09 April 2011

Britain's Future If We Vote NO to AV - Candidates Elected by Just 9% Of The Vote

Papua New Guinea switched from our present system (first-past-the-post) to the Alternative Vote (AV) in 2007. No2AV campaigners like to cite Papua New Guinea to try to ridicule the idea of switching to AV, but what is more interesting is Papua New Guinea's experience of using our present Westminster system (first-past-the-post) between 1964 and 2007.

Here is a fairly typical result in a constituency under the previous first-past-the-post elections in 2002.

KARIMUI-NOMANE OPEN
===================================================================
Candidate Party Votes %
-------------------------------------------------------------------
Max Tabel Bro PNGNP 3,827 07.3
Jeffrey Kama 3,788 07.3
Douglas Duguno Kaupa PPP 3,220 06.2
Posi MENAZ 4,841 09.3
Wei Mogerema 3,448 06.6
Robert Nai Somodili 3,450 06.6
Simeon Gaima Wai * PMCP 3,304 06.3
Mark Kaupa Yoba PLP 3,647 07.0
31 others 22,719 43.5
-------------------------------------------------------------------
Total 52,244
Informal 100
Total 52,344

SOURCE: Adam Carr's Electoral Archive.

As you can see the 'winning' candidate [Posi MENAZ] with the 'most' votes gets just 9% (YES, JUST NINE PERCENT) of the vote. Just 4,841 votes out of a total of 52,244 votes cast in the constituency. No wonder they switched to AV. As Adam Carr puts it on this wikpedia page about Papua New Guinea elections under first-past-the-post in 2002.
"Papua New Guinea has no real party system and most Members of Parliament function as independents, although they give themselves various party labels. This tendency is reinforced by the electoral system, which combines first-past-the-post voting in single-member constituencies with large numbers of candidacies, making election to the House largely a matter of chance."

22 comments:

  1. Except that Papua New Guinea didn't switch to AV (although they used it pre independence). Instead they switched to "Limited Preferential Voting" (the local term) which is a version of the Supplementary Vote.

    ReplyDelete
  2. Yup. The irony of the claim that more extremists will gain power, is that it's actually FPTP that can win extremists seats: get enough candidates, have them split quite evenly and you can win with less than a 1/3rd majority (as Caroline Lucas did in Brighton for the Greens).

    ReplyDelete
  3. Tim, you are right in that it is called Limited Preferential Voting, because they only rank the first 3 candidates 1,2,3 rather than as few or many as you like with our proposed version, but it is a version of AV not the Supllementary Vote which (as it suggests) is just ranking one supplementary candidate (like in UK Mayor elections). In Australia (and it's neighbours PNG) our AV is called Optional Preferential Voting.

    ReplyDelete
  4. Joseph, yep, the more candidates under FPtP, the lower the percentage the winner may need. 5 candidates and 20% may be enough, 10 candidates and 10% may be enough, 30 candidates on the ballot paper (like in Papua New Guinea) and the winner needs just 7%.

    With AV, no matter how many candidates, the winner needs to get a majority to prefer them to the second placed candidate. That, the fact you can finally show your true opinion, and the fact it makes tactical voting pointless is why it is far superior to FPtP.

    ReplyDelete
  5. @Joseph
    "The irony of the claim that more extremists will gain power, is that it's actually FPTP that can win extremists seats"

    I don't think that's been the claim. I think the claim has been that extremists will win a greater proportion of first preferences because people will not see their vote lost. AV will simply allow more people to place their first preference for a less mainstream party. I think supporters of AV have actually claimed this is a benefit of AV. But you should expect the larger parties to get a smaller share of first preferences than they do under FPTP.

    Which of course means that your claim about Lucas is moot, it is conceivable that she would have got more first preferences under AV, and transfers may have pushed her over the top.

    ReplyDelete
  6. @Neil

    "With AV, no matter how many candidates, the winner needs to get a majority to prefer them to the second placed candidate."

    Not under the AV system that the UK will introduce. In the Australian system, where voters are REQUIRED to rank ALL candidates, what you say is absolutely true. But under the Uk system, if a large proportion of third placed candidate votes are NOT transferred (as might well happen to Labour or Tory votes), then the winner does not require a majority of CAST votes, only a majority of the votes that transfer to the final round.

    Further, I don't really see that AV gives the winner greater legitimacy.

    Granted winning on 9% isn't very good. But even if this had been an AV election, then it must be true that the "winner", even though s/he gets >50% of final round transfers retains the SAME low level of primary support, right?

    And of course it's a false argument to claim that Papuan elections are similar to UK elections. Trying to make generalized arguments from extreme cases is something of a fallacy, don't you think?

    I would point out that under AV second placed Australian parties have got parliamentary majorities in 23% of elections. Yes, that's right, nearly one in four times the second favourite party (on first preferences) actually gets a MAJORITY in parliament. This has happened 11% of the time in the UK.

    In the system we will use in the UK it's even more alarming, it's the same as the system in Queensland, between 1998 and 2006 (four consecutive elections), the second placed party in terms of first preferences FAILED to be the second placed party in the parliament, in one case the FOURTH placed party had the second highest number of seats in parliament. That is a direct result of not forcing people to fill out their preferences in full.

    I'm no fan of FPTP, but pretending AV is somehow superior is to ignore the facts. Bot are really very bad systems, and there's little to tell between them.

    ReplyDelete
  7. @Neil
    "the fact it makes tactical voting pointless"

    Well that is another of the things that pro-AV supporters claim, but it's simply not true. Tactical voting can make a huge difference under AV, just as it does under FPTP.

    On Radio % last week they held a mock AV election. During this election the final three candidates were Labour (41%), Green (29%) and Conservative (27%). The Conservative candidate was eliminated, and the final result was Labour 49%, Green 41%

    It's clear that about 6% of votes that went Tory were lost completely, and that 10% of votes were lost overall. It's also clear that the Winner did not get greater than 50% of the votes cast.

    BUT, had SOME Tory voters put the Green candidate on their ballot, and others put the Green candidate ABOVE the Labour candidate, then the Green candidate would have won the seat, preventing Labour from winning. THAT would have been tactical voting.

    ReplyDelete
  8. Alan, AV allows people to show which candidate/party is their true first preference. You seem to argue this is a bad thing. I cannot understand why you would prefer a system that forces people to first try and guess the result and then vote tactically on that basis.

    Of course tactical voting is theoretically possible under AV, but practically very difficult, it is much easier and almost always more effective to just list your true preferences.

    On the radio 5 election, a majority who expressed a preference chose Labour over the Greens. Yes, some decided they had no preference between these two and decided to abstain. Including these abstentions as votes makes the total for Labour slightly less than 50%, but it makes no sense to do that. The difference is that these voters had a choice between the two candidates and abstained, whereas under our present system they would be denied any choice on the matter - their vote would automatically be wasted.

    ReplyDelete
  9. Alan, finally, I am having fun by citing Papuan elections where MPs are elected by 9%. But seriously how far away is Britain from this scenario? 10 years? 20 years? In the 1950s nearly all MPs and most councillors were elected with over 50%. Now there are councillors regularly elected with just 20 something of the vote and we have MPs elected with just 29%. As voters choose to vote for more parties this can only get worse. AV allows for multiple parties without increasing wasted votes.

    ReplyDelete
  10. Hi Neil.
    "AV allows people to show which candidate/party is their true first preference. You seem to argue this is a bad thing."

    That's not what I said. Joseph stated that there had been a claim that more extremists could be elected under AV, and that this was not true. I don't know if this claim has been made, but what IS true is that people WILL be freer to vote for extremist parties without losing their votes. This may well lead to a change in voting behaviour, and mean that extremist parties get a greater proportion of first preferences than they do under the current system. This may not give them more seats, but it may well give them more "power" in the sense that they will be able to claim greater levels of popular support.

    It's also worth bearing in mind that if AV means a substantive change in voting behaviour, with a consequent fragmenting of the vote, then we might see elections that are LESS proportional than under FPTP.

    For example, for the last two elections the first party has gained ~35%,the second ~30% and the third ~22%. It is conceivable that we could see first preferences more like ~32%, ~27% and ~19%, with the first placed party STILL gaining a majority in parliament. The very fact that AV encourages people to vote for minor parties as their first preference might make it less proportional than FPTP.

    I am not arguing that this is necessarily a BAD thing, but just that this is likely.

    For myself, I support PR, and under PR we would expect minor, even extremist parties, to get some level of representation. I'm relaxed about that. The Yes campaign seem to want to have it both ways, arguing FOR the greater proportionality of AV, while also claiming that extremists won't get elected. It's disingenuous to make both arguments at the same time, most supporters of AV also support PR. If we're going to have the PR debate, then we have to admit that PR (and possibly AV), might well mean some extremist candidates win elections. So what? Isn't that democracy?

    ReplyDelete
  11. "Of course tactical voting is theoretically possible under AV, but practically very difficult"

    The possibility of tactical voting is a function of the system. Tactical voting is MORE than just theoretical under AV. Indeed AV is a system that more or less makes a virtue of tactical voting. The rank ordering of candidates is a highly tactical process.

    But here's the thing, it kind of depends how you define tactical voting, right?

    We agree that AV allows people to express their preferred candidate as their first preference. But is someone who usually, say, votes Labour in a two way marginal voting tactically? They might be voting against the Tory, or simply voting FOR the candidate they find LEAST objectionable. And if they started putting, say, the Socialist candidate as their first preference (under AV), does that imply that they have stopped voting tactically? One could argue that the first preference vote for the socialist is a tactical vote, as the voter knows their vote can be transferred to Labour further down the list.

    No, the whole tactical voting argument is a red herring.

    ReplyDelete
  12. "Including these abstentions as votes makes the total for Labour slightly less than 50%, but it makes no sense to do that."

    Of course it does, if you want to claim that AV wastes less votes, then you MUST include votes that are not carried over. They are still wasted votes, by definition, right.

    "The difference is that these voters had a choice between the two candidates and abstained,"

    That statement implies that the Tories had a priori knowledge that the Green candidate would make it through to the final round. But how could they have? Now if the Tories had KNOWN that there was a chance the Green candidate might make it through to the final round, and would be against the Labour candidate, and that THEIR transfers would be the difference, don't you think they would have included the Green candidate on their ballot?

    The Tories didn't transfer their votes because they ASSUMED that their candidate would be one of the final two.

    The point is this, the Australian version of AV is much superior because it DOES force all voters to rank ALL candidates. This is a very big weakness of the type of AV that is being proposed. It DOES waste more votes than it needs to.

    ReplyDelete
  13. "I am having fun by citing Papuan elections where MPs are elected by 9%. But seriously how far away is Britain from this scenario... Now there are councillors regularly elected with just 20 something of the vote and we have MPs elected with just 29%"

    But my point is that it makes no difference, right? You'll still end up with winning candidates with ~20% of first preferences.

    The fact is that even under AV the winner in that case will command the popular support of less than a third of the voters, just as they would under FPTP.

    ReplyDelete
  14. "[YES are] arguing FOR the greater proportionality of AV, while also claiming that extremists won't get elected. It's disingenuous to make both arguments at the same time"

    Alun, it really isn't. The BNP and the communists are campaigning for a NO vote. While the Greens and UKIP are campaigning for a YES vote. Why do you think that is?

    Basically extremists (like the BNP and Communists) with no hope of ever getting 50% support reject AV, because they know FPtP might give them seats with just 20 something percent of the vote.

    While smaller parties with potentially widespread appeal (like the Greens and UKIP) know they can win more seats with AV.

    Also AV will reveal people's true first preferences. You argue that this might make AV more disproportional. But to say that FPtP is more proportional because people can't vote for who they really want so vote for the main parties is an alice in wonderland sort of argument.

    ReplyDelete
  15. To me tactical voting means trying to avoid a 'spoiler effect' scenario where 2 candidates with similar policies split the majority vote and lead to a candidate the majority hate being elected.

    This happens regularly and there is tremendous pressure to vote tactically under FPtP. What makes it worse is that under FPtP, a lot of tactical voting is still ineffective because it is difficult to guess how everyone else might vote, even using past results as a guide.

    AV definitely removes the spoiler effect because now you can use ranking to prevent it. That is not a red herring, it is the truth.

    ReplyDelete
  16. "The BNP and the communists are campaigning for a NO vote. While the Greens and UKIP are campaigning for a YES vote. Why do you think that is?"

    I think it is because both UKIP and the Greens support proportional representation, and both believe that AV is a step towards PR. Neither party actually supports AV.

    "Basically extremists (like the BNP and Communists) with no hope of ever getting 50% support reject AV, because they know FPtP might give them seats with just 20 something percent of the vote."

    Again you're trying to hold opposing positions at the same time. In your world AV gives seats to small parties that you like, but does not give seats to parties you do not like. Here's the thing, AV will NOT give seats to small parties UNLESS those small parties have a geographically CONCENTRATED support.

    That is a function of SINGLE MEMBER DISTRICTS, and NOT a function of how votes are cast or counted. For that reason both FPTP AND AV share the same weakness, parties like the Greens and UKIP, with reasonable levels of support find it VERY hard to get MPs elected, because it is very difficult for them to cross the threshold of wining a plurality. Caroline Lucas was the exception rather than the rule. But parties with a geographically CONCENTRATED support base CAN cross that threshold with relatively small levels of support. he obvious example in the UK is Plaid Cymru's result in 2010 compared to, say, UKIP's result. UKIP got more than five times more votes than Plaid, but they won no seats at all, on the other hand Plaid won 3 seats. AV will NOT change that result. The Greens in Australia have polled over 7% of the vote in the last three elections to the House of Representatives, and yet the last election (with >11% of the vote) was the first time they elected an MP. On the other hand the Liberal Party in the UK has NEVER been without parliamentary representation, and it's vote share has been really quite small in some elections. This has more to do with how votes are distributed, than it does with how votes are cast or counted.


    "While smaller parties with potentially widespread appeal (like the Greens and UKIP) know they can win more seats with AV."

    No they can't, they can only win more seats by concentrating their vote geographically. The ONLY way for them to win a FAIR number of seats comensurate with their level of popular support is to introduce multi member constituencies. And that is simply a mathematical fact.

    But as I say, you can't have it both ways.

    Most supporters of AV really want PR. AND it's disingenuous to pretend that AV (or PR) will help to elect the "good" extremists, but will keep the "bad" extremists out.

    Part of the problem the Yes campaign are saddling themselves with is that they are trying to be all things to all people. They are making wild and clearly untrue claims for AV. Claiming it is more proportional, but won't help elect extremists is just such a claim.

    People will just not trust the again if they can't be honest.

    I'll be honest, PR WILL help to elect extremist parties. With PR we may well see BNP MPs in parliament. So what? That's democracy. For every BNP MP there will be a Communist or a Green balancing them. PR will mean that people will have more choice of candidate, and will see a candidate from the party they vote for elected, more often that not, something that does not happen either under AV or FPTP.

    ReplyDelete
  17. "tactical voting means trying to avoid a 'spoiler effect' scenario where 2 candidates with similar policies split the majority vote"

    You see that is not how I would define tactical voting. I think tactical voting is about people voting to keep a candidate OUT. As Billy Bragg tried to do. He set up a website where Labour and LD voters could "swap" their votes in order to keep Tory candidates from winning. This was about keeping the Tories OUT, it was not about preventing the vote being split, the LDs and Labour actually being two different parties with very different ideologies.

    But LD and Tory voters will just as much engage in tactical voting to keep the most popular candidate out.

    The fact is that if political parties really wanted to prevent a spoiler effect, then they could have an electoral pact. In that case, the Labour party and the LD party could, for example, choose not to run candidates against each other in certain constituencies. The Tories and the LDs could do this also if they chose. This happened in a Canadian election some time in the 1920s, I can't remember the actual election off the top of my head, but it was a provincial election. The Conservative Party won something like 42% of the vote, and was far and away the biggest party in terms of vote share, but won no seats at all, because a plethora of smaller parties had chosen not to run against each other in each constituency. But Labour have always refused to do this, and the LDs are so stupid as to waste money putting up candidates in constituencies they cannot win.


    "This happens regularly and there is tremendous pressure to vote tactically under FPtP."

    Well, if you mean that sometimes a person will tend to vote for the candidate they find least objectionable, rather than the candidate they prefer, so that they don't waste their vote, then I agree. But I do think there is a fine line between a "tactical vote" and simply a realistic vote. The whole tactical voting thing is blown out of all proportion. All electoral systems allow some form of tactical voting, some systems can make tactical voting more purposeful than others. But in the end, for tactical voting to work, it has to be a positive decision by voters en mass to vote tactically, and the fact is, most people simply do not vote tactically, that's a fact.

    The fact is, we are seeing LESS tactical voting than ever before. When Labour and the Tories were getting 90% of the vote between them, THEN we were seeing tactical voting on a massive scale. Now we do not see it, and the fact we do not see it means that we need a system that allows third, forth and fifth placed parties to get a FAIR level of representation in parliament, because they are getting large numbers of votes. But AV will not provide that solution.

    ReplyDelete
  18. Alun, this idea that candidates get less than 50% support under AV is misleading. Voters who vote for losing candidates under the present system have no choice - their vote is completely wasted despite their best intentions to influence the result.

    Wheras with AV they could still continue to rank candidates but CHOOSE not to. Therefore they have consciously abstained, perhaps more so than people who do not vote at all. To count these votes in the final total makes no more sense than counting non-voters.

    ReplyDelete
  19. "Alun, this idea that candidates get less than 50% support under AV is misleading. Voters who vote for losing candidates under the present system have no choice - their vote is completely wasted despite their best intentions to influence the result... Wheras with AV they could still continue to rank candidates but CHOOSE not to. Therefore they have consciously abstained, perhaps more so than people who do not vote at all."


    Here is the definition of a wasted vote:

    1) Any vote which is not for an elected candidate.
    2)Any vote which does not help to elect a candidate.

    This is a standard definition. Under FPTP there are two ways to waste a vote. You can vote for a candidate that does NOT get elected, in some constituencies, then, more than 50% of votes are wasted, for example if a wining candidate only requires a plurality of the votes (e.g. 40%), then a majority are wasted (60%). The second way is to provide surplus votes to a candidate, so if a candidate wins 70% of the vote, 20% of their vote is wasted, added to the 30% of votes going to the unelected candidates, that means that 50% of votes are wasted.

    But under the form of AV we have chosen, there is a third way to waste a vote, by having your vote NOT transferred. This is still a vote that has been cast, and it has NOT gone to a winning candidate, so it is a wasted vote.

    An abstention is when someone DOES NOT VOTE AT ALL.

    Under AV in Australia about 50% of votes are wasted. This is the MINIMUM number of votes that can be wasted in any single member district election, and is significantly better than under FPTP in the UK, where as many as 70% of votes are wasted.

    But in the form of AV the coalition have chosen, it is likely that MORE than 50% of votes will be wasted, because there will e significant numbers of votes that do not transfer. These are IDENTICAL to votes that have gone to losing candidates in FPTP elections.

    Pretending they are abstentions is really not a very good argument. They are wasted votes, identical to wasted votes in FPTP.

    The ONLY way to make sure that LESS than 50% of votes are NOT wasted is to introduce multi member constituencies. THEN the threshold for election is reduced to BELOW 50%, and so any party with a supermajority will simply elect more than one MP, and any party with a substantial vote share, but less than 50% will still win a seat. The more seats a constituency has, the less the number of wasted votes there are. The size of the electoral threshold will depend on the number of seats and the formula used to determine the quota (e.g. Hare quota or Droop quota etc.) but roughly it will be something like this:

    3 seat constituency...33% to win a seat
    4 seat constituency...25% to win a seat
    5 seat constituency...20% to win a seat
    10 seat constituency..10% to win a seat
    20 seat constituency..5% to win a seat

    etc.

    I think 20 seat constituencies would be a good size, because it would prevent tiny splinter parties getting elected, but would allow parties with robust minority support to get elected. Having constituencies would also allow for regional parties with high levels of geographically concentrated support to get some representation (a single state wide constituency would not allow this).

    ReplyDelete
  20. "I think it is because both UKIP and the Greens support proportional representation, and both believe that AV is a step towards PR. Neither party actually supports AV".

    The BNP and Communists both support PR but oppose AV. UKIP and the Greens both support AV over our present system because they believe they can pick up enough 2nd and 3rd preferences to give them more of a chance of winning seats. The BNP and Communists know they won't pick up many 2nd and 3rd preferences (because they are extremists). It has nothing to do with me saying parties I like will win seats and parties I don't won't. It is to do with the facts, which is why the parties themselves are holding the positions on AV they do. You are trying to suggest that the parties don't know what they are doing. They do and it is backed up with research from the British Election Survey and IPPR research.

    ReplyDelete
  21. Alun, we both know that AV is actually a way of exercising an 'exhaustive ballot' without the cost and time involved. So we have election round after round wittling it down to 2 candidates or when someone has a majority.

    If someone doesn't turn up to vote, they have not voted in that election. If someone decides not to rank beyond a certain point they have not voted in that election round.

    When a vote is wasted under the present first-past-the-post system, it is as a result of how the system works, under AV it is because someone DECIDED not to rank a candidate. Can't you see the difference?

    The voter had the choice under AV to continue their vote, but CHOSE not to. That surely is exactly the same as deciding not to vote, they are abstaining. It is not the same as your vote being wasted.

    ReplyDelete
  22. Alun, by the way I agree that PR in multi member constituencies is much fairer, but AV is progress and it is all we are being offered. The establishment concede through gritted teeth as little as they can get away with. But if we don't take it, we will have to wait longer and still only be offered AV.

    Are you suggesting we vote to keep the present first-past-the-post, or are you arguing not to vote at all?

    ReplyDelete