New Zealand flag finalist…

The first of the two rounds of deciding on a potential new flag for New Zealand has been completed. The winner is Silver Fern (Black, White, and Blue). It will go head-to-head (hoist-to-hoist?) against the current flag in a second postal ballot next March. The proposed alternative shares with the current flag the stars of the Southern Cross.

This round selected from five alternatives, using a preferential (ranked-choice) ballot. The winning option did not lead on first preferences. That leader was ” Silver Fern (Red, White and Blue)”, which had 41.76%, against 40.16% for the eventual winner. The leader in first preferences was the only one of the five designs without black in its color scheme. Apparently, via the sequential elimination and then transfer of preferences, the consensus was fern and stars with some black.

This first round originally was to be among four alternatives, but a social-media campaign of especially passionate supporters managed to get a fifth design, Red Peak, included. Red Peak, however, did not fare so well, coming in a distant third with 8.69%. Another silver fern option, which was all black and white, and a black-and-white Koru design ended up way back.

The preferences round by round suggest the Red Peak was by far the second choice for Koru supporters. When black and white fern was eliminated, its ballots put the black-white-and-blue fern design in the lead, narrowly, for the first time. This made Red Peak the swing bloc, and those ballots split 54.34%-45.66% for–perhaps a little ironically–the option with no red on it–but with that consensus black. The final count was a vote division of 50.53%-49.47% between the two fern-and-stars design.

The long list from which the five ballot contenders were selected by a panel had 39 options. The panel had 10,292 designs to sift through.

11 thoughts on “New Zealand flag finalist…

  1. Part of me is mildly amused that NZ chose to use AV for a referendum after the Royal Commission that recommended MMP in 1987 explained so carefully why preferential voting systems are bad.[*]

    But then referenda expose the flaws of FPTP very starkly. Being supported by one-third of those casting ballots – which would be a “clear victory” for Tony Blair or David Cameron in an FPTP election – would look rather embarrassingly like the sort of “thumping defeat” that AV underwent in 2011. So you either have to gerrymander the choices on the ballot to restrict them to only two (Australia, constitutional referenda) or where there are three or more, else use preferential voting (NZ: Australia in advisory plebiscites), runoffs (Switzerland, Puerto Rico, NZ again) or a disguised form of approval voting (California).

    [*] For reasons I’ve explained elsewhere, I (like MSS) accept that if you use a compensatory MMP system, you need to use FPTP for the single-seat component.

    • Rather loose with your terms there, Tom. By that terminology, virtually all legislative votes are ‘gerrymandered’ into being option X against the status quo or against the proposal on the floor if X is an amendment to it.

      Yes, I suppose these flag referenda break with NZ’s electoral system referenda in more than one way. Had New Zealanders had a vote in this first round on whether the flag should be changed at all I suspect there would not be a second round.

      For my part, I remain thoroughly unconvinced of the preferability (much less ‘need’) of using plurality for MMP’s district component. I, too, have made my arguments elsewhere, but Tom already makes the first one with his point on the “flaws of FPTP”…

  2. No, JD, in a legislative assembly any individual member who can find a seconder can propose amendments (sometimes there are higher thresholds, but even so 5% or 10% of the legislators will almost always be enough under the Standing Orders to get a motion onto the floor). If one or more counter-motions are moved, the assembly deals with them, usually using some form of truncated Condorcet method.
    Outside the emergency provisions in France’s Gaullist Constitution, it is rare to find a democracy where the executive can tell the Parliament or Congress “This is the only option. Take it or leave it,” at least on any important legislative decision.
    True, there are often exceptions for disallowing regulations and other quasi-executive decisions — eg fast-tracking trade deals, closing military bases. But even here, when a legislature is given an “up or down” choice, it can almost always revisit the matter a short time later. And Parliamentary or Congressional backbenchers can make it clear to the executive and/or the floor membership that they voted NO this time but will consider voting YES if an amended version with changes A, B and C is put.
    This works because, and only because, a legislative assembly is small enough for the members to meet and negotiate personally, and to make handshake deals stick.
    But in referenda, by contrast, voters can be (and often are) presented with a yes or no choice on a single package deal, with the implication that if they vote NO, that’s the matter off the table for another generation. (Europe’s practice of repeatedly calling votes on EU integration until the voters get the answer “right” is one response. Australia tends to take the opposite approach — before the poll, the NO case will urge voters “Even if you support this idea in principle, vote NO if you’re less than 95% happy with the legislative details of this particular proposal, so you’ll get a better deal next time”; but after the poll, they switch to “The People have resoundingly rejected the very idea of any change in that direction.”
    The fact that MMP “needs” FPTP for the district tier (ie, that attempts to replace it with runoffs, AV or – I assume – Approval, would be too open to manipulation) is a mark against it, in your scribe’s humble view.

    • Either way, it has nothing to do with gerrymandering.

      I think it’s very unfair to MMP to hold something against it which isn’t even true. If anything, MMP with AV is less open to manipulation as it’s more tenable to require parties to nominate a certain minimum of district candidates (to combat the decoy list problem), as they won’t split the vote.

      • JD, I’m using “gerrymandering” in its widest sense to mean any deliberate rigging or selection of the detailed rules and procedures of a poll (which voters vote when and how, and among which options) to favour a pre-ordained result.

        True, the Platonic ideal of “the will of the people” cannot be distilled in its pure essence. The best we will ever attain is that “these N individuals voted on this date among these options.” A different set of rules might have produced different results. As long as these rules were selected because they were genuinely and reasonably believed to come the closest to the Platonic ideal – and not to favour a particular side or outcome – they pass muster.

      • I, too, would prefer not to stretch the meaning of “gerrymandering” in this way. Tom is talking about agenda control, which is a fact of life in any collective decision-making process. And, in fact, many countries’ legislative rules do let the executive make take-it-or-leave-it proposals, at least under certain conditions.

    • MSS’s blog, so MSS’s rules. I’ll drop “gerrymander”. But not all agenda control need be agenda manipulation. Ideally, agenda control (structuring?) should be aimed at eliciting some clear outcome (rather than nooise or loops/ cycles) and one that is preferred by the majority over as many alternative outcomes as possible. It becomes manipulation when it is deliberately rigged to favour one result over another.
      I agree that the leader(s) of the majority party or coalition in the legislature (the prime minister in a parliamentary system, the Speaker/ President of the chamber in a presidential system) often has more power to present an option as an up-or-down, unamendable proposal. I still maintain that the backbenchers and/or opposition party(ies) usually have far more opportunity to put a counter-proposal on the agenda to be voted upon – formally or informally (‘we’ll never confirm nominee A, but nominate B and we’ll think about it”) – due to the sheer difference in scale between a legislature with dozens or at most hundreds of members, and a mass electorate with at least dozens of thousands. This applies even when some form of voter-initiated referendum is available and a fortiori when one isn’t.

  3. If its correct that keeping the current flag, either through first a yes-no referendum on whether the flag should be replaced at all, or at least as one of the five in a preferential vote, was not an option in this vote it is a serious flaw. If only due to status quo bias, the current flag would have had considerable support and there was a good chance a majority would have preferred to keep it.

    Personally, I would have voted to replace the flag due to the current flag looking too much like the Australian flag. And maybe this is the sort of technical issue that you don’t want the voters to screw up. But I question having any election without including what is almost certainly the first choice of a considerable number of voters.

    Interestingly, I can’t think of any other instance where a change of national flags was put to the voters. One technical bit of trivia is that the flag of the USA has changed over thirty times, all without any vote at all, though the basic design admittedly changed only once.

    Of the five finalists, I prefer the flag with the fern and without the southern cross. Including the southern cross along with the fern makes the flag look too busy. The red peak design looks too much like something a corporation would clothe its lower level employees in, and the koru has the problem that no one outside New Zealand would have any idea what the design is supposed to represent.

    • Ed, I think (CMIIW) that this round was just a “primary” ballot to select a single challenger which will go up against the status quo in a later up or down referendum. It isn’t final.

      New Zealand used a similar method to select MMP to replace FPTP in 1992-93, except that that time the first round was FPTP instead of AV — not that it mattered because MMP, the version recommend by the Royal Commission, won over 50% of the votes anyway. (Unless this was distorted by tactical voting?).

      Not a bad way to run a multi-option referendum, although it has the flaw of encouraging opponents of change to “raid the primaries” by voting, on the first round, for a change option that is less likely to prevail head to head against the status quo later. But then this could get too clever — if Change Option A wins the “primary” with a large majority on the first round, this doesn’t legally preclude it from still losing the “runoff,” but politically it would probably give the “yes” case a big boost for the campaign, and move the Overton Window closer to the direction of change.

      Possibly one could guard against this sort of tactical primary-raiding by stipulating that if one change option is supported on the first round by at least 50% of the valid ballots and also — say — 33.34% of all persons entitled to vote, it is deemed adopted without a second ballot being needed. If no-one gets that, count preferences to see which change option goes into the playoff.

      (I’ve sometimes thought of a combined preferential/ runoff system for presidential elections… ie eliminate the lowest candidates one by one and distribute their preferences until either one candidate has 50% of the votes and 33.34% of the voters [elected outright on the first ballot], OR only two candidates remain [they go into a runoff]. Like top-two á la California but without the danger of SNTV-style tactical vote-splitting. In normal circumstances most Red voters would number all and only the Red contenders for the presidency, and most Blue voters would number all and only the Blue contenders, but in (say) a wartime situation more voters would take the trouble to number all candidates so the first ballot would be decisive and the time and cost of a runoff could be saved: c/of the Irish political parties’ agreement to avoid contested presidential elections in 1977 and 1984.)

      • “I’ve sometimes thought of a combined preferential/ runoff system for presidential elections” what’s the advantage of that over regular AV?

      • Why not simple AV? Well, first, AV with runoff allows a turnout quorum, which seems to be important to a lot of societies with voluntary voting.

        I have heard some political scientists (and Law colleagues, watching from the field adjoining) praise runoff as superior precisely because it gives a second round of voting and thus more opportunity for deliberation. Voters go to the polls knowing the choice will be a “straight fight” between incumbent President Charles de Valery and opposition challenger Mack Cleighton. They are not surprised to find that their little-thought-about fifth and sixth preferences decide the result. (Nicholas Renton, for example, makes this argument in his manual of meeting procedure http://www.amazon.com/Guide-meetings-organizations-reference-societies/dp/0455137102/ref=asap_bc?ie=UTF8).

        But unlike a single FPTP (or AV) ballot, the field is not legally “crowded” with no-hope candidates: unlike FPTP, the field of politically viable candidates is not arbitrarily reduced by “playing chicken” – “squeezing” tactical votes, candidates doing deals to “stand down” for each other – based on opinion polls that survey only a few thousand lucky voters. Who gets eliminated is decided by an open, transparent, publicly-run ballot in which every citizen can vote.

        Simple AV is my first choice (as it were), but I have to acknowledge that in the Darwinian struggle for survival among electoral systems, Runoff is a clear favourite in the marketplace of ideas. Numerous societies that used to use either FPTP or some kind of electoral college [one more “indirect” than the US version, that is, ie with the electors viewing themselves as trustees rather than delegates] (Argentina, Austria, Brazil, Chile, Finland, France, Indonesia) now choose their presidents via popular elections using runoff: newer democracies with directly elected presidents have chosen it on a blank slate (Egypt, Portugal, Poland, Romania, Russia, Serbia, Timor Leste). Electing Presidents by FPTP (Iceland, Nigeria, South Korea, Taiwan, USA) or AV (Ireland, Sri Lanka) is now quite rare.

        Only the rejection of SNTV and MNTV comes close as a clear-cut verdict in this otherwise endlessly mired area of debate.

        Curiously enough, FPTP supporters who loathe preferential systems (particularly STV/AV) seem to regularly propose some sort of runoff system as an alternative acceptable to them. I saw this numerous times in the online debates over the UK’s AV referendum. “AV is bad because it counts a second or third preference as equal to a first!” “Well, so too do you Conservatives, when electing your party leader, or the Speaker of the House of Commons.” “Yes, but that’s totally different. That’s an exhaustive ballot – nothing like STV/AV at all.”

        From a mathematical point of view, this argument is incoherent, but I suspect it carries considerable political or symbolic weight, for two reasons.

        First, because (as mentioned) an X for your least-hated candidate in a second (or later) ballot is seen as more conscious and deliberative than a 5 or 6 or 7 next to their name in the first ballot, not knowing how the final race will look. (I do think this criticism of STV/AV has the most force when voting tickets are used, where voters can find with surprise after election day that their explicit first preference for Left-wing Major Party A helped elect a Senator from Right-wing Minor Party B via their implied tenth or twentieth preference: with explicit preferences only allowed, especially when these are optional, voters should be much less likely to expressing the “wrong” preferences, although they may regret expressing too few if it turns out that some little known third- or fourth-party candidate ended up surviving to the final bossfight).

        Secondly, and this reason is purely symbolic: Xs on ballots in a runoff ballot look for all intents and purposes indistinguishable regardless of whether they’re cast by enthusiastic first-choice fans, or by very grudging nose-holders (eg French Socialist supporters backing Gaullists against the Front National in 2002 and 2015). I suspect this is why, for all their talk about how second and third preferences shouldn’t be count as 1.000 of a vote, FPTP supporters are more than happy to accept de facto second and third preferences from minor parties by way of tactical voting. (“Polling Shows Lib Dems Cannot Win This Constituency! Vote Conservative To Keep Labour Out!”). As long as it takes the legal form of an X for a major party candidate (in that constituency),

        That was why, for all the Tories’ talk about how preferential and/or proportional voting would enable the BNP and National Front to swing elections, they were always happy to accept votes from BNP and National Front supporters, as long as these were cast for Tory candidates. (We can be fully confident, of course, that the Tories made no concessions whatsoever on policy in the hope of enticing BNP or NF supporters to turn out and vote Conservative, rather than voting BNP/ NF or staying home on election day. That sort of outreach to minor-party supporters never happens under FPTP, of course, as long as such minor parties don’t show up as winning any seats).

        So, in recognition of the popularity of runoff voting – as a tolerable second best for AV supporters and (surprising though it may seem) to FPTP supporters as well, I would propose the best possible version of it, and that would involve using AV to narrow the first-round shortlist rather than using FPTP on the first round (which has the weaknesses demonstrated in the 2002 French presidential election) or on the second round (which has the weaknesses demonstrated in the 1925 Weimar German presidential election https://www.en.wikipedia.org/wiki/German_presidential_election,_1925). Myself, I’d rather the tactical voting be on the second ballot than the first.

        To reiterate, I myself think single-round AV is fine but I recognise that this preference is not universally shared and if the trend in the world is towards some form of second ballot I would prefer it be a better rather than a worse version of second ballot – ie, ranked in descending order as follows:

        [1] Only the final two remaining candidates (chosen by an AV count) on the first ballot may proceed to the second ballot.

        [2] Any candidate who (a) is one of the two highest (non-withdrawing) candidates on the first ballot, and/or (b) polled over 15% on the first ballot, can proceed to the second round, which is decided by plurality.

        [3] Only the two highest (non-withdrawing) candidates on the first ballot may proceed to the second ballot.

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 )

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