Trudeau, get going!

An analysis in today’s CBC says that if Canada’s new Prime Minister, Justin Trudeau, is serious about the party’s platform commitment that this year’s election would be Canada’s last under first past the post, the reform plan needs to “get going”.

43 thoughts on “Trudeau, get going!

  1. I thought that was a fairly silly article. Mostly it was was an exercise in the urgent need to use the word ‘complex’ at least 17 time and to warn of things that go bump in the electoral dark at least 23 times in any item that discusses electoral reform. AV would not require a redistribution or complex legislation.

    • Alan understates the complexity of actually advancing a reform process, especially if they want (for whatever reason) to include PR options rather than just go for AV. Moreover, if the government is not serious about advancing a process, it has every reason to make it even more complex.

      Sure, they could just adopt AV pretty easily, using their majority, and keeping the existing ridings. I don’t think anyone expects that to happen–even if AV is their sincere preference (so to speak).

      • The article does actually claim that adopting AV would necessitate new ridings. Really it strikes me as the opening shot in the kind of FUD campaign that unfortunately dominates constitutional referendum politics. Tom has given some examples.

        In the 1977 referendum on simultaneous elections for the Australian senate the No campaign began, with admirable restraint, by claiming the amendment provided for simultaneous dissolutions, not simultaneous elections, of the senate. By the second week of the referendum campaign they were claiming that simultaneous dissolution would enable the government never to hold senate elections and to abolish the senate by the backdoor.

        Possibly referendums need machinery like the Irish referendum commission which can give definitive answers on the legal effects of particular questions. Electoral reform referendums definitely do not need supermajority thresholds like the BC STV referendums. Logically, an entrenching measure like the BC 60% threshold, should only be enacted by the same supermajoirty that they require.

      • The polling swings of the last six months have clearly shown that Canadian voters want more than two choices, if they can vote for what they want. This leaves a vast number of voters, frustrated at having to vote against something on October 19, saying “never again — next time I want to vote for my first choice.”

        Polls taken the day before the election correctly predicted the outcome, but until that Sunday, polls had shown a Liberal minority government. Some newly elected Liberal MPs had been encouraging NDP supporters to help elect a minority Liberal government. They were apologetic: “I was expecting a minority government.” Many people echoed a Globe and Mail columnist who wrote “What have we done?” A voter’s anguished letter to Justin Trudeau went viral: “I did not vote for you. I voted against the alternative. . . . Change the electoral system.”

        So the first effect of using proportional representation would have been to end “strategic” (negative) voting.

        The Liberal platform promised to “Make Every Vote Count:” to study electoral reform, including both options: proportional representation, and ranked ballots in single-member districts, and implement one of them in time for the next election.

        Fair Vote Canada welcomed Justin Trudeau’s promise to make every vote count. “Never again should we face a one-party, one-man government elected by a minority of voters. We urge you to work with all Parties and enact voting rules for a true and modern representative democracy in time for the next election. . . We are calling on you to design a voting system for Canada in which every ballot delivers equal representation.”

        Preferential ballots, of course, are intended to help centrist parties, who hope to be everyone else’s second choice. But in Atlantic Canada, where the Liberals swept all 32 seats, the preferential ballot would be a cure without a disease.

        Stephane Dion’s 2013 line was “Preferential voting . . . does nothing to correct the distortion between votes and seats and the under-representation of national parties compared to regional ones.” This has never been more obvious. Surely very few in the Liberal Party will expect such a partisan quick fix to be acceptable. Only proportional representation will make every vote count.

        The preferential ballot would not have helped western Liberals. Only two ridings in Saskatchewan elected Conservative MPs by less than 50% of the vote, and the NDP was second in both. In Alberta I see only one riding where it might have helped Liberals.

        Out of all Liberal voters across Canada, the 20% in the GTA elected 26% of the Liberal caucus, while the 11% in Atlantic Canada elected 17% of the Liberal caucus. Western Liberals cast 25% of the Liberal votes but elected only 16% of the caucus. Liberal women are especially under-represented in the West. Since Justin Trudeau will make his first cabinet half women, gender parity has arrived in the federal Liberal Party.

        In Alberta, the 25% of the voters who voted Liberal deserved to elect eight of its 34 MPs, yet elected only four, all men. In Saskatchewan, the 24% of voters who voted Liberal deserved to elect three of its 14 MPs, yet elected only Ralph Goodale. In the nine ridings of the BC Interior and North, Liberal voters cast 30% of the votes, but elected only Stephen Fuhr, while the Conservatives elected five MPs. In the seven ridings of Vancouver Island, Liberal voters cast 21% of the votes, but elected no one, while NDP voters cast 33% of the ballots yet elected six of the seven MPs.

        As for the referendum defeats of MMP, the Law Commission of Canada model (top-up MPs elected from regional open lists) has never faced a referendum; the 2007 Ontario model and the PEI model both had closed province-wide lists, rejected soundly outside Toronto and Charlottetown. And BC-STV was soundly rejected in the second referendum when voters saw the map with DM 4.25, too large for local representation, too small for good proportionality. Polls have shown for years that 70% or more of Canadians want proportional representation.

  2. I think a priority for opponents of FPTP (ie, supporters of proportional and/or preferential voting) is to work on arguments why such an important change should not be put to a referendum before being introduced.
    Superficially, the argument seems attractive. Even if the electoral system is not entrenched in the Constitution and even if the Constitution doesn’t require a referendum for amendments, democratic theory would seem to demand a direct vote for two reasons:
    1. It’s an important change, and
    2. It affects the basic rules of the game.
    Against that, however, is the point — revealed by the Canadian and British referenda — that a new electoral system will often seem extremely complex to voters who aren’t familiar with it. This gives an unfair advantage to the No case, who can simply stir up confusion — or even lie outright, like the British NO2AV campaign (“Australia needs to use voting machines!” “AV means BNP supporters will get six times as many votes as everyone else!”) — and win by default.
    In fact, the very fact that many voters didn’t want a referendum on something that thought was confusing was easily elided into an argument for voting “No.” The side that pulled off “This AV referendum is costing us X million pounds, so vote NO!” has no standing to then demand a referendum: they’ve painted themselves into a corner.
    There were no referenda before the Great Reform Act, or women’s suffrage, or the Parliament Act 1911: want to argue those were illegitimate? I’ve got no objection to a subsequent referendum two or three terms down the track to confirm whether to keep the new system. [*] the experiences of Ireland, New Zealand and the ACT shows that once voters become familiar with preferential and/or proportional voting systems, and can evaluate them on a level playing field, then in almost all cases preferential voting systems beat non-preferential systems and PR systems beat winner-take-all systems. (Unless you’ve got Tammany Hall or Jack Lang’s Labor machine backing reversion to SMDs, like New York and New South Wales in the 1930s, which — no surprises — meant STV-PR ended up sleeping with the fishes).
    Moreover, so far as importance is concerned, voters are probably even more concerned about the tax system than the electoral system, yet the British Tories in 1988, the Canadian Tories in 1993 and the Australian Tories in 1999 introduced controversial new taxes without a referendum. A plurality-elected majority in the legislature was deemed sufficient. No super-majorities of popular votes, or spread of popular votes across a super-majority of electoral districts, was considered necessary.
    ============================================

    [*] My preferred ideal Constitution would not require a referendum before every new change, as Australia’s or Switzerland’s does, but would make it easy for 5% to demand a referendum to block changes that are actually controversial. Saying “Sign this petition to demand a $50 million referendum to veto the Government’s plans to let dual citizens sit in Parliament” or “… to move the Senate terms changeover date from 1 July to 1 August” or “… to allow direct federal block grants to local councils” is a much harder sell for the snake-oil mongers than “The government is holding a constitutionally-mandated referendum to let dual citizens sit in Parliament/ to move the Senate terms changeover date from 1 July to 1 August/ to allow direct federal block grants to local councils… there’s a hidden agenda buried there and, if you don’t feel you understand these changes completely, you should not merely abstain but should vote NO!”

    • Personally, I think a referendum is a bad option, but it is better than letting Parliament decide, especially in cases like Canada where there was no specific proposal from the current government at the last election. It’s less important with a switch to AV, because that won’t mean an especially dramatic change in the way government is run.

      The comparison with tax is an interesting one, but I think it’s unfair. A tax cannot really be introduced for political advantage, (in terms of disadvantaging the opposition), while an electoral system can be introduced that disadvantages the opposition and allows the government to stay in power (France 1986, for example). I hardly think, too, that single-member plurality is a democratic injustice on the scale of male-only suffrage or the House of Lords having absolute veto over legislation; it’s not good, but it theoretically treats voters and parties fairly.

      Ireland’s referendum was hardly a ‘level playing field’, given that STV had been used for 40 years at the time of the first referendum. I see large problems with the argument that ‘Well, if they liked it in the ACT and in NZ, they’ll obviously like it here, so why bother asking them?’; indeed, they provide evidence that voters are willing to vote for change.

      • 1. The British Columbian referendum of 2005 would count as a landslide victory for proportional representation anywhere except British Columbia.

        2. MMP was rejected by the BC citizens assembly and largish popular majorities in Ontario and Prince Edward Island.

        Naturally this unhappy history is leading some advocates of proportional representation towards demanding an MMP referendum as the only acceptable acid test for the sincerity of the Trudeau government on electoral reform. I look forward to chatting with everyone here about Canadian electoral reform sometime in 2115.

      • I don’t think that proportional representation would succeed in a referendum. As has been noted before on this site, people only seem interested in PR after SMP has massively failed. It helped the STV campaign in British Columbia in the 2005 referendum that the last two SMP elections had been first a case of a ‘wrong winner’ , and then the election of a two-seat opposition. The same goes for New Zealand. The 2015 election was hardly a massive failure of SMP; indeed, it performed what is generally expected.

  3. It is refreshing to see the nod to political scientists who “seem keenest on the mixed member proportional (MMP) system”! If the reform process does go forward, the balance of input from political scientists and other experts vs. input from partisan and local interest groups will be interesting to watch.

  4. As Wilf notes, the Law Commission of Canada came up with an MMP model years ago, and it was a well thought-out proposal. (Full disclosure: they draw on my work quite a bit, so I could be a little biased.)

    As much as I would like to believe that district magnitude and list type were critical for various referendum defeats, I can’t buy it. As Henry says, you need a pretty massive public sense that the current system has failed before you will get a majority for something few voters understand–a new electoral system. I don’t believe there is much evidence that voters approach these referenda with a lot of thought about detailed provisions, but just whether “change” is needed or not. In BC’s first vote, which got 57% in favor (but need 60%) the research is pretty clear that a large block of voters did not even know it was going to be on the ballot. But the electorate was in a feisty mood, and prior knowledge was correlated with voting yes. By the time of the second vote, on the same proposal (albeit with more detail, as Wilf noted), the FPTP system had gone back to “normal”–a fact I noted on this blog before the second referendum took place.

    In my own research, where I try to develop objective criteria for systems in “failure”, PEI and Ontario were not clear-cut cases. That the referenda failed–badly–was less than surprising ex ante. New Brunswick, where the objective case was strong, never got to a vote–a new government canceled the referendum. That it was a government elected in a plurality reversal only makes it ironic. And a little sad. But they got away with it. So maybe the objective basis for the reform process was not so strong, speaking now ex post.

    One final response:
    “There were no referenda before the Great Reform Act, or women’s suffrage, or the Parliament Act 1911.” I suppose democracy has changed, maybe even advanced, since then. I still maintain (but can’t prove) that it would be highly “illegitimate”–a word I will admit I can’t really define–for a government in an FPTP jurisdiction to implement a major electoral reform in the present era without an endorsement via a referendum.

    • Denis Pilon notes:

      With the benefit of hindsight one can trace how the provincial governing elites came up with numerous innovative ways to structure the reform process to create barriers against change, while the conventional political elites modeled how to trump such populist reform initiatives without appearing self-interested. Ultimately, over the course of the four referenda – from British Columbia to Prince Edward Island to Ontario and back – a process of policy-learning appeared to be going on as the various provincial governing and political elites added new twists picked up from each preceding campaign to help ensure defeat of the referenda.

      One difference between the two BC referendums was that the small group of political apparatchiks who controlled the No campaign were granted half a million dollars (the same as the Yes case) to prototype the sort of facepalmable nonsense that No2AV would use in the UK referendum.

    • Why was Proportional Representation embraced in the Scandinavian countries, the Netherlands, Belgium, Switzerland, and others?

    • An seemingly little-appreciated factor contributing to the massive disparity between the 2005 and 2009 BC Referendum results was the question wording:

      2005: Should British Columbia change to the BC-STV electoral system as recommended by the Citizens’ Assembly on Electoral Reform?
      2009: Which electoral system should British Columbia use to elect members to the provincial Legislative Assembly?

      • It really is remarkable that the 2009 referendum was influenced by district magnitude and the form of the question but not by:

        1. an ineffectual public information campaign where the material distributed by the referendum information office resembled regular marketing inserts in newspapers and appears to have reached almost no-one

        2. the opposition by the BC liberals, who wanted to retain FPTP, the NDP and the Greens who appear to have been equally divided between retaining FPTP and opposing STV because it was not MMP

        3. a deeply undemocratic supermajority requirement that made negative votes worth more than positive votes

        5. submerging the reform issue into a regular election campaign

        4. a FUD campaign run by and for political insiders that raised tiny amounts of money in 2005 but was granted CA$500 000 in 2009

        Dennis Pilon notes:

        A post election assessment by province’s chief electoral officer noted how declined and rejected ballot totals for the referendum were significantly higher than the general election, a result he suggested “may reflect a general discomfort among the electorate with their level of understanding of the new electoral process.” The report also cited a Nordic Research survey conducted after the referendum that found over half of those that voted against STV claimed they did so because they did not feel “knowledgeable” (Elections BC 2005b; 38).

        FUD works. One Tony Abbot famously asked people in the Australian republic referendum too one NO if you they did understand everything in the republican model.

        Electoral reformers who persuade themselves MMP is the only possible option, like electoral reformers who believe any particular system is the only possible option, do no service to the cause of reform.

        Personally I agree that AV alone will not establish proportional representation. AV is not proportional representation because it is in fact, not proportional representation. I generally find it a bad idea to try and make toast in my coffee machine, but I do not denounce the coffee machine for that reason.

        The question is whether AV is better than FPTP.

        Canada would benefit from AV, and benefit much more from any one of MMP, STV or List PR. A position that says MMP or nothing is a position that objectively supports retaining FPTP. Possibly Tom’s idea of an act that provides for some form of PR followed by a review referendum after 1 or 2 elections is a better plan. Ideally Canada would emulate the NZ process of having two referendums, one to adopt a particular model and one to decide between the model and the existing system.

      • I don’t agree, Alan. If the Liberals were to have a referendum on, say, AV, that would be a campaign in which the incumbent government were in favour of the change. This would give them a powerful platform compared to the UK, where the pro-AV campaign was led by the unpopular Nick Clegg and Ed Miliband (but not most of his party), and was completely hopeless. Goverment support for change would reduce the advantage offered by the lack of usage of AV.

        Certainly, from your point of view a switch to PR or AV would be ‘good’ for Canada, but this is hardly an objective truth; not everyone wants proportionality. I think that a switch to PR would be such a dramatic change in the way Canada is governed that, despite the fact that a referendum (even one where the government is backing change) often leads to a loss, a referendum is the only way to ensure these issues are properly debated.

      • Henry, where have I ever argued against a referendum? I have criticised the lack of public information and the supermajority requirement in a particular referendum. I have also questioned rather unlikely explanations for the 2009 result.

        A change to PR certainly should be by referendum. I am not 100% convinced that AV is a large enough change to require a referendum, but I’m open to argument either way. I think Tom’s proposal has some merit but only for dealing with AV.

        Just to make things absolutely clear, I would prefer PR over AV and I would prefer AV over FPTP. I would prefer STVPR to ListPR, and ListPR to MMP, but I would prefer any PR model that was on offer to FPTP.

    • “Wilfred Day, a lawyer who is a member of the national council of Fair Vote Canada, which promotes electoral reform, says the Liberals won a majority on a promise to make every vote count and that means they already have a mandate to bring in proportional representation”

      Wilf, the Liberals made no specific promise and did not win a majority of the vote. They have as much of a mandate to introduce AV as they do to introduce proportional representation. “Make every vote count” is, if it was what the Liberals promised, deliberately vague. No practical electoral system can actually make every vote count.

      • I also said “Since 63% of voters voted for parties that promised to make every vote count, this House of Commons already has a mandate to bring in proportional representation.”

        As for the Liberals, their platform item was headed “We will make every vote count.” Most people read that as meaning proportional representation. The ranked ballot in single-member districts alone still leaves half the ballots in the trash. (And it was also the campaign slogan of Fair Vote Canada. Cheeky of the Liberals to steal our slogan if they didn’t mean it, eh?)

      • What is the evidence for the proposition that most people read that “We will make every vote count.” to mean proportional representation? From a quick review of some of some comment threads, Canadian electors just do not have that precise an idea of how electoral systems work. I’d prefer proportional representation but that is not authority to rewrite the Liberal mandate for them.

      • All during the election campaign, all over Canada, I kept seeing comments from the public, and even from many media commentators, referring to the Liberals as having promised PR. It was a very common understanding, or misunderstanding. And of all the Liberal candidates who commented or answered questions, only one (a man in London Ontario) said he supported the ranked ballot but not PR.

        A poll taken after the election showed just over 25% of those who voted Liberal would have preferred a minority Liberal government.

        That should be no surprise. The E-day Liberal vote of 39.5% comprised the 25.7% they had pre-writ, the 12.2% picked up from the NDP support decline, and 1.6% from the Green decline. That means 35% of the Liberal E-day voters had decided during the campaign, many near the very end, that this was how to stop Harper. Some of them were originally Liberals who had decided, back before the first debate, to cast a strategic vote for the NDP, and then came home to the Liberals. But most were new strategic voters.

        The day before the election, pollsters were still expecting a minority government. So were many Liberal candidates. So most new Liberal MPs know how they won.

      • And yet as far away as Australia somehow journalists knew that Trudeau supported AV, not PR. MSS has noted previously in this thread that electors do not vote on the finer points of electoral systems like district magnitude. Slicing and dicing the Liberals’ support does not prove that they must introduce PR any more than claiming that everyone understood or misunderstood what was promised.

        The full plank included, at para 3: ‘We will convene an all-party Parliamentary committee to review a wide variety of reforms, such as ranked ballots, proportional representation, mandatory voting, and online voting.’

        I just don’t think you can treat a somewhat original of Para 1 to be as fixed in stone as the laws of the Medes and Persians but conveniently overlook Para 3 which is only 2 lines and 19 words away.

      • Many Canadians who voted Liberal want assurance that never again will a divisive political leader win 100% of the power with 39% of the votes. And never again will voters have to vote against something, or vote for a less-preferred candidate to block the election of one even less preferred. The system of single-member districts using “ranked” or “preferential” ballots (Instant Runoff Voting or the Alternative Vote) still means voters are voting against something, voting for a less-preferred candidate to block the election of one even less preferred. It is still a winner-take-all system

        And as Eric Grenier’s really illuminating piece showed, it advantages the Liberal Party in a big way. It has been called “First-Past-The-Post on steroids.” It is obviously a scheme for partisan advantage. It is disrespectful to voters’ genuine preferences. The 25% or 35% of those who voted Liberal, who voted to stop Harper, would regard it as a betrayal of the Liberal promise.

        As Stephane Dion says, the preferential ballot in single-member districts alone ”does nothing to correct the distortion between votes and seats and the under-representation of national parties compared to regional ones.” It makes only half the votes count.

        I cannot believe most Liberal MPs will support it. Furthermore, they are well aware how they won. They would not dare get caught in public supporting a scheme for partisan advantage.

      • Stephane Dion wrote:

        Preferential voting is a step in the right direction. However, it does nothing to correct the distortion between votes and seats and the under-representation of national parties compared to regional ones. Other changes are needed to end a voting system that best ts the Canadian context.

        The step in the right direction is explicitly listed in the Liberal platform. Obviously there are reasons, which most of us would advance, to prefer proportional representation, in any of it’s valid forms, to a mere step in the right direction. Wally there are valid reasons to a argue that AV may be unfair in one country in a way that does not apply in another country.

        Those are completely different arguments from imagining that the Liberals are subject to a binding mandate in favour of only one choice in a list of four in the text of their platform. It is even more imaginative to argue that by a step in the right direction Stephane Dion meant a step in the wrong direction or that the Liberals have a binding mandate against something they explicitly listed in their platform.

  5. I think we’re getting a bit ahead of ourselves in this discussion. Please don’t misunderstand me: I would agree that the latest developments seemingly give credibility to the notion that the Liberals’ plan is to 1) switch to AV, mainly because they believe that system will work best to their advantage; and 2) skip a referendum on the new electoral system, ostensibly to prevent its popular rejection by a combination of supporters of the FPTP status quo and proponents of PR unwilling to line up behind AV as a “step in the right direction.” That said, if the Liberals’ main motivation is to introduce a system which will work best to their perceived advantage (and that’s one very big if), one must also admit the possibility that a year-and-a-half from now what constitutes “perceived advantage” to them may be very different from today.

    For example, if Justin Trudeau and his party were to sink in the polls over the course of the next eighteen months, to the point that not even AV could possibly save them from certain defeat, but a switch to PR could potentially result in a hung Parliament, might they not reconsider their options if their actual motivation behind changing the system is merely the party’s perceived advantage (as was the case with the French Socialists and the short-lived switch to PR there in 1985-86)?

    As for the Liberals’ platform on electoral reform, I simply don’t see any exclusive commitment to PR there, although I can see how some might want to interpret “make every vote count” as such.

    • The advantages to the Liberals of AV are so great that it would have to be a very dramatic drop in support indeed for proportional representation. While it could happen, I view it as highly unlikely. It would have to be quite a significant event that got NDP preferences to flow to the Conservatives and vice versa.

      • I am somewhat alarmed by the crude cui bono analysis being applied to electoral reform. Politicians, and their advisers, rarely understand electoral systems and often make serious errors when they attempt to game the system. The NSW Right, a Labor faction that prides itself on knowing everything there is to know about gaming STV electoral systems, has consistently shown titled as hopeless as the rest of us at predicting both the future and STV. I will give you some examples. Bear in mind I am speaking about professional operators who have been voting preferentially for their entire lives.

        1. Between 2007 and 2011 the NSW Right deposed the premier not once, but twice. This led to the unhappy experience of a negative swing of 13.43%, the loss of 20 seats in a house of 91 and the loss of government. This impending catastrophic defeat was obvious was obvious in the polls onwards. That did not prevent the vat creatures of Sussex St applying the same highly effective strategy of leadership change in 2010 with almost identical results at the federal level.

        2. Some years ago a cousin of mine was dropped from the winnable third spot on a Labor municipal ticket to the unwinnable fourth spot. (His ward had 4 seats and even in inner Sydney Labor cannot take over 80% of the popular vote) He approached me and asked how he could get re-elected. I gave him the sad news that he would lose the election unless he could take first preferences from the Labor candidates above him and that he would probably be expelled if he campaigned for that to happen. He nevertheless persuaded himself that he could be re-elected from the fourth spot while asking his supporters to give the top 3 candidates his top 3 preferences. I mention my cousin has a degree in mathematics. I do not need to tell you the election outcome.

        3. When the Turnbull government briefly mentioned the reviving the recommendations of the Joint Standing Committee pm Electoral Matters Senator Sam Dastyari, former general secretary of the NSW Branch and former leader of the NSW Right, declared it an attempt to stack the senate. He was unable to produce any analysis that supported this unhappy conclusion and simply ignored the psephological consensus that rapidly emerged against his alarmed outcry.

        I have no doubt politicians are into cui bono up to the armpits, in Canada as in Australia. I am by no means convinced that they know enough about electoral systems to be able to make their nefarious schemes effective. Nor do I think that electoral reformers should allow themselves to be drawn into this analysis and ignore the significantly more important question of quod pro bono publico.

      • I am not at all sure that AV will advantage the Liberals as greatly as most people think.

        That argument is based on people voting the same way under FPTP and AV which, in Canada of all places, is simply extraordinarily unlikely. It is not tenable to argue that the Liberal electorate includes many voters who were forced there by the electoral system, and then argue those people will vote the same way if the electoral system does not force them there. The first and most expected result of adopting AV would that the Liberal primary vote would decline considerably to reflect NDP and Green voters who would be free for the first time in their lives, to vote a sincere first preference.

        The advantage argument also rests on the validity of polls which have asked Canadians to express a second preference. There are good reasons to believe those polls are meaningless. Australian pollsters must not only produce a raw prediction of the primary vote, they must also predict the distribution of preferences. Two methods exist for predicting preference flows 1. last election and 2. respondent answers. Intuitively, one would expect respondent answers to be the most accurate predictor. They are not. For whatever reasons, distributing preferences according to the preference flows at the last election has a vastly more accurate record. The reason seems to be, although there is no definitive work on the subject, simply that ‘Who would you vote for?’ is a much easier question than ‘If you voted for X, who would you vote your second preference for?’.

        If Australians, who have been casting preferential votes all their lives, are unable to accurately say what second preference they will vote, how could Canadians, with no experience of the system, do better?

        The open and shut case advanced in this thread is simply not as open or as shut as has been claimed. Nor, as I have already argued elsewhere, is cui bono the best test of any electoral reform.

      • Alan, I don’t think I fully understand what you’re trying to convey with your comments on cui bono. Are you saying it’s futile to predict the course of change based on politicians’ self-interest? Or that we should not focus on that? Or simply that what truly matters (I thought it went without saying) is the bono publico?

      • jd

        I guess I am arguing all three.

        An ambition to game the system does not necessarily invalidate anything done for those reasons.

        Proportional representation was introduced to the Australian senate purely because the ALP believed, correctly, they were about to lose government and they wished to retain some influence in the senate. It was partisan advantage pure and simple. Does it therefore follow that Australia should abolish proportional representation in the senate? Proportional representation was often introduced in Europe as a way to contain a surge by social democratic or labor parties. Does that mean the Netherlands or Sweden should abolish proportional representation? For that matter, New Zealand, on one argument, had a referendum on proportional representation only because a major party over-promised in an election campaign. (Australia had a republic referendum for the same reason.) Does that mean New Zealand should abandon proportional representation?

        I accept that the Canadian Liberals probably think they will gain most from AV. That does not necessarily mean they are right and I would argue that those who gained most would be Canadian electors who would be free to vote a sincere first preference. I have tried to show why I think that any advantage the Liberals may gain are grossly overstated.

        While I agree with Stephane Dion that AV is no more than a step in the right direction I think we should always take steps in the right direction when they are on offer.

        I would hope that the position of indigenous Canadians is a central part of any debate on electoral reform. The model advanced by the Canadian law commission would merge the three territories into a single common district to elect list MPs. I am not sure that is compatible with the Liberals’ promises about a new relationship based on nation-to-nation.

        Lastly, abstract notions of the public good may be difficult to define but I thought it remarkable that the public good had simply drowned in a tsunami of cui bono arguments. I am happy to advance a rights argument in favour of AV over FPTP but I will leave that to a later comment.

  6. Alan says “I would hope that the position of indigenous Canadians is a central part of any debate on electoral reform.” As would I. I have been biting my tongue for a long time. I have all kinds of ideas on indigenous representation. But I am not aboriginal.

    Alan says “The model advanced by the Canadian law commission would merge the three territories into a single common district to elect list MPs.” I don’t know why they didn’t say that, but what they actually recommended was to give each of the three territories a second MP.

  7. Pingback: Imagining the 2015 UK general election under AV | Fruits and Votes

  8. One thing I’ve noticed in New Zealand elections since the introduction of MMP almost twenty years ago is that in some electorates the differences between the single-member district vote and the party list vote are quite remarkable, in fact much more pronounced than in Germany. As I see it, these often dramatic differences are in large measure due to the fact that the party list vote is not affected by the narrow tactical constraints which normally distort voter preferences under FPTP – to be certain, the party list component has its own tactical constraints as well, but these operate in a far more subtle manner – and in this regard New Zealand stands as a very good example of how a new electoral system can impact voter behavior. With that in mind, I wholly agree that if Canada ends up adopting AV for House of Commons elections, there would be major first preference shifts in many ridings (relative to the preceding FPTP election), once voters realize that the first preference count won’t decide the winner unless the candidate in first place secures an absolute majority at that point.

    Another thing that comes to mind is that in Canada, as in New Zealand before 1996 (but unlike in Germany) FPTP has been for a very long time an integral element of a deeply entrenched political culture. In the mid-to-late nineties, shortly after the adoption of MMP, party list MPs in New Zealand were perceived by many as second-class parliamentarians; as this view went, the “real” representatives of the people were those chosen by FPTP. Perhaps that was inevitable: a sizable minority of the electorate had voted to retain FPTP, and many of those probably perceived the proportional compensation introduced by MMP as an unwelcome “tack-on.” It seems to me that this sentiment (also evident for a while in Scotland and Wales) has subsided over time, but I wouldn’t be surprised to see it surface in Canada should PR be adopted there; I could also imagine a variation of that sentiment with AV in place, directed against MPs elected on the basis of second and successive preferences, after having placed second or lower in the first preference count. I don’t think either scenario would be a show stopper, but they could cause quite a bit of friction early on – all the more so if the new electoral system is imposed by a parliamentary vote while bypassing a referendum.

    • For similar reasons, I would think the first president or governor elected on preferences if the US or any of its states adopted AV would face some questions of legitimacy.

      • Alan @5:38:

        Agreed, and this tends to confirm my theory (expressed recently in a different thread) that runoff avoids this perception for cosmetic reasons. “Only 38% favoured the winner, and the other candidate lost with 43% support” seems unfair to these people in a way that “the candidate who trailed by 5% on the first ballot nonetheless win the runoff 51-49” doesn’t.

        Query, though, how much one should bend the institutional rules to cater to the mathematically illiterate. The justice system does allow “apprehension of bias/ loss of public confidence” as grounds for disqualifying a public official from a particular decision, even when no-one is alleging any actual bias, and for excluding “improperly obtained” evidence, even when no laws were actually broken: but in other areas, the rule-makers insist on a more robust, filtered set of rules even if a cruder version fits better with the initial intuitions of someone with no expertise in the topic. (Eg, most PR systems use either the Droop Quota, or highest averages, even though the inevitable response of a layperson who is told for the first time “the quota in Australian Senate election is 100% of votes divided by one more than the number of seats” is to ask “Why not just divide by the number of seats, pure and simple?”)

      • Yes, explaining the Droop quota takes an extra effort. I do it in my classes by writing out the formula and saying, now let’s suppose there is one seat. It does not take long for students to get that 50%+1 is a good winning threshold, and that more than one candidate could have 50%. Then I just say, generalize to M=2, M=10, etc.

      • MSS, how do you then explain the wasted votes at the end of an STV allocation? I’ve seen it described as ‘the reason STV is less proportional’ than list systems, though I’m sure it’s a very minor one if at all. Is it best to compare it to the rounding mechanisms (ie like highest averages or largest remainders)?

      • jd

        STV is semi-proportional if any only if a cardinal vote under truncated voting is exactly the same as an ordinal vote that expresses a first preference under transferable voting. As soon as you state the premise, it collapses. STV is proportional to the last distribution of votes, not the first distribution of votes.

        The ‘wasted votes’ generated by the Droop quota are a bit of a furphy. STV works with the Hare quota where all votes are counted and no votes are ‘wasted’. However, the Hare quota disadvantages candidates elected on a surplus, because more of their aggregate vote is locked up with electing each candidate. Votes transferred ‘down’ from surplus candidates are devalued in relation to votes transferred ‘up’ from excluded candidates.

        The Droop quota is the number that both elects a unique set of candidates and is neutral between candidates elected on surplus and candidates elected on transfer.

    • Much of the difference in votes across tiers in New Zealand is due to strategic voting: Green voters often give their local constituency vote to the Labour candidate. There is also the fact that the constituency MP gets more resources for offices (most list MPs also set up offices, but they have less funding to work with). There is, as best as I can tell, no such resource or status imbalance in Germany, although even so ticket-splitting has been on the rise in the past two decades (i.e. post-reunification, and with the more complex party system that came along).

      It is also very much true that a large minority in New Zealand still supports FPTP, as was evident from the referendum in 2011, and this is highly correlated with preference for the National Party.

      • “and this is highly correlated with preference for the National Party”
        Well, that’s convenient, because if 50% of the electorate ever decide that the country needs a single-party National Party government, they can attain that very easily, by, uh, giving their list votes to the National Party. No need to lobby MPs to hold a referendum to revert to FPTP, or to gather signatures for a voter initiative, or to campaign in a referendum.
        In other words, PR systems are far more generous to FPTP supporters than FPTP systems are to PR supporters. (The main reason why I would have no moral qualms about abolishing FPTP without a referendum if that were politically feasible. If FPTP supporters want the “strong” single-party government they crave, and manage to persuade a majority of the electorate, then with PR voting they don’t need a referendum to get a single–party government – and if a majority of voters then change their minds and decide that coalition or minority governments are less bad for the country, they aren’t stuck with a voting system that keeps giving them single-party governments that only 36% of them want).

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