posted above. I’m sure to be older than he (or she) is.
As others have pointed out, the Yes and No campaigns are both pedalling misleading nonsense.
Here are some proper arguments against AV – It can be LESS proportional than FPTP. It is prone to a certain amount of donkey voting where voters rank candidates randomly. Under certain circumstances a shrewd voter can get a better result by lying (eg a Tory in a Tory-held seat gives his first preference to Labour in the hope that the Labour candidate will beat the Lib Dem into second place, meaning that Lib Dem voters’ second preferences will be counted rather than Labour voters’).
With the exception of my example in brackets, these are direct quotes from the Electoral Reform Society’s website -before they removed that page to campaign for AV. Admittedly the removed page also gave arguments in favour of AV, but it’s a shame that even the ERS can no longer be trusted to give balance.
Old Codger Chris, your arguments would work if we had a choice of proportional representation in front of us. But we do not. We have only AV or the current system. A vote against AV is not going to be interpreted as vote in favour of proportional representation, because the whole “No to AV” campaign is dominated by opposition to AV on the grounds it is a change too far rather than a change no far enough.
AV corrects one anomaly in the current system, which is that if the people of a constituency prefer A to B as their MP, they can get B as the MP not because of anything A or B said or did but merely because C decided also to stand. It does not correct the single-member constituency problem, which is that only local majorities get representation, which I see as a more fundamental failing.
It is a sign of how little our influence is in this government, and hence of what a mistake it has been for those at the top of our party to claim we have made a “huge” difference or other such rot, that all we got in terms of electoral reform was a referendum on AV, a miserable little compromise as someone once said. If we really had a huge influence, we would have had a referendum on STV – the Tories might not have liked it, but for heaven’s sake, it’s a referendum, not actually imposing the system, so agreeing to the referendum and then calling for a “No” vote would have been a reasonable compromise for them, had we been in a position where we really had significant influence.
I take your point that AV can be fiddled tactically, but it’s less open to that than FPTP. Given the difficulty we have persuading people how to vote tactically under FPTP, and given that our society is so innumerate that top commentators proudly boast that they find the mechanics of AV beyond their comprehension, I think if we had AV it would not be with anyone’s effort to try and mount a serious tactical vote line. The “donkey vote” argument is more an Australian thing, where they have a stupid rule (does anyone know why?) which forces voters to use ALL their preferences.
If all you have on the table is a miserable little compromise, than practicality means you should go for it unless you are sure you will be in a situation soon to demand something better. I might feel it worth voting against AV if I was reasonably confident that after the next general election a government committed to having a referendum on STV would be in power. As I have no such confidence, I go for the compromise, it’s better than nothing.
Matthew Huntbach, I agree with much that you say, but I just don’t like a system which counts second, and even third, choices until a candidate gets 50%+1, disregarding the second preferences of voters who favoured candidates with more first preferences. Bad though it is, I think our current system of a simple plurality is less unfair.
There would be a better case for the Supplementary Vote as used in direct mayoral elections, with all candidates except the top two immediately eliminated and no third etc choices – I don’t think third choices are worth much.
As for the chances of getting any form of PR (I would favour FPTP plus top-up as in Scotland, Wales and London – but that’s a purely academic debate in current circumstances) I fear that’s unlikely anytime soon and the AV v FPTP referendum should be treated purely on its own terms.
Old Codger Chris:
As I have pointed out in another thread, this is not a practical strategy. I assume you are thinking of the situation where the result for first preferences is something like Con 42%; LD 30%; Lab 28% (in my model, these are the only candidates standing). Then, under AV, the Labour candidate is eliminated; nearly all of their preferences go to LD, and clearly the LD candidate wins easily. If 3%-worth of Tory voters switch to Labour, then the LD would be eliminated first. And let’s assume that the 2nd prefs of LD voters split 50:50 between Con and Lab. Then the results would be:
1st round: Con 39%; Lab 31%; LD30%
2nd round: Con 55%; Lab 45%
But there are major problems with trying to engineer such a change. The gap between the Labour and Lib Dem shares of the vote in this scenario is within the standard margin of error of opinion polls. This makes it virtually impossible to predict. It is very difficult to know a tactical Labour vote is worth casting, or whether it’ll have the desired effect. If too many Tories take this strategy, it would backfire by putting Labour in a position to win.The tactical switch above has already brought Labour within 8 percentage points of the Tories on first preferences. Bigger swings happen during election campaigns. How can
Let’s change it to the situation where the LD lead over Labour is more decisive in the first round: Con 42%; LD 32%; Lab 26%
Now, to bring Labour into the lead over Lib Dems, there needs to be a swing of 7% from Con to Lab, as:
Con 35%, LD 32%, Lab 33%
Out of 100 voters, 7 need to switch from Con to Labour for this to happen. Or, to put it another way, about 1/6 of Tory supporters need to switch. How do you explain this on the doorstep? A tactical voting strategy under FPTP is straightforward: a bar graph and “X can’t win here” is all you need. How can the situation we’re discussing be boiled down to something so simple? What sane campaigner would even want to try this? How do you prevent too many Tory voters from switching to Labour, enabling an easy Labour win? Already, with the above figures, the tactical switch has brought Labour close enough for the matter of who wins to be effectively in the hands of Lib Dem voters. A few more switchers, and Labour are first placed in the first round.
If a few misguided individuals try tactical voting under AV, then it is unlikely to have a material effect on the result. Basing a campaign strategy around it would be absolute madness.
“Conservative Councillor: Those opposing AV should stop their ‘misleading statements'”
Uh-oh, I can guess what’s gonna happen next…
“[AV] ensures that a winning MP has the suport of at least half of those voting”
Dang!!
I’m starting to think that AV is a bit like quantum mechanics. Anybody who thinks they understand it very probably doesn’t.
Old Codger Chris: And as you are no doubt aware, the ERS concluded that AV was “not suitable for Parliamentary elections”. It is obvious why the ERS are now suppressing this information – senior figures from the ERS are running the Yes campaign, telling us that AV is fantastic when a few months ago they were dismissing it as a lemon.
It’s a shame that anybody has to win this referendum – both campaigns deserve to fail because they’re both peddling lies.
Matthew: “AV corrects one anomaly in the current system, which is that if the people of a constituency prefer A to B as their MP, they can get B as the MP not because of anything A or B said or did but merely because C decided also to stand.”
Not so I’m afraid. On this point, David Allen Green is absolutely right.
Consider this scenario (highly simplified for easy of understanding) :-
Candidate A gets 30,000 first preferences. These 30,000 voters all choose candidate C as their second preference.
Candidate B gets 20,000 first preferences. None of these voters chooses a second preference.
Candidate C gets 15,000 first preferences. Of these, 2,000 choose A as their second preference, and 13,000 choose B.
Under AV, the winner is candidate B. So far, so predictable. AV apologists will not lose sleep over the fact that A got 10,000 more first preferences, because supporters of C obviously like B, and added together they provide a clear mandate for B – don’t they??
Well no, they don’t. If we were to eliminate A rather than C, and distribute A’s second preferences among B and C, we find that C gets 45,000 “votes” while B only has 20,000. So the electorate prefer C to B by a margin of more than two to one – yet as far as AV is concerned, B is the winner!! According to you, this should not be possible.
If you tot up ALL the “votes”, i.e. every preference put on every ballot paper, we find that while A and B are acceptable to 49% and 51% of voters respectively, C is acceptable to far more – a whopping 69% of voters. AV’s main selling point is blown out of the water – it fails dismally to deliver a victory for the candidate with the widest appeal. In fact, it looks very much to me (though this is subjective) that AV has managed to come up with the LEAST fair result out of the three possible outcomes.
Obviously the numbers in my example are somewhat unrealistic, but when I fiddled about with it in Excel, using much more plausible figures (i.e. second preferences more evenly split, applying second preferences to B, factoring in some votes with no second preferences etc) it was VERY easy to come up with scenarios in which B was acceptable to fewer voters than both A and C, yet still won.
AV is a dog’s dinner. It treats different votes unequally (some transfer, some don’t; some have second preferences that matter, some don’t; etc); it is confusing to its own proponents, never mind the voters; and it doesn’t even deliver the two main things its supporters claim for it (a guarantee of 50% support, and candidates with the widest support winning).
@Stuart: No-one claims that AV will always “deliver a victory for the candidate with the widest appeal.” The point of AV is that it delivers a victory to someone with the support of at least 50% of those who voted. So your analysis, while interesting, is irrelevant, as it is based on a strawman.
AV does not always find the Condorcet winner; AV supporters are aware of this. But a Condorcet election becomes unworkable with more than a few candidates. It might be just about workable with 3 candidates (you need to compare 3 ×2 × 1 = 6 pairs of candidates), but with 4 (4 × 3 ×2 × 1 = 24 pairs) and 5 (120 pairs) it’s not practical. Also not every election produces a Condorcet winner.
@Alex Macfie
“The point of AV is that it delivers a victory to someone with the support of at least 50% of those who voted. ”
Not actually true. It only means this if everyone whose preference candidates are eliminated place 2nd, 3rd of potentially more preference votes. It doesn;t follow that they are supported by 50% of voters even if they do get non preference votes, it means they are tolerated more than someone else, there’s a big difference between supporting someone and considering them the lesser of two or more evils…
The big lie is that every vote counts. In about 1/3 of seats at the last election anyone not voting for the incumbant wasted their votes and will continue to do so under AV….
Stuart Mitchell, As you point out, the – now removed – page of the ERS website stated that AV was not suitable for Parliamentary elections. In fairness, they think the same about FPTP. The ERS say they decided to support a Yes vote after asking their members – I wonder which party is the most heavily represented in that Group?
I could stomach the ERS saying that AV is slightly preferable (though some of us disagree) but their enthusiasm for it could set back the PR cause, especially if AV wins in May. To sceptics, it will look like reformers keep trying to move the goalposts.
I have 3 reasons for favouring the Supplementary Vote, as used in the direct Mayoral elections held in some English cities and towns –
(1) As with AV, it should reduce negative and tactical voting to some extent.
(2) A maximum of 2 choices implies a serious decision not a mathmatical game – the main purpose of a general election is to elect a government.
(3) If no candidate achieves 50% +1 of first preferences, the second preferences of voters supporting ANY candidate below the runner-up are re-distributed – insofar as these are for either of the leading candidates.
The Supplementary Vote is a miserable compromise (compared to PR) for several reasons. But I find it less objectionable than AV or FPTP and would have voted Yes to it if given the chance.
Stuart Mitchell, all I’m saying is that AV corrects one common anomaly. I’m not arguing that AV is better than any other system, far from it. This is, of course, the problem with the current referendum – AV is like your candidate C, few people will make it first choice but many will go for it as second choice.
The most common anomaly under FPTP is that a minor candidate can “split the vote” and thus result in the election of a candidate who would not have got elected had the minor candidate not stood. As I have already said, underneath, this anomaly IS the basis of the argument against AV. Because this anomaly exists it means that people who might be tempted to vote for minor candidates or for candidates from new political movements where it is not sure what sort of support they have, feel it is better to stick to the acknowledged lead candidates, those from the two major parties. The party labels serve as rallying flags for the best place to go to avoid “splitting the vote”. This combined with the distortion of the local-majorities-only representation generally ensures there is a Parliamentary majority for one party and stability as the same parties go on and on. The anti-AV people say this is a good thing. It’s a legitimate argument, but I wish they’d be honest about the mechanics.
Your point that AV excludes the “everyone’s second choice” candidate from getting elected is a reasonable one, but it’s part of the compromise of AV. Conservative commentators in both the big parties have attacked AV on the argument it does mean “everyone’s second choice” (and they mean the Liberal Democrats) will tend to get elected. In doing so, they only reveal their own innumeracy – I find it shocking that people like this are not ashamed to exhibit their failure to comprehend a little mathematics in public. AV does mean that in order to get elected a candidate must have a large number of voters who put him/her first choice, so it has this property the FPTP people think is valuable and wrongly attack AV for not having. AV is in fact a fairly minor variant on FPTP, in practice it rarely gives a different result in a constituency to the one which FPTP would give.
Under your scenario, candidate C needs to run a campaign appealing to A voters to put C first on the grounds that if C beats B, C gets elected on transfers, whereas A has strong minority support but is never going to get enough transfers to win. I have already tried this argument in Mayoral elections. Maybe if we did get AV in this country and our thicko commentariat were able to get their heads round it, they would be able to see it, but I didn’t see any commentator on the London Mayoral election who seemed able to grasp it and talk about it sensibly.
I think, however, there are enough intelligent people in the Liberal Democrats who can think through the algorithm of AV to see that if it were in place, this would be our new tactical vote appeal. It was beyond Jeff Rooker, leading light in the “No to AV” campaign, who had a ludicrous letter published in the Guardian about how Liberal Democrats might call for tactical voting under AV showing clearly he has no mathematical sense whatever (or that he is a liar, if in reality he knew what he wrote had no mathematical sense). Under FPTP, our tactical vote appeal is when we are in second place to squeeze the third placed candidate. Whereas under AV, our tactical appeal would be from third place trying to squeeze the first or second placed candidate’s votes for a tactical first preference. It’s a challenge, but I think I can already see the bar charts that could be used to argue the point in our literature.
Oh No Matthew Huntbach, not the Lib Dem bar-charts!
When asked to deliver Focus’s I regularly moaned about those bar-charts which, in my area (a) reminded everyone that we live in a Tory constituency / county / borough, and (b) probably confirmed to the anti-Labour portion of the electorate (most people in the streets I covered) that voting Tory was the best way to cage the Socialist Beast.
Most embarrassingly, the charts were sometimes – how shall I put this – a little misleading without reading the small print (estimated figures for the last General Election based on the new boundaries and voting in the last Council elections blah blah).
Labour’s Red Rose flyer was sometimes silly enough to respond with iffy bar-charts of its own. I’m sure the local Tories were laughing their heads off.
Alex Macfie: “The point of AV is that it delivers a victory to someone with the support of at least 50% of those who voted.”
AV must be pointless then, since it doesn’t even do that. Consider :-
50,000 people vote in an AV election.
10,000 votes are discarded before the final round of counting (since all the preferences on them are for eliminated candidates).
Therefore to win it is only necessary to get 20,001 votes in the final round – that is, 40% (+1) of all votes cast.
As a general rule, for every 1% of votes where the votes are discarded before the final round, the share required to win goes down by 0.5%. In practise there won’t be a single constituency in the land where it is necessary to poll 50% of the vote – in many the target will be much, much less.
So where has this 50% fabrication come from?
Some AV apologists try to justify the claim by saying that it only applies to the final round of voting, but that is a completely ludicrous and meaningless observation. What sort of electoral system DOESN’T guarantee a majority share for the winner in a two-candidate contest???
Interesting to note that Clegg (who has obviously cottoned on to the fact that the 50% claim is inconveniently untrue) has now changed his tack slightly – he’s talking now about candidates “having to aim for” 50%. This is the same kind of dissembling language he employed in his infamous “Tory VAT bombshell” speech before the election.
Politics sucks.