47 thoughts on “on the advantages and disadvantages of a single nationwide constituency

  1. Well since I started the whole problem by mentioning, in relation to what I still think was a legitimate concerns about senatorial election flaws, that I would prefer at large national elections, I guess I should help the move over here.

    What I said on the other thread was that in a perfect world, I would want a bicameral parliament. The upper (non-government forming) house would have members elected from SMD’s. The lower (government forming house) would be elected proportionally, preferably from a single nationwide constituency.

    Why? Because the house that controls and forms the government should represent the people, all of the people. Right now, Tony Abbott is claiming that the Australian people gave him a mandate to repeal the Carbon Tax. Is justification is that he will control the new House of Representatives. The problem? Somewhere in the neighborhood of 55% of Australians voted for something other than Abbott’s Liberal National Coalition. 45% a mandate doth not make. (Just for clarity, whenever I use the word “majority” I mean more than half and only more than half.)

    Now don’t get me wrong, I understand a desire for political stability. If I were able to create my ideal system, Tony Abbott would have every right to be actively forming a government with the expectation of taking power. In fact in my ideal system, as soon as Kevin Rudd publicly conceded defeat, Tony Abbott would have been called to form a government by the Governor-General because his party clearly has the best possibility of successfully surviving a proportionally elected House.

    I personally do not like the idea that the government might need to actively show that it has majority support at every given moment. That is not saying that I think votes of no confidence shouldn’t be allowed, I just think that only actual votes of no confidence should mean anything. A government should be assumed to have the confidence of the House until the House clearly states that isn’t so in a specific vote of no confidence or by refusing to pass any sort of budget that the government can live with.

    But at the same time I hate false majorities even more. A minority government and a system that allows them is something I would sacrifice in a heartbeat for a system that prevents a party from getting a majority of seats without a majority of votes. (And equally importantly, only a minority of seats from a majority of the votes) The people cannot be said to be the ones deciding on a government if parties have exaggerated or undervalued strengths in parliament. Parliament should not just be an electoral college that also happens to be the legislature. Parliaments should be the representatives of the people overseeing the government of and for the people. A parliament that does not represent the people cannot truly do any of that.

    If Australia had an election for an independent president alongside its parliamentary vote the other day, there is no doubt in my mind that Tony Abbott or whoever the Coalition candidates was would have won. I have no problem with Tony Abbott having the chance to govern. I have every problem with him having a majority without a true majority, an exaggerated majority from his two-party preferred vote and a system that reduces the voices of everyone who did not vote for the Coalition. The LNP, campaigning only in Queensland, has 21 seats on 8% of the vote. The Greens, campaigning nationally, have 1 seat on almost the same percentage. That is the same number of seats the Country Liberals have from campainging only in the NT and receiving 1/20th of the Greens’ vote. The Nationals have 9 seats on half of the Greens vote. Palmer’s United Party has one seat with 6% of the national vote, so the under-representation falls on both sides of the political spectrum.

    Why are they not nearly equally or fairly represented? Why are the voices of people who are spread out undervalued compared to the voices of people who all live next to each other? Why does being the most popular (or least unpopular given the election system)candidate in an area make one more qualified to determine the fate of the nation than the people a larger group would support if they were all lucky enough to control an electoral division?

  2. In answer to the question that seemed to have been asked without my notice, I am torn on the question of thresholds. Ideally I would not want one, but if the chamber is large enough they may be necessary. If I had full control to create a system, I would use optional preferencing, optional above the line tickets, optional ranking of above the line tickets, and the single transferable vote. All of which would make thresholds seemingly unnecessary. Though even then, I would at least have to listen to suggestions of automatically eliminating parties collectively receiving less than a certain percentage of the votes and transferring their ballots. Of course, as STV is designed to elect individuals and not groups, that would strike me as being unfair.

  3. I guess my support for single, national districts comes from (in my opinion) the elegance of having a blank slate filed along whatever lines the voters deem appropriate. If regionalism is a factor, regional parties will play an important role. You have enough proportionality to express very fine details, if needed. Thresholds certainly act against this, but they might be necessary anyway.

    I’m quite fond of the idea of a run-off, in place of thresholds. If needed. Allow for a post-first round coalition-building phase, then have only the two complete proposed coalition governments compete, with the seats divided proportionately between the two based on the run-off, then internally based on the results of the first round. With a very strict constructive-only confidence system, and if you had cabinet members assume office directly as proposed in the coalition agreement should their coalition win 50%+1 with no opportunity for post-second round negotiation, you could probably get away with fairly diffuse coalitions.

  4. The “coalition runoff” is an interesting idea, but if one were to do that I might suggest having a separate election for an executive council, for lack of a better term.

    Decide on the number of seats needed, which I would say should be more than the number of necessary cabinet posts but not necessarily much higher, and have an election by some form of proportional representation. Whoever gets elected to the council governs collectively. If run right, it may be possible to have the best of both worlds of regular cabinet government and Swiss style consensus government. Whatever party/coalition controls the most council seats and the big posts gets to have a major influence on policy making but the opposition can do more than simply talk if it possesses real votes in the cabinet room.

    Or you might just get a really small majority coalition making every decision by forcing a majority vote and ignoring the opposition.

  5. Above the Line voting for sane people, with all independents given their own box and the Above the Line section given its own page.

    For the insane people who would want to rank hundreds of candidates, they can ask for the full ballot, which might get several pages to make it legible.

    Alternatively, I can accept that STV has its limitations and accept another system. Preferably not closed lists though.

  6. As Mike noted, threshold or no threshold is a big question here. At least as important are assembly size and territorial extent.

    It is one thing to have an assembly of 120 or 150 elected in one district when your territory is fairly compact. It is quite another when either your assembly size up in the hundreds or your country spans vast territory (or both, as in Ukraine in 2006 and 2007 or even Russia for a while).

  7. More to the point, this question came up in a discussion of Australia, and Alan offered this thread as a place to discuss it outside of the Australia thread.

    And that makes sense, because a national district for either chamber in Australia doesn’t. With the world’s sixth largest geographic area among independent countries, and a federal form of government, Australia would have to be one of the least likely places for a district consisting of a full chamber.

    It seems Colombia is now the largest country using a nationwide district (for one of its two chambers). It is a pretty large country, and a rugged one with distinct regionalisms (and, although not federal, increasingly decentralized). It is a very unlikely place for a single district. Yet even Colombia has an area about 1/8 that of Australia.

  8. I can’t imagine Australia ever moving to a single district, but if you consider the fact that every state is already a single district for the Senate, and that NSW and SA use statewide upper house districts, the idea of each state being a single lower house district isn’t all that unreasonable. Divisions like Farrer, Durack, O’Connor, and Lingiari are already gigantic, so I can’t see district size being a key factor there.

  9. Yes, in writing my previous comment, it occurred to me that if we included sub-national jurisdictions, Australia would give us two of the largest examples of single jurisdiction-wide districts (NWS and SA state second chambers).

    And as along as we are playing the subnational game, Brazil’s states all have unicameral legislatures with statewide districts. A few of the states are huge in area or population or both (and at least one has a district magnitude of over 100).

  10. Is this the place to mumble about the 221 councillors of the City of Cape Town? They’re elected by MMP but that does make the list district very large indeed in terms of magnitude.

  11. There are 111 wards and 110 list councillors, elected on closed lists as far as I can tell.

    As far as I can tell, opposition councillors are almost entirely powerless and the Mayoral Committee (cabinet) does most of the governing.

  12. Mark Roth;

    That sounds a lot like the Australian system. How would the independents be ranked?

    Combining STV with party tickets (or other kinds of ticket, for that matter) is an idea I’m interested in, so let me throw this out there:

    The basic system is CPO-STV (http://en.wikipedia.org/wiki/CPO-STV). Because CPO-STV is a Condorcet method it can easily take tie votes, and that’s what a ticket will be: a big tie among all its members.

    When the voter gets a ballot, he creates a rank ordering that freely mixes tickets and individual candidates. Tickets can represent parties, or be free-standing, and can overlap. (I picture party tickets that do not overlap with other party tickets and non-party tickets that overlap both the party tickets and each other.)

    The ranking provided by the voter is translated into a rank-ordering of candidates: 1) Any candidate listed individually gets the rank given by the voter. 2) Tickets are treated as tie votes at whatever place the voter put them in. 3) If overlapping tickets appear on the same ballot, candidates who are on both get the higher of the two. 4) Candidates who are neither ranked individually nor as part of a ticket are all tied for the highest unfilled rank, which might be higher than certain filled ranks.

    In another place someone suggested that candidates who belong to tickets but don’t campaign independently might get hopelessly tied, and that this might be avoided by having each ticket provide an ordered list, either of its own members or of all the candidates, and that a bullet vote for a ticket should be read as that list. That’s the only place I’d allow for an ordering that isn’t entirely voter-generated.

    I’m picturing a touch-screen with some kind of drag-and-drop interface, hopefully as a means of generating a paper ballot rather than as a system unto itself.

    I know this strikes people as complex, but from the perspective of the voter it’s really not; all you need to know is how much you like the different options, and then you put them in order.

    (You can also make this a form of MMP. First, elect district candidates using your favorite single-winner ranked preference method. Next, CPO-STV works by comparing completed sets of winning candidates. You can decide in advance to only look at outcomes that include the candidates who win in the districts. Don’t worry, the winning outcome will be one that corrects for any distortions caused by the district votes.)

  13. @ 14, What I proposed is pretty much the Australian system but with a few ideas (more would be needed) on how to clean up the muck in the system.

    For independents, I believe they should be treated, individually, the same way that party tickets are. Each independent should be able to get his own box above the line, which, when coupled with multiple above the line preferences, would enable a voter to select Candidate A, Party Ticket B, Candidates C, Party Ticket D, etc. as he chooses.

    For the tickets themselves, I would have the election officers view them as voluntary lists between candidates. The lists can be named, they can consist of all of the candidates nominated by the party, they can be coalitions. All that would really matter is that they are technically a group of candidates who agree to campaign together and pool votes.

    A vote for a list wouldn’t be a vote for the list. It would be a vote for all of the candidates on the list in the order of the list. So a vote of Candidate A, Ticket B would be read as a vote for Candidate A, Candidate B1, Candidate B2… There would be very few ties as most votes would be for the list leader. His surplus would then transfer. Of course below the line votes may and could change the order.

    FWIW, MMP is one probably my second choice if STV is unwarranted. In the specific case of Australia, I am not sure how the High Court would view the list members unless the lists are all open lists. I have also often wondered if MMP would work with FPTP districts and at-large STV seats forming the two systems instead of FPTP and party lists. I cannot figure how to seat a candidate who won a local seat into a list of individually elected STV candidates.

  14. Strange, Mark. You would use a national district, but you wouldn’t do away with above the line and compulsory preferences? Surely, if you dislike closed lists, you’d do away with above-the-line voting, which makes STV quasi-closed list.

  15. Any functioning proportional system carries the risk that marginal forces that you don’t like will get elected. Thresholds increase people’s comfort level with the result (by reducing the opportunities for broadly unpopular or indeed unsavoury groups to get representation).

    One can’t defend thresholds on any other ground than “we choose to sacrifice the representation of the unpopular or unsavoury to keep the overall notion of proportionality legitimate in the eyes of the overwhelming majority”. So the question in the Australian context is-is this a disaster for democracy that will deligitimise the part of the Australian parliament whose makeup actually resembles the balance of opinion in the Australian electorate, or will these protest parties simply crest and break, as has happened before and will again?

  16. @15
    Given the district size, I’d imagine the results would be indistinguishable from pure closed lists in practice. Closed lists have never bothered me, to be honest. Coalitions are likely to be quite fragmented, so encouraging intraparty competition might not be a good thing.

    Here’s a thought. You could require each party to field a unique list of candidates in each state/region. You could then divvy up all the seats to each party nationally first, then divide up each party’s seat total would among the states/regions in proportion to each state/region’s fraction of that party’s vote total.

    The districts would likely be small enough for pure open lists. If not, you could subdivide further. It wouldn’t have an impact on party proportionality at all.

    You could also have regional thresholds, which may be a reasonable compromise.

  17. @16, I have no issues with above the line voting, as it is done in NSW and would continue to use it. Voting for a party above the line indicates a selection for that party’s candidates only. A voter may then give subsequent parties a secondary preference. I would not require full prerefences at all. Unlike many in Australia, I would allow a vote with only one candidate selected to be counted.

    @18, Even if the results turn out to be 100% indistinguishable from closed lists, I would not use closed lists for two reasons. Firstly, I think individuals should be elected not political parties. An independent should have as much chance, at least in theory, as any party candidate. And a party candidate should be able to get in if enough people prefer him to the party’s preferred candidates. I sometimes see a closed list election as an election to see how many votes each party leader will wield at the negotiating table that writes the scripts for show debates and votes in parliament. The second reason is simply that STV allows for “wasted” votes to accumulate, not simply count for nothing.

  18. Mark, when you said it would be “insane” to vote below the line, it sounded like you were proposing compulsory full preferencing.

    DC: no, there is plenty of rationale for thresholds. Too much proportionality is a recipe for instability and disproportionate power for small ‘kingmaker’ parties. Only I would opt for districts instead of thresholds, with similar results due to the effective threshold, but with additional advantages of better representation due to fewer safe seats and a clearer connection between voters and their representatives.

  19. Professor Anne Twomey, who was instrumental in reforming the NSW legislative council system after the tablecloth ballot of 1999, on the current brouhaha.

    On other matters, I would oppose thresholds. Any system that involves binning the votes of anyone is to be avoided if possible. The microparties problem disappeared after the NSW reforms without any use of thresholds and there is no reason to think the same will not happen with the Senate after reform.

    At minimum I think OPV will be introduced and I suspect ticket preferencing will be introduced as well.

  20. @20, No. I am absolutely against any requirement that might allow someone’s vote to go to a person they do not support because they have to rank that person. I believe that a ballot should be formal and should count if it has as a single preference. “I like Candidate A and no else” should be a perfectly reasonable statement, albeit one that comes with a clear understanding that if Candidate A doesn’t need/can’t use your vote, you have no more say in the election.

    I just would not expect the vast majority of people to vote below the line, especially in an election for a national body without any districts. Even a ballot for a nationally elected U.S. Senate would consist of at least 100 candidates and probably more.

    I might vote below the line, especially for a party that I support but whose leader I am not a fan of. I would never force anyone else to vote above or below the line, they should just have the option. I have a general preference for making ballots as idiot proof as possible by allowing the greatest range of voter choice and means of marking a valid vote.

  21. Fully optional preferences might work better with the rule (used in Northern Ireland and, back in the 1980s, by the University of Queensland Senate before it reverted to multiple-X voting and no, I am not sure why this particular method should so appeal to dour Orangepersons in two different hemispheres) that exhausted votes are deemed divided equally among the remaining candidates. This avoids the need to reduce the quota – which, I’m reliably told, is an administrative nightmare – while still ensuring that every candidate is at least notionally elected on a full quota rather than a near-quota “plurality” (ie, largest remainder).

    One partial compromise might be to reduce the quota until an elected candidate’s surplus is distributed, and freeze the quota after that. In Australian STV elections, however, this would be a moot point because one or two candidates are almost always elected with surpluses on first preferences anyway – those with the most popular support in Tasmania and the ACT, those who have been in bed with (in varying senses) the most party executive members for the federal, NSW, WA and Victorian upper houses.

  22. I seem to be in a distinct minority when it comes to preferring at large national elections. If I may, I would like to make two more points.

    Firstly, I wish to reiterate that my system is not as bad some believe. Only one house, albeit the government house, would be elected at large. The other house, which should have all of the same powers except the vote of no confidence, would be elected from single member districts. At the same time, certain aspects of the government would change. Minority governments should become clearly permissible, formed by either the largest party or the largest coalition. If at some point in a term the governing coalition collapses or another, larger coalition forms than the government should change. At the same time, the government only has two responsibilities: to not lose any express votes of no confidence and to pass a budget. As long as it manages those two things, it is presumed to have the confidence of the house.

    Now to my other point, if a nationwide PR election is not desirable, would there be more support for MMP with the list seats assigned on a national basis, as in New Zealand?

    And on that note, I have a question relating to Australia. Would a New Zealand style MMP system be constitutional in Australia? If not on its face, could it be made constitutional if instead of voting directly for a list, voters chose a specific candidate on that list and then the seats were awarded to: the winners of the electoral divisions, the list candidates who received full quotas on there own, and then to lists as a whole via “transferring” surplus votes and votes for candidates not directly elected to the next candidate on the list?

  23. Mark: I believe MMP would be constitutional in Australia if:
    1.All seats (including lists) would be divided by state.
    2. There were no overhangs (so that no state is overrepresented) – wither an elaborate compensation system or the British-style ‘AMS’ MMP
    3.Lists would be open.

  24. @21 I don’t think a threshold would bin anyone’s vote, I think it would transfer their vote to a continuing party. A 1% threshold would eliminate many of the microparties from contention for a seat.

    It is possible that if microparties are being used as front groups that they would still be on the paper to funnel votes to larger groups (for instance, even with a threshold, the Stop The Greens Party and the Republican Party would probably stick around to funnel votes to the Liberal Democrats), but the incentive to form a new microparty because you think you could bundle preferences to get elected is significantly lessened.

  25. Thresholds are a solution in search of a problem.

    If thresholds, unlike ListPR or MMP thresholds, are transferable then they are ineffectual and there is nothing that a threshold could achieve that would not be achieved better by giving electors control of their own preferences through OPV and ticket preferencing.

    If thresholds are not transferable they are fundamentally undemocratic.

  26. @28, Well said.

    STV comes with its own unique system for dealing with low supported parties, votes for those parties naturally transfer to someone else. Everyone is elected with a full quota of first or later preference votes. In other words, every senator has or will have at least 14% support to get his/her seat.

    The problem is when the average voter needs to read hieroglyphics to figure out where his vote is going. And that is assuming that the party he chose as a single ticket and his vote can be traced.

  27. I don’t see how a threshold is ineffective if transferable. If a 4% threshold were used (that seems to be the most common number suggested in Australia), if your first preference group fails to get 4%, your vote would transfer to your highest preference above 4%. That would substantially reduce one of the incentives for microparties to stand for election–that the “preference snowball” will convalesce around your party and you’ll manage to get elected even though you only got 0.22% of the primary vote. It would not eliminate LDP-style “affiliate parties” which try to attract voters with a catchy slogan and funnel them on to another party, but it would substantially reduce the number of parties on the ballot without binning anyone’s vote.

    I don’t think a 4% threshold is necessary–it strays too hard from the principles of “make the ballot more manageable” and “make sure candidates are elected by popular will rather than chance and backroom deals” and becomes “let’s keep the nutters out and get an extra seat in every state for the duopoly.” I don’t think a 1% threshold would be out of line though. You would still see small parties that have legitimate bases of support like the LDP, Sex Party, Family First, and PUP survive the threshold elimination, but there would be little incentive for the Sorts Party and Motoring Enthusiast Party and Bullet Train Party to stand for election as they aren’t likely to get even 1%.

  28. What would happen if an Australia state for it’s Senate and/or Australian Federal Senate instead of using STV changed it’s electoral system towards a CPO-STV system? Would this reduce vote harvesting?

  29. Australian states don’t have the legislative power to adopt their own system for electing senators. The High Court has ruled on a number of occasions that the Commonwealth Electoral Act covers the field and pre-empts any state legislation with respect to federal elections.

    Unless you disqualify all groups below the threshold and transfer their preferences immediately, a threshold simply makes no difference. If you are advocating a pre-emptive threshold, then the candidates are being elected by the legislation, not the people. One of the things that made modified D’Hondt in the ACT so unpredictable and a horror to count was a 5% pre-emptive threshold.

    That system was also extensively gamed and also led to the election of unexpected candidates with minuscule support.

    I understand how the thing would work. What I don’t understand is how it makes the election of senators fairer or more transparent, or why you would advocate thresholds when the preference harvesting issue can be better and more fairly addressed by OPV and ticket preferencing.

  30. @31, The only theoretical power that states have over senate elections is that in theory a premier could advise a governor to hold the state senate election on a different date. If a premier did that, and the governor allowed it, it would be a constitutional crisis.

    @32, Wasn’t the original ACT election system complicated even more so that candidates who were provisionally eliminated for falling below the threshold could be “uneliminated” at some point if certain conditions were met, screwing up the already transferred ballots that preferenced those candidates. Or am I just misinterpreting a confused system?

  31. The counting system would be exactly the same as now–enter all the votes into the computer and hit the button, and the computer tells us who gets elected. As far as what the computer is doing, it’s pretty simple: after counting first preferences, exclude all candidates from any group that failed to reach the threshold, and any ungrouped candidate who failed to reach the threshold, and distribute their votes to the highest remaining preferences. From that point on, it’s standard STV. I certainly wouldn’t recommend a quota as high as 5%–I’d go with either 1% of formal votes or 10% of a quota (20% for a double dissolution).

    I was under the impression that the biggest problem with modified d’Hondt was that votes bounced back and forth between parties and candidates after the threshold cull, which constantly changed the number of seats parties got and which particular candidates got them. If they had simply done threshold, transfer to highest remaining, and then d’Hondt or threshold-transfer-STV, I think it would have worked fine. Instead, the Senate tried to combine d’Hondt and STV and failed miserably.

    I think it achieves two key goals: reducing the number of groups on the ballot paper, and making sure that the candidates elected have at least some popular support rather than simply winning the luck of the draw of the micro party preference snowball.

    I’m far from completely opposed to ticket preferencing, but especially with a 14.4% quota, you’d see huge numbers of candidates elected at far below a quota, which is certainly not a proportional result.

    Under a threshold system, the minor parties with legitimate support would continue past the threshold, and the micro parties would transfer their preferences to bigger parties. What you wouldn’t have is the Sports Party coming 4th from the bottom, and then progressively gaining the preferences of 18 larger parties to win a seat in the Senate. They’d be eliminated at the threshold cull and their ballots would transfer to a party with at least some community support, or more likely, they don’t even nominate candidates in the first place.

  32. The Sports Party would also have no hope at all under the NSW system. Why then do we need thresholds? The problem with the ACT thresholds was not that it made counting difficult, but that it made voting difficult. Microparties were elected in considerable numbers. The only difference was that the snowball hit microparties who met the threshold. There were still weird and wonderful preference harvests and there was still collusion between parties that had nothing in common.

    Seriously, I’m not looking for explanations of how the Sports Party would lose out under thresholds. I’m looking for an explanation of why you’d adopt thresholds instead of OPV plus ticket preferencing.

    Thresholds have snowballs. They melt under OPV. Your proposal reduces the choices the electors can make. OPV increases them.

  33. There were no microparties elected in either modified d’Hondt election. There were local parties elected, but they all had substantial popular support. The anti-self government movement can hardly be called microscopic when it got 19% of the first preferences in 1989 (in two different parties).

    A threshold doesn’t have to make voting more difficult–if you don’t get rid of GVTs, it’s still “put a 1 against the party of your choice.”

    I’m thinking of a system which gives minor parties a chance while discouraging micro parties from entering. The problem with ATL-preferential in the Senate is that it will result in disproportional largest-remainder style results and that it will be exceptionally difficult for minor parties to establish themselves. The Greens likely never would have emerged under such a system. I don’t think entrenching the existing party system by using disproportional elections is good for democracy.

    I do think it’s good for democracy to make sure that the parties elected actually have community support rather than being the micro-party that won the lottery, which is the kind of distinction that a threshold can make. The preference snowball is greatly lessened by the fact that micro parties that have no chance of reaching the threshold are substantially less likely to contest the election in the first place.

    OPV does give more power to the voters, but it will likely take away from proportionality. There is no perfect solution, but a threshold would be a “do the least harm necessary” reform that makes the system more democratic and less stasiocratic without impacting proportionality.

  34. Sorry, both modified D’Hondt assemblies included a number of groups with very low support that existed only from election to election. An example is the Abolish Self-Government Coalition whose sole platform was to abolish the ACT itself.

    To quote Malcolm Mackerras:

    The system sounds quite reasonable until it is realised that the cut-off severely penalises independents. Although the top-scoring candidate, Rosemary Follett, was elected to the Legislative Assembly, the second, third and fourth most popular candidates were not. These were the independent candidates Bill Mackey, Tony Fleming and Ken Fry who got 5686, 5269 and 3977 personal votes respectively. In contrast, David Prowse won only 54 personal votes, yet was elected and ultimately became Speaker of the Legislative Assembly.

    I say someone threw a snowball at Prowse, but of course that is impossible under a threshold. Ummm…

    It is impossible to imagine how you think what you are proposing is not a radical and drastic change to the electoral system or how you think that avoiding the supposed evils of OPV justifies this wild leap in the dark.

    It is no accident that the people of the ACT decisively rejected this system by referendum and entrenched OPV by a large majority. It is also no accident that, as far I can tell, no existing STV system uses thresholds.

  35. @37 per Alan:

    “Although the top-scoring candidate, Rosemary Follett, was elected to the Legislative Assembly, the second, third and fourth most popular candidates were not.”

    While STV is not and doesn’t claim to be an FPTP system (like SNTV), when electing seventeen individuals you want a reasonably high overlap between the two Venn Diagram circles labelled, respectively, “candidates who poll the most first preferences” and “candidates who end up getting elected.” Ironically, the Hawke Government’s stated reason for not adopting STV – that it had led to candidates with low first-preference support getting elected to the Senate ahead of higher-polling candidates – applied in spades to the Modified D’Hondt system it devised for the ACT, with the examples Mackerras notes. Canberrans (who on average are the best-educated and most politically savvy jurisdiction in Australia) disliked the system from the start, voted by a large majority to replace it with STV, and voted to entrench STV just three years later.

    “It is also no accident that, as far I can tell, no existing STV system uses thresholds.”

    I seem to recall once reading that one or another of the US cities that uses STV – Cambridge, Massachusetts perhaps? – has, or at one point had, a rule that any candidate with fewer than (I think it was) five hundred first-preference votes was automatically excluded on the first count. Further, I have some idea the rationale for this was that when that city adopted STV to replace MNTV, it also replaced nomination by petition signed by 500 voters with a much lower threshold (say, 10 or 20 voters only). In other words, the 500-votes seats threshold was meant to reassure voters that no candidate who didn’t have at least five hundred friends and admirers could end up on the city council.

    This may have been in Hoag and Hallett, or John J Humphreys, or one or another of the various older pro- or anti-PR works I speed-read-ed in various town and university libraries across Australia two decades ago – sorry I can’t source it any more precisely.

  36. @36, While I have repeatedly stated a preference for STV, I feel that it is odd to say that there is something unfair about Largest Remainder. If voters choose not fully employ their options, isn’t giving the last seat to whoever has the most votes the best available option?

    And to anyone, is there a webpage out there that explains how the ACT D’Hondt system worked?

  37. Alan, the Abolish Self-Government Coalition got over 7% of the first preference vote, the lowest elected in 1989. That’s not something I would call a “microparty.” Minor, certainly. Prowse, on the other hand, was the 3rd candidate elected for the similarly-named No Self Government Party, which earned 11.5% of total first preferences. His election was no more ridiculous than a Labor or Coalition senator being elected with only a few hundred personal votes, and substantially less ridiculous than Ricky Muir or Wayne Dropulich being elected with less than 1%.

    @38 Cambridge does indeed have a threshold, but it’s only 50 votes, which is the number of unique individuals who had to sign the nominating petition anyway. The logic is that if you don’t get 50 first preferences, it’s as if you weren’t even nominated to stand for election. This cull comes after the surpluses for any candidate elected with over a full quota are distributed, IIRC.

    The full electoral law is available here: http://www.rwinters.com/docs/chapter54A.htm

  38. @39, there is nothing in general “unfair” about largest remainder. That doesn’t mean it’s particularly proportional when m=6. With so few voters distributing preferences, you’re going to see the final seat elected with around half a quota, which is around 7% of the vote. This is substantially different than in NSW, where the 21-seat election means that only 2-3% is needed to win the final seats.

    OPV does not at all diminish the average voter’s reliance on how-to-votes to determine their preferences, and can result in many voters being somewhat disenfranchised by following cards which recommend only a single preference above the line rather than distributing their preferences to a party they may not know much about but which is broadly ideologically similar to them. The GVT system has the disadvantage of being gamed by parties to some extent, but at the very least it does not result in voters’ votes being binned as exhausted. This will be magnified if exhausted ballots are held as part of the quota and those where people distributed preferences are not.

    The CDP’s HTV cards for the 2011 election even stated “voting 2 will not do,” which is somewhat counter-intuitive as those surplus votes can absolutely make the difference (http://www.electionleaflets.org.au/leaflets/990/). When the system essentially has parties misleading voters into exhausting their ballots rather than giving them the option to vote for multiple parties, it’s hardly more democratic.

    The GVT system is far from perfect, but the ticket-preferencing system is not a perfectly democratic panacea either.

    If the stated goal is to make the ballot paper more manageable and to ensure that parties without at least some community support are not elected, a small threshold would solve that problem. I don’t think that the DLP, Family First, and Christian Democrats all sending preferences to each other, and their voters only needing to vote 1 above the line to do so, is a particular problem for democracy.

  39. @40: Good to see Abdul Alhazred getting some work at the AEC…

    General question about these opportunistic tickets-can a party be included in a preference against its will…I don’t know, can the Australian Fascist Party include the Australian Communist Party in its preferences without the latter being able to object?
    Has there been any notable political impact when a party, particularly one of the larger parties preferences a minor party that may rankle with its supporters/potential partners (someone cited the ALP in Victoria preferencing the Greens and then Family First-two groups that I would imagine do not get along under normal circumstances?)

  40. @43, Since parties HAVE to preference every other party on the list, even if last, no party can really object to where another party places them. If the Wikileaks issue revealed anything, there isn’t even much that can be done if one party doesn’t live up to its end of a mutual preferencing deal.

    As to your specific example, if the Animal Rights Party can go out of its way to preference against the Greens, I don’t think anyone would bat an eye if the Communists and Fascists struck favorable deals with each other.

  41. DC @43: No, there’s no legal recourse to getting a Like from unwanted admirers. (It sometimes happens, eg, if the Greens, who want to reduce or stabilise the population of the whole planet and also to take in refugees, receive unsolicited preferences from Australians Against Further Immigration, who want to reduce or stabilize the population of Australia, primarily by keeping out refugees).
    Since Senate tickets, like individual voters, are required to number every candidate (if anything, the conditions for tickets are slightly more stringent since below-the-line voters are forgiven if they number at least 90% of the candidates), there is no bright line between “We’re putting Team X second-last because we hate, despise and detest them nearly as much as Team W, whom we’re putting last” and “We know our seventh to seventy-seventh preferences are going to go through these no-hoper fringe candidates like an enema through a goose: what really matters is that our preferences will end up parked with Team X, whom we’ve ranked ahead of Team W, who are our only real rivals.”
    (As recent elections have shown, the assumption that minor candidates will not survive long enough to inherit as beneficiaries under your will may prove over-confident, but that’s a political judgment, and as a legal matter it would be impossible for a court or electoral commission to distinguish the two motives for putting a team second-last).
    Ie, it’s not like joinder of lists under PR-List (which Senate STV resembles in some of its effects) because the recipients don’t need to formally ally with – or even,for that matter, much like – each other.

  42. @44 & 45: So in effect there are no political consequences for the parties making the decisions about where they preference who…could it be that’s why not much progress has been made on reforming the situation?

  43. Not having any legal recourse is not the same as not having any electoral recourse.

    I suspect,for instance, the LNP suffered for their decision to preference the Greens below everyone else, including almost neo-Nazi microparties. There is some evidence that KAP suffered for the same decision. The parties do get criticised, held to account, whatever, for their preferencing decisions.

    The reason there is so little legislative action has more to do with the irresponsibility of the major parties than anything else, especially under the leadership of Julia Gillard and Tony Abbot. Bear in mind that when the Gillard government proposed automatic enrolment to the parliament the opposition denounced it as an attempt to rig the election.

    Bob Brown took the preference harvesting issue to Julia Gillard early in her prime ministership, proposing OPV+ticket preferencing, and was told that it could not happen in federal elections.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.