Is MMP in Ireland’s future?

The Constitutional Convention of Ireland is considering proposing a new electoral system for parliamentary elections.

The 100-member Convention strongly favors a mixed-member proportional (MMP) system, with 69% preferring it over other options. A “proportional list system”–not clear whether open or closed was specified–wins 29% support, and a paltry 3% would like FPTP. (And, yes, those numbers sum to more than 100.)

The news story does not offer information on preferences for keeping the current system vs. change, either in general or any specific replacement system. It does note that there will be a further round of deliberations next month on the exact model that the Convention will recommend.

Ireland is, of course, the main model we have of Single Transferable Vote (STV). MMP and STV are usually the two models most preferred by reform activists (at least in current FPTP jurisdictions) and by political science expert in electoral systems. It is very interesting to see an Irish process possibly leading to STV vs. MMP as choices for the country.

50 thoughts on “Is MMP in Ireland’s future?

  1. The Irish don’t like getting a choice among different candidates of the same party?!

    Like most other proposals to change Ireland’s electoral system in ways that would exploit the assumed advantages of party lists over STV, this one could – if the political will was there – be introduced by ordinary statute, without needing a referendum to amend the Constitution. Keep STV but bring in party-ranked ballot-papers, as in Australia, with each party’s lead candidate in the constituency at the top with their name in bold type. If elected, they serve as “the Deputy” for Dunkilcarnigon West. The other two, three or four TDs for the constituency serve a similar function to list MPs in NZ and Germany. If you want a higher degree of proportionality, raise the district magnitude.


  2. That would be very said if Ireland abandon STV for an MMP system. Why would Ireland move from MMP to STV especially if the list component is closed? Maybe if it were opened, that might be a consideration.

    Perhaps Ireland should add a list component to the MMP system by using the first preference vote as a vote for a party and require a 3% nationwide threshold for the seats, but then Ireland may end up getting the Australian Capital Territorial D’hondt STV system.

    Maybe the Seanad should not be abolish but reformed so that it is completely elected at the same time as Dail elections, and a different electoral system is used from the first chamber.


  3. Tom: I think the Irish probably DO like getting a choice among different candidates of the same party. Whether their leaders like offering that choice, however, is a different question…

    A simple statute change will not be enough – any change to the electoral system will need a referendum, as STV is codified in the Irish Constitution.


  4. Sorry, I should clarify:

    A legal change to an explicit party list system would indeed require a referendum to amend the Constitution.

    But adopting features of the Australian mainland upper house systems – which keep the shell of STV while functioning 95% [*] the same as preferential list systems – could be done by statute, with the exception of compulsory full preferences. [**]

    [*] I understand there are a few cases where the 2-10% of Australians who vote “below the line” do indeed decide whether Major Party Candidate #3 or Minor Party Candidate #1 takes the last seat out of five or six in a State.

    [**] Which in turn may have a flow-on affect limiting the impact of the other features (party columns, ranking by the party machine, ticket-voting) – especially the last since voting a straight ticket will be enticing to fewer than 90-98% of voters when they can cast a valid vote among individual candidates just by writing a number 1, and can cast an effective vote among all the candidates in a Dail constituency just by writing fifteen to twenty numbers (as opposed to several dozen in an Australian Senate election).


  5. I expect the sixth and last senate place to be decided by very small margins in a number of states. Voting below the line will have more than usual force.


  6. Yes, electoral-system change would require a constitutional amendment, which is why it is a topic of the Constitutional Convention. The Convention has already proposed referenda on various constitutional reform projects, including abolition of the Seanad (senate) and abortion legalization.


  7. Ireland’s Constitutional Convention is a very interesting model of an electoral reform process. It includes 66 randomly selected citizens on the model of a citizens assembly, 29 MPs, 4 Northern Ireland MPs(!), and Chair Tom Arnold; plus an Academic and Legal Support Group. This is a possible process for Canada to follow. In the past, only Quebec has held a process combining a select committee with a citizens group, while Ontario did both in sequence.

    It voted on what to discuss in the second weekend June 8 and 9. The crucial vote was whether to include consideration of an entirely new electoral system: yes, by 59% to 41%.

    The 59% vote is more interesting when you know that the presentation on alternative electoral systems was by Prof David Farrell. Farrell has said “I’ve always been upfront about the fact that my personal preferences are for STV.”

    When considering changes to STV, they decided to consider firstly the size of constituencies, and secondly the number of MPs. The Dail has already recommended scrapping most 3-seaters.


  8. I’m not unfamiliar with the attraction of MMP. I felt it myself when I first started studying electoral systems. It retains single-member electorates! It keeps out parties under five percent!

    At the same time, STV seemed a very unattractive option. In the version I was then familiar with – the Australian Senate model – it seemed to have little to commend it. Voters had to laboriously number as many as twenty or even thirty candidates – every single one, no gaps or repeated numbers – and for what? To anoint the one or two candidates preselected for highest positions on each party ticket. “Proportional representation” inevitably means closed party lists, I reasoned – not an unreasonable assumption, in that pre-Internet age where every single news report on PR focused either on Italy (not mentioning that was an open-list system) or Israel (always mentioning that was a closed-list system). And PR inevitably means elections at large, across an entire State or even country, without any smaller local or regional electoral districts. Given these constraints, why not adopt a system that (a) retains “local MPs” in the only form possible – SMDs, and (b) makes it easy rather than hard to make your choice among closed party lists, ie one tick instead of thirty or forty numerals?.

    I also came across a few other Australians (not Tasmanians, obviously) who had heard about the Bundestag system and would rave about it as the best electoral method ever devised.

    I actually didn’t realise that such a thing as non-closed lists could exist until a Swedish exchange student told me how in Riksdag elections one could cross off names. Then at last I made contact with some Proportional Representation Society of Australia heavies who explained what the Hare-Clark version of STV – then used only in Tasmania, although Malta is a close relative – could offer. Optional preferences, rotated ballots, countback for vacancies, party columns and ballot labels but no ticket-voting.

    As a result, Tasmania (and now the ACT) has one of the very few legislative chambers on the planet where every single Member sits there because voters consciously supported him or her. No safe districts, no safe positions on the party ticket.

    They also challenged my rather lazy assumption that single-member districts are preferable to (manageable) multi-member districts. I can see why FPTP and winner-take-all supporters are so fond of repeating this meme, but I am puzzled why professed supporters of proportional representation would want to contribute to its circulation.


    • To be clear, no specific legal threshold, or any threshold at all, is a defining feature of MMP. Technically, neither are single-seat districts for the nominal tier, although various alternatives do not make much sense (to me).

      As for STV, it took me a while to come around to understanding the Australian terminology. As far as I can tell “Hare-Clark version of STV” means what the rest of the world–or at least the international political science world–would call simply, STV. It is, of course, not helpful that this same world also calls the Australian Senate system simply STV, notwithstanding that the usual use of above-the-line (ticket) voting really renders it not so much different from a closed list.


  9. Just to be argumentative,and with no offense meant:

    1) As far as I know, every system that uses MMP does have some sort of threshold in the 3-5% range. I would be shocked if Ireland doesn’t adopt that. In fact, should Ireland adopt MMP, I would also be shocked if they don’t have single member districts and closed national lists. The cynical part of me thinks that MMP would be designed in whatever benefits the bigger parties: thresholds and single seaters.

    2) I feel I am the only person who doesn’t mind Above the Line voting nor thinks it is incompatible with STV. The problem with the Australian Senate system (as with most elections) in Australia is compulsory voting and compulsory preferencing. Take those away and I would clearly see the advantage in allowing voters to choose to make one mark and direct all preferences with that one mark. I would admit that the NSW system where you can vote for multiple parties above the line probably is better.


  10. I had a somewhat similar intellectual journey to Tom Round, in that MMP was beguiling at first until you got into the details. For me the deal breaker was when I checked German sites after a Bundestag election to look up which members had not been elected from the districts. They weren’t listed, because members not elected from districts under the German system are not elected at all, they are party appointees.

    MMP seems to depend on introducing a number of non-elected appointees into the lower house, to compensate parties whose popular support is greater than the number of districts they could win. I don’t have a huge objection to this, but I don’t see the attraction against using STV or open list PR with smallish multi-member seats, and ensure that every member is elected? The gains in having single member districts (with one third to half the members representing no constituencies at all) vs smallish multi-member districts, and the finer granuality of the proportionality don’t seem to be worth it. And actually it strikes me that you can get to the same route with a bicameral system of one chamber elected by single member plurality, and an appointed second chamber, when there is a convention that the appointments to the second chamber will be made to both boost representation of parties with popular support who could not win many districts, and deny a majority in both chambers to a party that has won a majority in the lower chamber on a lowish nationwide plurality.


    • I would completely reject Ed’s notion that members elected on party lists (closed) are “appointed” instead of elected. Lists are assembled, and voted for, in a context in which it is public knowledge who is on them and in which parties can’t get away with getting a reputation for odious candidates. They get voted in like any nominal-district winner, but via a different process.

      One might not like closed lists, but one should not redefine the election process as not an election (of specific candidates to be legislators) just because one does not like the electoral system.

      Ed’s other points about low-M STV or OLPR are valid ones, though I can still imagine reasons to prefer MMP over the former options. (I am not advocating here for any specific system; all of them have flaws!) I would caution, however, that having two chambers elected by different rules is likely quite different in its ultimate impact on representation and policy than having a single chamber whose members are elected by different rules (i.e. a mixed-member system).


  11. Will Ireland embrace a one vote or two vote MMP system?

    Will it use FPTP in conjunction with a closed party list corrective element like NZ and Germany?

    How about a one vote MMP system using an optional preferential voting system in the SMD and the first preference vote is for a party. Could that work? or the party vote could be determine after the final distribution of preferences?

    Is it possible to invent or does it exist a party list preferential vote system? How does that work in practice? Didn’t Australia used it in Southern Australia?

    Would it be possible to create a party list preferential vote system that is always open? Vote for parties, and within parties as well?


  12. I’ve always thought of a different type of MMP system. The % for the winning party determines the number of seats chosen proportionally and the remaining seats would be in single-member districts.

    For starters, there could be 2 ways to implement this: first, you could allocate the ratio of seats in both tiers according to the % of the winning party/list’s vote in the previous election or second, you could divide the country up into purely single-member districts (respecting any provincial and regional lines of course!) but with 2 additions: all candidates would have to run in 3 districts and the % of list seats is according to the % the winning party gets nationwide.


  13. @ JD, I stand corrected.

    @Derek, I believe that someone proposed something similarish for Canada right after the last federal election.

    Basically the number of ridings won by the party with the most votes would be all the seats that party gets. Everyone else’s seat totals would be adjusted to the smallest degree necessary for plurality.

    If the leading party got 38% of the vote and won 200 ridings, Parliament would increase so that 200 seats is 38% percent of the total (527 seats) If the results ended up being roughly proportional there may only be a few seats. If they were totally nonsensical, Parliament could end up being very large.

    Or, to look at it another way, MMP achieved with SMDs and overhang seats only.


  14. (MSS @9) “To be clear, no specific legal threshold, or any threshold at all, is a defining feature of MMP”

    True. However, until (I think) NZ became the second adopter in 1993 (what about Hungary?), MMP was referred to as “the German system” or “the Bundestag system” and German federal practice was assumed to be normative. The five per cent threshold was often praised in other countries as a good way of preventing Weimar-style instability: eg, I recall National Review in the Wm F Buckley era advocating a similar rule for Italy.

    It was (and is) less widely noticed that MMP means that elections often turn less on how many votes the two largest parties get, but on whether a few thousand voters help the third- or fourth-largest party cross the threshold (Germany) or win a district (NZ), as this can tip the entire balance of the parliament by a substantial number of seats.

    I suspect that the presence of SMDs makes a relatively high threshold (5% in Germany and MMP) under MMP more palatable – small parties still have a chance to keep the district seats they win, and may even be treated as if they had cracked the threshold nationwide. Very few democracies seem to have the stomach for setting high (over 4%) nationwide thresholds if there are no SMDs – if crossing the threshold nationwide is all or nothing. Thus Israel, Holland, Cambodia and South Africa all have very low quotas (under 2%, or even simply one quota) with nationwide allocations. Russia’s 7% and Turkey’s 10% stand out precisely because they seem so merciless to small parties.

    (MSS @9) ”
    As for STV, it took me a while to come around to understanding the Australian terminology. As far as I can tell “Hare-Clark version of STV” means what the rest of the world–or at least the international political science world–would call simply, STV”

    Basically, yes. Outside Australia STV does not mean candidates appearing in party-ranked order on the official ballots. It’s either alphabetical or by lot. As a result, parties’ attempts to enforce a single ticket order are half-hearted and unsuccessful (parties seem more interested in trying to rotate their votes equally to achieve a D’Hondt-like allocation of seats). So when a British or American political scientist advocates or opposes “STV”, they are thinking of a system where party seats totals are broadly proportional within medium-sized constituencies but where the parties cannot ensure in advance that particular candidates are elected.

    Within Tasmania, the two different paths STV has taken – managed tickets vs Hare-Clark – end in opposite directions. Hare-Clark means fewer safe seats and more voter choice among individual candidates than SMDs do. The Senate/ “mainland upper house” system means less.

    Anecdotally, talking about “the Senate electoral system” with the average Australian voter usually elicits an ambiguous response along the lines of: yeah, good that even small parties get some representation and the Government can’t just ram through whatever legislation it wants. On the other hand, bad that parties can give uninspiring candidates safe seats for life (or even appoint replacement Senators no one ever voted for); that tiny minor parties can win seats with 1% of the vote due to the lottery of preference allocations no one knew about when they voted; and that you have to number dozens of candidates or else blindly give a party your proxy.

    As a result, Australian PR advocates have found it very productive to distinguish very sharply the version of STV they favour from the lacklustre version used for the Senate, by pointing out that Hare-Clark keeps the good parts but removes the bad elements noted.


    • Interesting on attitudes towards STV variants, Tom!

      As for Hungary, it is not, and never was, MMP. But the system was indeed adopted before New Zealand adopted (and to the best of my knowledge gave us the name) MMP.


  15. How about the following MMP variant: both constituency and party-list votes are ranked. The constituency contest happens under AV. The party-list vote falls under a threshold (logically at 4% or more, but it could be lower). Parties that don’t meet that threshold in terms of first preferences have their votes redistributed to parties that did.


  16. Actually, the proposal I’m considering is a system where all candidates must run for many district seats and the number of seats awarded proportionally is based on the % of votes the winning party got.

    Say that we have an 100-seat legislature in, say, California, with the following results:

    Republicans got 45%
    Democrats 44%
    Libertarians 6%
    Greens 5%

    But, let’s remember that, before the election, California is divided into 100 districts and parties would usually nominate more than 2 candidates running in its district.

    So, the Republicans got the most votes and 45% of the 100 seats, or 45 seats, will be allocated proportionally (no threshold).

    The remaining 55 seats would be allocated in single-member districts, I’m guessing under some sort of Instant-Runoff-Vote scheme (gotta check that one!)


  17. MSS @19:

    I’d semi-agree that party-list legislators are still “elected” (at least when the lists are published in advance of polling: it’d be more debatable if the names on them were kept officially secret, as in Iraq in 2005, or if it were just very hard administratively to get a list of the names, as I understand anecdotally can be the case in Germany).

    However, “being directly elected” is a question of degree. There are institutional factors that strengthen or weaken how “directly elective” a position is – how much of a representative mandate it has.

    An example par excellence is the term of office. A monarch or life-president who was elected for life is still “elective” but after 25 or 30 years in the palace has a weaker democratic mandate than someone who was elected only 3 or 4 years earlier. Likewise if the field of potential electoral opponents is drastically narrowed by legal qualifications for residency, property, professional qualifications, etc. So if, say, the President gets re-elected every twelve years because only candidates personally worth over $100,000 and supported by 20% of parliamentarians may stand against him or her, and because a challenger legally requires 60% of the popular vote to unseat an incumbent, then he or she is still “elected” but less “directly” than most other presidents in democracies.

    I submit that having several candidates bundled together on a party list also dilutes the degree of “direct election.” Not enough to become “undemocratic,” but arguably undesirable, unless outweighed by countervailing factors (eg, “You cannot expect a largely illiterate electorate to vote individually among candidates for 300 National Assembly seats at large” or “It is better to have the Governor and Lieutenant Governor both elected as running-mates on the same ticket so they don’t spend their term trying to sabotage each other politically”) – just as terms over, say, five years can be justified in special exigencies (“it’s war time and the Luftwaffe is bombing London,” say, or “It’s only the Upper House and you still get to vote for one-quarter of the seats every three years”).

    That this is a demerit (albeit not a fatal one) for an electoral system applies particularly when the names are packaged together in an unalterable order, but also (to a lesser degree) when the order can be varied by the voters but, nonetheless, giving a tick to Nadine Dorries or Olympia Snowe at position #15 or #20 on the party ticket can still end up electing David Cameron or Rick Perry at #1 or #2 on the ticket (or vice versa, depending on your mileage).

    (Even in the US, where the presidential “ticket” is only two names long, many voters nonetheless seem rather startled to find that they voted for FDR or JFK but ended up with Truman or LBJ in the White House after a couple of years. You can test even political scientists by picking a given leap year pre-FDR and asking who was elected Vice-President on that occasion… )

    As for the two divergent variants of STV used in Australia, an analogy might the difference between, say, 10- to 15-seat districts using (a) closed-list PR with Hare Quota, joinder of lists, and largest remainders versus (b) open-list PR with D’Hondt highest averages and no joinder allowed. Even though both variants can be pigeonholed under the category of “party-list PR with no threshold” (which is even more specific than just “Single Transferable Vote”!), they are still going to offer very different incentives for behaviour by voters and parties.


  18. I’ve always wondered what would happen in the U.S. Presidential Primaries if all candidates had to choose their running mate before the primaries. It surely would be an interesting piece of work I assume!


  19. The big drawback with STV is that it becomes increasingly difficult to conduct an election the larger the district magnitude.

    Larger magnitude means longer counts. Australia is the only place I can think of where counting routinely lasts more than a day. It’s pretty tough to convince politicians and the public to wait three weeks for results when they’re accustomed to results the day of or day after.

    The ballot papers also get very large-the 1999 NSW ‘tablecloth’ comes to mind; if you sacrifice party columns, it’s harder to find candidates on the paper.

    I wouldn’t recommend STV for larger than a DM=9. Otherwise, it seems easier just to use lists.


  20. In the NSW legislative council, with magnitude 21, the result for all but the last 1or 2 seats is generally known within 24 hours. I’m, perhaps quixotically, attached to magnitude 21, but the count is not really unknown for 3 weeks. ACT Elections’ largest magnitude is 7 and all seats are usually known within 24 hours.

    It’s almost tempting to suggest US STV advocates thing about trying to promote the NSW model for a state senate. California could have a senate of 42 with half elected by the whole state every 4 years. I think rotation systems are an unnecessary complication. A non-rotative California senate of 41 with magnitude 41 and a four year term could work.

    They could contract ACT Elections for the electronic voting software.


  21. Would it be acceptable if, say, California wanted to allocate its 55 electoral votes in a two-tier fashion? You know, allocating 9 electors statewide and the remaining 46 electors in multimember districts?


  22. The best way to allocate any state’s electors is to the winner of the popular vote. Having one or more states abandon the unit rule in a situation where other states retain the unit rule is just a technique for rigging elections.


  23. STV has led to that distinctively Irish phenomenon of relatively large numbers of independent parliamentarians-one would have to imagine this would be less prevalent in an MMP system, what with the incentive to vote “usefully” in the district neutralising the regionalistic or personalistic appeal of a lor of these independents (“I’ll give him a preference because he’s a lical and will deal with the specific problem we are having”), and any threshold making quasi independents difficult to elect through a national list.


  24. @27, could list systems be modified to provide more nearly equal treatment to candidates and voters who prefer to believe they don’t need political parties? What about putting independents on a separate list (open, irrespective of whether the party lists are open or closed) and using STV to select the winning independents? The number of winners and threshold would be determined by the number of votes cast for all independents. Who wins would be determined by STV. Seems to me this might work.


  25. @28: Sounds a bit complicated-a candidate is a candidate, it doesn’t make a lot of sense to create a parallel system for those who choose not to be a member of a party (or indeed are only quasi-independents-running in an officially non-partisan capacity but in fact associated with micro-parties or interest groups).

    On a general note, I should point out that I’m not advocating the position that every politician be an independent (its arguable that the independents have been blocking political change in Ireland as much as anything by channeling protest votes into an ideological-and government-formation-dead end). Its just that for issues where there is a high degree of local consensus, its one method Irish voters have used to get the attention of a highly centralised state. It might be difficult to persuade voters that the potential benefits of MMP outweigh the flexibility of STV as it has evolved in an Irish context.


  26. Ideally an electoral system does not discriminate between parties, between candidates or between parties as a class and independents as a class.


  27. Ah, but it’s harder than it looks to decide what’s “discrimination” between parties and independent candidates.

    SMP elections may seem fair on this criterion, since they draw no explicit distinction between candidates who belong to parties and those who are independents. But suppose Marge Maverick (Independent), the well known public intellectual, has 25% support across the nation. She’ll end up receiving only a fraction of one percent of the total vote, since only residents of a single district can vote for her, and will likely not win a seat. Parties don’t have this problem, since they have multiple members—so SMP is only friendly to independent candidates with highly localized support.

    In a List-PR election, our candidate at least wins one seat easily, even if the rest of her votes go to waste. Strangely, the best system for this independent might be Above-The-Line STV, since it would let her direct her excess preferences to candidates she approves of.


  28. @30, that’s exactly my premise. STV comes close to this ideal but has other limitations. List systems and MMP overcome those limitations to some (variable) degree but do not provide a level playing field for independents.

    @29, yes it’s a parallel system, one in which the voters determine how many seats are filled are filled by party representatives and how many seats are filled by independents. And, yes, it’s more complicated. I’m sure that when MMP was invented after World War II, it must have seemed complicated in a similar way.

    I don’t think my idea is perfect. I’d like to hear better ones.

    I should add for purposes of disclosure that I am a strong believer in political parties (I am on the State Central Committee of a small one). I think supporters of non-partisan elections and independent candidacies are mistaken. But they are, at least in the U.S., very numerous. Their views should be reflected in the design of the electoral system.


  29. @31

    And yet even the Australian senate, with its listy STV system, regularly includes independents, as does the NSW legislative council. It would be fairly easy for Marge to be elected to the senate and a doddle for her to be elected to the NSW legislative council. How many independents sit in the Bundestag or the the NZ House of Representatives?

    While I agree that there is a magnitude limit in both Hare-Clark and listy STV I also think there is a good case to accommodate Marge (more accurately, to accommodate Marge’s electors) by having one chamber where magnitude is very, very large. Obviously accountability declines as magnitude rises. The AEC argues that the gains in proportionality above M=9 are far outweighed by the loss in accountability. So I’d have a small magnitude chamber and a large magnitude chamber.

    There may also be a case for the small magnitude assembly to have a very small number of national representatives, say 9, who are elected by the whole country.

    Go Marge!


  30. @32: At a local level in Germany (and at Lander level in Bavaria) there are so called “Free Voters”, which are lists of candidates that do not have the status of a political party. This sort of structure might allow for adepts of non-partisanship to organise outside of political parties, without having to weld together two quite distinctive electoral systems.


  31. @34 I’m working on it.

    @35 I do not see the solution in creating a pseudo-party of independents. There are much easier solutions like Hare-Clark which does not discriminate between parties and independents.

    I am not discussing independents only because of my absolutism that elections are about electors not parties. I would also argue quite strongly that systems that favour parties over independents entrench the power of parties in general and those parties capable of forming a government in particular. ‘Throw the bastards out’ gets a lot harder if you are restricted only to a choice between bastardries.


  32. I expected this thread to be popular, but it has gone beyond my expectations!

    I think Bob’s idea (#28) is one of the most interesting innovations in electoral-system design that I have seen in a long time. I have not thought it through very much yet, but it should not be dismissed (for the reasons he gives at #32, and notwithstanding DC’s good points at #29).

    I don’t think, by the way, that Bob’s idea should be considered “parallel” in the way that term is used for MMM systems. As best I can tell, it’s list PR, except that the pooled votes for independents affect only the total number of independents, rather than the order of their election. In this way, it seems it would get around the problem Alan notes at #36, the creation of a “pseudo-party of independents”. Or at least it does so to a significant degree.

    Maybe STV is still better. I do not know.

    As for the question of how many independents there are in NZ under MMP, it depends on how strictly one defines “independent”. In recent elections, there have been a few “parties” that actually elect only their one candidate who is capable of winning a single-seat contest. They register as a “party” so as to attempt to win additional seats via the list. But that’s pretty close to being an independent, de facto.


  33. On MMP and thresholds (raised in comments 9, 10, 12, and 16), it is correct that Lesotho has no threshold despite using nationwide compensation. The MMP systems where compensation is carried out in districts rather in the full jurisdiction represented by the assembly likewise tend not to use legal thresholds. This would include Bolivia, Venezuela, and Scotland. Of course, all of these have many relatively small compensation regions, and hence significant “effective” thresholds. (There is a small national compensation district in Venezuela, with a low threshold, but the main action takes place at the state level.)

    I agree this is likely moot in Ireland, as I’d expect a threshold. That is not based on any actual information of what is being discussed. It just seems highly likely.


  34. Regarding single-seat NZ Parties, I’m fairly sure that Party Leaders also get additional salary and Parliamentary expenses compared to independents, and you get additional allowed electoral expenses. You do need 500 financial members.
    So there are strong inventives to form a party in NZ.


  35. @37 Yes, in retrospect, the use of the term “parallel” in this context is confusing. And MSS is right that “the pooled votes for independents affect only the total number of independents, rather than the order of their election”.

    As far as I can tell, my proposal for STV within the list of independents could be incorporated into either party list or MMP systems. In MMP, independents with a local base of support could run for district seats instead of list seats.

    The advantage that I see over STV by itself is that reduces the length of the list of candidates each voter has to consider. One disadvantage is that it doesn’t accommodate the voter who wants to support an independent as her first choice and partisan candidates as second and third choices. I’m not sure how important that is in practice.

    Finally, if I were implementing this, I would want to give each party the option to use STV within its own list. It’s possible that this would reduce the formation of splinter parties in response to very low thresholds. It would make more sense (to me at least) for such folks to operate as organized tendencies within a larger party.


  36. Before trying to merge STV and MMP it really behooves people to consider the unfortunate ACT experiment of trying to merge STV and ListPR. Imagine you are the electoral commissioner under this system. Then draft voting instructions that invite the electors to vote MMP for parties and STV for independents.


  37. To judge by the ACT experience you would certainly get people attempting to cast transferable votes for parties and nontransferable votes for independents. Ditto nontransferable votes for parties that make their lists STV and transferable votes for parties that don’t. Drafting the legislation to guide electoral commissioners and judges on how to resolve those votes is going to be an equally interesting process.

    If the object is transparency, as the call for shorter lists implies, it is hard to see how a system could be designed that was less transparent.


  38. I would never dignify Pauline Hanson as a public intellectual of the calibre of Marge Maverick.

    Indeed I’m mindful of Bob Katter, asked about similarities between Hanson’s movement and his own, saying: ‘Pauline was a fish and chip shop owner’ and contrasted that to his own record as businessman, state minister, cattle grower, etc etc.

    Marge’s distinguished record in the life of the nation does not necessarily guarantee that she has the skills to operate a party structure, as Pauline clearly did not.

    A system that mandates independents running as parties is just going to subject Marge to the same organisational requirements and possible criminal convictions that Pauline was unable to surmount. It is, in other words, an informal threshold to Marge’s election and a limit to the choices available to the electorate.


  39. Alan, do you have a link to where the AEC argues for 9 as maximum district magnitude? I could really use that in a discussion I’m having the day after next.


  40. I thought Marge was real… That Bob Katter bloke is obviously fictitious, though (“… cross Charlie Wilson with The Man From Ironbark. And have him punch a jornalist.” “Right away, Mr Murdoch, sir!”).

    As for Mr Dunne, today this text has been fulfilled in your hearing.


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 )

Connecting to %s

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