Party-preferential voting

I had missed most of the following discussion earlier, and as it occurred in a thread on Lower Saxony, it could easily have been missed by others as well.

I am going to reproduce two comments that describe interesting ideas for coping with thresholds.


there’s no reason you couldn’t make thresholds less discontinuous, by combining them with preferential votes.

The mechanics would be a tweaked STV, with the following differences:

1. Because of large district magnitude, a full preferential vote for all candidates would be impractical. Instead voters would list parties in order of preference, with the intraparty order being fixed as in List-PR.

2. To effect a threshold of T seats, a party would not be awarded its first seat until it has accumulated T quotas.


Vasi, that sounds essentially like the NSW Legislative Council electoral system.

The election is by STV, with 21 vacancies at an election. One has the option of either voting below the line, by ranking at least 15 candidates, or voting above the line, by voting for one or more party tickets. Unlike federal Senate elections, voting for a party’s ticket does not result in a vote for a preset preference ranking of every candidate; instead, it only ranks the candidates of that party, in the order they appear on the ballot paper. Voters have the option of marking multiple parties above the line, unlike federal Senate elections. So, for instance, the Labor how-to-vote cards in the last election suggested that their supporters vote 1 Labor, 2 Greens above the line. That means they essentially ranked every Labor candidate, followed by every Greens candidate, and if all of them are elected or excluded, their ballot is exhausted.

The Australian group voting ticket essentially operates like closed-list PR, with the exception of in very large elections. The NSW Legislative Council used to use the same ticket style system that the federal Senate uses, but after the 1999 election resulted in a ballot paper the size of a tablecloth (almost 1 sq. m), and a candidate from the “Outdoor Recreation Party” got elected with 8,000 first preference votes (something like 5% of a quota), they changed the group-ticket system to the single party ticket system now in place.

Stephane Dion, the former Leader of the Official Opposition in Canada, also is advocating a version of party-preferential voting. though in ridings which would be only 3-5 seats (1 seat by AV in the territories). In a 4-seat riding, the threshold would be 25% + 1 vote. If all remaining parties are above the threshold, seats are awarded to them by largest remainder (I believe). If there are any parties remaining below the threshold, the party with the least votes is eliminated and their supporters votes transferred to their highest remaining preference. His system is OLPR, with each voter able to cast a preference vote for one candidate of his first-preference party.

I think the most proportional system possible would be party-preferential with a low threshold and a large district magnitude (the most proportional would obviously be a single national district). You could either exclude parties one-by-one (hopefully with block exclusions) until every party remaining was above the threshold, then distribute seats. Otherwise you could simply exclude all parties below the threshold and distribute their voters’ preferences to remaining parties. It avoids the huge numbers of voters wasting their votes by being below the thresold; for instance, even with a relatively low threshold of 3%, 19% of the valid votes in the May 2012 Greek election were cast for parties below the threshold. In this system, the only voters who do not have either a first preference or a transfer vote elect an MP are those who deliberately choose not to rank any parties that make it into parliament.

I also think a novel way to build a stronger government while remaining representative of votes would be to use preferential ballots, but with multiple thresholds. In a 120 seat legislature, 60 seats could be awarded to those parties above 2%, with voters below the threshold transferring to their highest placed remaining party. Then a further 30 seats could be given to those parties above 5% (including transferred votes), then a further 20 seats to those parties above 10%, and then the final 10 seats to the party which wins a majority by transfers. This means that even voters who vote below the threshold are represented, and parties with a decent amount of support have representatives in parliament, just not proportionally to their first-preference votes. You also get larger parties at the top, making a stable government more likely, but unlike supplemental member, or the Italian/Greek plurality-winner top up system, the larger bloc is distributed based on all voters’ preferences, retaining a much larger degree of proportionality than other semi-proportional systems.

I am not necessarily endorsing this concept, although I do find it very interesting. I would be interested in further discussion.

The thread has a lot of other interesting comments on the relationship of thresholds to democratic theory (particularly the last several comments posted as of 4 February). I re-posted the two comments above simply because they refer to proposals for an alternative way of coping with thresholds in electoral-system design.

(I did note the Dion proposal before.)

77 thoughts on “Party-preferential voting

  1. Wouldn’t giving the three equal first preferences each a whole vote violate the ‘single’ part of STV, and advantage these voters over voters who vote 1-2-3? It also seems like it would be a nightmare for determining who gets excluded.

    I could see that working perhaps in a SMD (though the uneven weight between voters is disturbing), but you still have problems with whom to exclude. One could use a variant like Coombs I guess (if no candidate has a majority of highest remaining preferences, the candidate with the most last preference votes is excluded, even if he’s in first for first prefs as well).

  2. Multiple first preferences, or multiple preferences of any sort could be equated as a voluntary semi-abstention. Perhaps the counting rules could state that if candidates are equally ranked with a voter’s highest remaining preference, the voter is stating that he doesn’t care which one wins. If all but one are eliminated, the vote now counts in that candidates column.

  3. I think the biggest issue with the approval/range voting obsessions is that even American electoral reformers are fixated on single winner elections. This may have been a result of Bush-Gore and the drive for picking a majority winner; I’m too young to remember anything before the 2000 elections clearly, but I haven’t seen much evidence of converted electoral reform efforts from before that (except local semi-proportional fixes bc of voting rights lawsuits).

    It’s somewhat understandable given that most elections in the States now are single winner (even when electing multiple members to the same office, such a numbered posts in citywide council elections), and the few multi winner races are by plurality (for m vacancies, each voter may vote for up to m candidates, and the top m are elected).

    This one-sightedness leads our reformers to come up with more and more convoluted ways of electing a single winner ‘fairly’ when multi winner proportional elections are fairer almost by definition.

    I see no need to have single winner races except for executive positions and very remote and large areas (it’s understandable, for instance, that Alaska is one SMD and Hawaii two SMDs rather than those three seats being elected in a giant three member seat). Beyond AK and HI, and perhaps places like Montana and West Texas, we should be trying to steer the conversation away from single winner (unless talking purely theoretically).

  4. Re the “STV with mutliple preferences allowed” model I described @52 – I should clarify that, if and once the highest of the three reached quota and was elected, the other two would see their votes drop accordingly.

    So, assuming a 5-seat race with 999 voters and quota of 167 votes, if the Purple Party runs three candidates and most of its 400+ supporters vote 1, 1, 1 (or X, X, X [*]) for all three, their respective totals on the first count may be (say) 411, 403 and 397 votes, but as soon as the first one is elected (candidates would have to be declared elected maximum one per stage of the count, even if several are temporarily over quota), the totals for the other two would drop 167 votes each, to 236 and 230 respectively. Then, once the second was elected in his or her turn, the third would be left with 63 votes only.

    Giving a lot of candidates the same preference would be like declaring that henceforth every US dollar is worth one zillion quatloos. It’s an accounting device but it doesn’t change the relative value of what’s there, and eventually mathematical gravity will catch up. Such a preferencing strategy does – I was told by these online Approvalistas – make it more likely that a candidate you do support will keep out one you hate, although it means you’re sitting out the question of which of the supported candidates you actually like.

    (I actually have some time for Approval. It’s about my fourth-favourite voting system – or, as Approvalistas would say, it’s one of the three to six I would tick.)

    [*] I think an X in an election (if there are no numbers, or at least no 1) is pretty clearly a tick or first preference. In a referendum, though, it’s murkier, especially if the ballot only has crosses and no ticks for all propositions. Note that “crossing out” very definitely denotes disapproval…

  5. @Tom

    You are not allowed to have a fourth preference for Approval voting. You can only list it as one of 4 equal first preferences.

  6. I’ve never actually voted with a tick or x before. At high school student elections, we used optical scan cards. At university, we do online voting, by clicking as many boxes as one has votes (I guess checking the box makes a tick on the screen, so I should say I’ve never made a ‘written’ tick or x). At all polity-level elections I’ve voted for save two, I used a touchscreen during early voting; one of the remainder was a party presidential caucus, where I signed in on the list of the candidate of choice (the only PR election I’ve ever voted in, though with an almost obscene 15% threshold), and at the final one, I used an optical scan card on election day (thankfully, only three races on the ballot; some of our elections have over 60 contests, making it far easier to just mark the “straight ticket” and vote for all of the candidates of one party).

    The optical scan ballots seem to be an almost exclusively American obsession (far better than punch cards with hanging chads). They don’t lend themselves well to STV, though both jurisdictions in the US which use STV do use scan cards. In Cambridge, voters get a matrix of 400 or so bubbles, with candidates in the rows and preferences in the columns, and have to darken one bubble per preference. Are there any jurisdictions in Australia (or Eire or Malta) which use anything other than ‘write one or more numbers by hand’ for STV elections

  7. The ACT has electronic voting for some electors. The AEC and NSW use OCR to scan hand written ballots (there is extensive cross-checking with a physical count) and then the captured data is processed eelctronically to priduce a result.

    There was a famous occasion in NSW when, after the scanning was done, the electoral commissioner pressed the big red button to initiate the electronic count and nothing happened. The program executed correctly on the second try.

  8. Chris@59, thanks, duly noted. I use “tick” as a synecdoche for all methods of indicating simultaneous, equally-weighted support for two or more candidates (or unique support for a single candidate only). Lazy and inaccurate at times, l admit, but “crossing” is ambiguous (in pre-MMP NZ, you struck out the names of all candidates except the one you liked, and this was often described as “crossing off” or “crossing out” the disapproved names), while “darken a circle” or “punch out a chad” is tedious if repeated. But our Grand Wiki will clarify all these provisos, and more…

  9. Elections ACT imposed three conditions of tender: (1) successful tenderers had to be prepared to offer all intellectual property tot he electoral commission; (2) the software could not be platform-dependent; (3) the software and hardware tenders could not be awarded tot he same or related tenderers.

    A number of US election management companies attempted to meet these conditions and could not. Their attempts to have the conditions of tender amended were unsuccessful.

    Elections ACT then published the operating code and invited public to find if there were any ways that the program could execute incorrectly.

    Private administration of elections isnot only dubious in terms of security, transparency and accuracy. You get an inferior service at considerably higher cost.

  10. Back around 1991 or so, Australia Post introduced standard sized and positioned squares on envelopes so that handwritten postcodes (basically, 4-digit zip codes) could be machine-read. I wondered then why that couldn’t be adapted to reading numbers handwritten in squares on ballot-papers, eg, so all the “1234” votes get sorted into one pile, the “1243” votes into another, and so forth. Count the piles, enter the numbers for each permutation into a spreadsheet, know your new Senators in time for the 9 o’clock news. I assume I was being way too optimistic here…?

  11. That’s roughly the process for the Senate and NSW legislative council scrutinies. I don’t know about the South Australian legislative council.

  12. Cambridge MA finishes the count on election day because of the scan cards (excluding ballots with write in candidates,

  13. …which are entered the next day, and postal ballots, which have 10 days to arrive). These hardly ever effect the final count, so most candidates know if they’ve been elected that night or the next morning.

  14. It seems odd that the NSW Legislative Council scrutiny takes 3 weeks when they use optical character recognition. It it because they process all 4 million ballot papers at a central count center?

    The US Postal Service also uses optical character recognition, though without standard positioned squares for zip codes. I would think such a technology would be easily adaptable to handwritten ranked votes. The problem would be making the technology cheap enough for multiple counting machines to be used. Variations of the “darken the circle” machines used in most of the US are cheap enough that many schools have them to make it easier for teachers to grade quizzes and tests.

    In Texas, any voting system (meaning machine/computer) used must be capable of informing a voter if they’ve overvoted or undervoted in any race. I know in Minnesota, they’ve handcounted ranked ballots because the approved systems just tally the number of votes cast per race, and “first choice” and “second choice” are configured as separate races. Therefore the first preference results are known on election night, as are the total number of second preferences received by each candidate, but not which first preferences flowed to which candidate.

    It seems the simplest solution might be to move to postal ballots and put the postal service in charge of running the counts. They already have the character recognition technology in their sorting centers, after all.

  15. I believe the Australian delay is attributed to several things. As you said, the centralized count does slow things down. The very liberal postal voting laws in Australia probably mean that any sort of proportional counting would be difficult to achieve with eligible postal ballots still to arrive. There is also a degree of manual scrutiny, as I understand things, to ensure that the machines are reading what is actually written on the ballots and will count the votes accurately.

    And also, unlike the U.S. and to a lesser degree Canada and the U.K., the various election agencies have little interest or mandate to produce “immediate” results. There is no fundamental right to be entertained by results on election night nor any right for a fast count to be desired over an accurate count.

  16. If the “‘count” is run by telling a computer, eg, “We have 1000 ballot-papers filled out for ABCD, 7000 for ABDC, 300 for ACBD…” etc, presumably one could press a button and generate a provisional result on election night even if there are still votes being flown in from Afghanistan or faxed in from Antarctica. On the other hand, counts have been manual for decades which may had entrenched an understandable reluctance to start too soon if you may have to scrub everything and re-start when the plane from Basra or Australia House, London gets in.

    With IRO-AV, one can at least start and proceed as long as any outstanding latecomer ballots arn’t numerous enough to change the order of elimination. With PR-STV, by contrast, you can’t because even if the order of elimination doesn’t change, the mere fact that late-arriving votes exist may change the size of the quota, the size of a candidate’s surplus, and therefore the transfer value of his or her ballots, and everything that flows from it.

    Australian election nights are usually pretty entertaining, still. Because small rural polling booths phone their results in first, there is invariably an early “swing” pf 20-30% to the National Party around a quarter past six that evening – but political gravity eventually catches up with it. The guest commentators on TV are usually retired (or pushed) former politicians who can sometimes barely conceal their schadenfreude when a victorious rival within their own party leads it to an electoral drubbing. Antony Green, he of ABC website fame, is usually on the ABC.

  17. Reflecting on the “equal first preferences” rule, it actually sounds relatively similar to a “party-preferential” system using OLPR like Dion proposes (“I’ll take any New Democrat over a Liberal, and any Liberal over a Conservative”), though not as sophisticated/complicated (probably for the best, as I think any counting system allowing equal first preferences is bound to have lots of errors in the tabulation). A closed list system (essentially, group voting tickets) also has a similar effect, though without the ability to give members of a party equal preference as Dion’s system would allow.

    Actually, for all intents and purposes, the NSW Legislative Council already is party preferential. Only 2.5% of voters voted below the line, and the vast majority of BTL votes were 1 preferences for the top candidate in a group, and most of those were probably voters ranking the candidates of one party straight down the paper, which is the same as marking only 1 for one party ATL). No ungrouped candidate has ever been elected, nor has any candidate ever upset the party’s ticket order and been elected. It would not have a significant impact to abolish BTL voting and make the ballots follow the same format as the assembly papers.

    Doing so might encourage more voters to transfer preferences (though I understand the plumping rate in assembly races is quite high), and eliminate confusion thinking that one can only mark 1 party above the line, or that a 1 vote fully distributes preferences. It would also reduce informality due to voters screwing up below the line.

    Of course, voters screwing up below the line was probably the only thing that saved NSW and Australia from Pauline Hanson’s triumphant return. Had all the BTL ballots with just a 1 next to her been ruled formal, I think she would have passed a quota. Only she and John Hatton, the other independent with a “Group J” ticket above the line, received any significant portion of their votes below the line.

    There is also the issue of needing to hold a referendum to eliminate below the line voting.

  18. Without really intending to, NSW has ended up with a constitution that is arguably superior to the federal constitution.

    The electoral cycle is 8/4 rather than 6/3. The legislative council has a higher effective number of parties than the senate. Ditto the legislative assembly and the house of representatives. The independence of the courts and monitory institutions like the electoral commission, the ombudsman, the auditor-general and the anticorruption commission is better protected than their federal counterparts and there is no federal anticorruption commission.

    Now if we could just get multimember districts in the assembly…

  19. I would say the equal preferences rule as proposed would not distort the final result (in terms of allocating greater voting power to any faction that voted 1, 1, 1 instead of 1, 2, 3) any more than using divisors 1, 3, 5… in St-Lague makes a difference from using 0.5, 1.5, 2.5. However, in my experience the proposal does tend to frighten the horses when first put on the table. My main reservation is that it would make counting more complex.

    Re “No ungrouped candidate has ever been elected, nor has any candidate ever upset the party’s ticket order and been elected” (Chris @69) –

    1. Someone or other noted that while some Independents had been elected to the Australian Senate a few times from the “Ungrouped” column at the end in the 34 years before ticket-voting was introduced in 1984 (Townley, Sid Negus), none has in the 29 years since. (Fielding, Woods/ Dunn, Xenophon and Chamarette were head of minor-party multi-candidate teams and so “deserved” their own column). The only case where order of election did not follow order on the party ticket was in Tasmania (surprise!) in the 1960s (Senator Archer) and even then it didn’t change which candidates got in, just the order in which they reached the quota.

    2. Don’t forget that below-the-line votes can change the result among parties if the election is close. Antony Green has nominated a few cases. The BTL voters might not succeed in replacing ALP #2 with ALP #3 but they can and do replace him or with with Green #1.

    3. One might say that the same result can be accomplished with preferential voting among tickets. But sometimes minor parties want to pick and choose among candidates of the major parties. Eg, in 2010 the Sex Party (a front for the porn industry) and the Greens put Senator Stephen Conroy last even though he was high on the Labor ticket (which they otherwise favoured), because they opposed his plans for a heavy-handed internet filter. (I have some vague idea that Labor did the same with Neville Bonner, Australia’s first Aboriginal Senator, in preference to other candidates on the Qld Liberal/ National ticket… Alan?). If parties are required to preference an allied party’s candidates in the same order the latter lists them, the former may be more reluctant to preference the latter at all. If they are allowed to vary the order, it raises the question of why the parties can but the voters can’t.

    4. Originally the NSW upper house filled casual vacancies by taking the last-eliminated candidate from the same team as the vacating MLC. This was changed, in either 1984 or 1991 (in conjunction with either the 4-year terms referendum or the 8-year terms one, I forget which) to Senate-style party appointment. One advantage of the “last running-mate out gets the vacancy” rule is that, even with very rigidly stage-managed party tickets deciding who gets in at the general election, voters would retain some control over who fills vacancies. Even if only 5-20% of voters number candidates below the line, this would often be enough to make a real difference in filling a vacancy: ie, it wouldn’t change which of the party’s candidates come first, second and third, but it may make a difference (once the quotas are deducted and the party’s “remainder” is depleted to single-digits percentage) who comes fifth or sixth.

  20. “If parties are required to preference an allied party’s candidates in the same order the latter lists them, the former may be more reluctant to preference the latter at all. If they are allowed to vary the order, it raises the question of why the parties can but the voters can’t.”

    I was suggesting that the voters choose which parties they prefer, and get all those candidates in the order they appear on the ballot paper. This is how ranking multiple parties above the line works in NSW now, and what Bob Brown has been suggesting for the Commonwealth Senate. So parties don’t preference allied parties’ candidates at all, voters do, but both lose the ability to alter that order (as parties had before 2003 and as voters still do). Some voters might dislike losing this ability, and it might make them less likely to rank a second or third party, but as I pointed out, only a very small percentage of the already small percentage of voters who vote below the line choose to do so as it is.

  21. I ran 2 student representative council elections at a Sydney university in the 1980s where the council was wise or foolish enough to let me write my own rules.

    I set up a unified ballot with no division between above-the-line and below-the-line where. Both tickets and individual candidates appeared in random order. This meant that ticket candidates effectively appeared twice on the ballot, so I specified they could only receive a vote once.

    The electorate seemed quite happy with it. The SRC almost had hysterics because they ‘knew’ there had to be a split between ATL and BTL voting.

  22. They use tickets for student elections? That seems bizarre to me. Are they affiliated with political parties, or just groups of like-minded friends?

    Of course, at my university a student was disqualified from the student body president election last year for using photographs of student assembly candidates which “implied they were endorsing him,” which was a violation of the rules. He then proceeded to sue the university, claiming such a restriction was a violation of his free speech rights and that it could “irreparably harm his chances of one day becoming president of the United States.” As one might expect, he was subsequently pilloried in the campus paper.

  23. They are like-minded tickets. Official party-affiliated tickets are pretty much unknown. That you find very few cross-party tickets or that many people go from student politics to ‘real’ politics is purely the operation of chance and circumstance.

  24. The advantage of simply prohibiting reversed pluralities is that you immediately remove a mountain of perverse incentives from politicians about gaining advantages through manipulation of the electoral code.

    Contemporary bills of rights are very explicit about the nature and conduct of elections, e.g. Kenya:

    Article 81 General principles for the electoral system
    The electoral system must comply with the following principles:
    (a) freedom of citizens to exercise their political rights under Article 38;
    (b) not more than two-thirds of the members of elective public bodies must be of the same gender;
    (c) fair representation of persons with disabilities;
    (d) universal suffrage based on the aspiration for fair representation and equality of vote; and
    (e) free and fair elections, which are:
    (i) by secret ballot;
    (ii) free from violence, intimidation, improper influence or corruption;
    (iii) conducted by an independent body;
    (iv) transparent; and
    (v) administered in an impartial, neutral, efficient, accurate and accountable manner.

    It would be a fairly simple step to add that electoral results do not stand when they are reversed pluralities.

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s