What just happened? (V) - Macleans.ca

What just happened? (V)


Alice Funke questions the vote-split theories.

If this analysis of the numbers bears out, the conclusions are straight-forward:

  • Where the Liberals lost votes mainly to the Conservatives, they lost seats to the Conservatives.
  • Where Liberals lost votes to the NDP as well as the Conservatives, they lost seats mainly to the NDP.


What just happened? (V)

  1. Liberals need to realize they lost, not because of the NDP or the Conservative attack ads, but because of boneheaded leadership that is out of touch with middle-class Canadians, but in touch with media pundits and Bay Street lobbyists and lawyers.

  2. From my perspective there was more than just a one way vote split happening here, and since the analysis doesn't consider the other major factor, it's not terribly insightful.

    Besides Quebec, Ontario was the most active province for riding shift.

    Ontarians have routinely voted for both conservative and liberal provincial governments, which in my opinion points to a populace tending towards center to center right voting generally speaking.

    Much like many of the people I know, the idea of Layton forming a coalition as PM is far more scary than a Harper majority, and as a result many of them voted CPC, despite being traditional Liberal supports.

    Quite simply, many Liberals followed the wave and voted NDP, while other Liberals freaked and voted CPC to stop that wave.

    So of course it was a split vote. It just split into more than two parts, and goes to show why the NDP will never form government.

    • Dude, the linked article says exactly the same thing you're saying, but it looks at the numbers and takes one more step, saying that in ridings where Liberals went to the Conservatives, the riding went to the Conservatives and in ridings where Liberals went more to the NDP, the seat went to the NDP.

      • Doesn't seem so to me. My understanding is that she's suggesting the vote split essentially didn't happen. That there was merely movement one way or another.

        I'm saying that it was 3D split, wherein both the NDP rise in support AND the perception of that rise, drove voters to either join the rise, or resist it by voting CPC, both leading to abandonment of the LPC position.

        Perhaps the main the difference is what we attribute to causing the shift. She seems to think Liberal voters were "converted" whereas it seems clear to me they were running like hell from Jack. LOL

  3. Not sure I buy her assumptions. Just because the Liberal down and the Conservative up percentages are close, doesn't mean that was due entirely to a Lib/Con switch. Nor can you assume that the NDP support in those ridings was all coming from Green switchers and new voters. It is just as possible that the Conservative increase came from a combination of both Liberal switchers and new voters (who may have been motivated about concern of the NDP rise), and that the NDP increase came from a combination of the same sources, thus the Liberals lost votes to both the CPC and NDP and both may have gotten new voters.

    Furthermore, those who advocate the "don't split the left vote" usually expressed it as – one of the Liberal/NDP should not put up a fight in a given riding to preserve a "majority" of left leaning MPs. Therefore, in all of the ridings she refers to, the left vote was already "split" between the Liberals and the NDP, what happened in this election just made it worse – for the left.

    Incidentally, I saw reports that the CPC was hoping that the left vote in Ontario would split just enough to push enough seats their way, and in fact they were counting on it for a majority. I suspect they know what they are doing, as it seemed to have worked.

    That being said, I suspect (though what do I know, I'm no expert) that the Liberals lost votes both to the NDP and to the Conservatives and I would attribute it to the fact that the Liberals were fighting on two fronts, didn't pick a battle, and didn't fight either of them very well. I also think Anon has a point – the LIberals were out of touch with almost everybody outside Toronto (where they had 34 seats – almost half of their 77), Vancouver, and Montreal. They didn't do anything to change that and they lost ground with the support they did have. Rebuild, big time, would be my prescription.

    One thing nobody has talked about is the fact that the weekend before the election, the CPC "leaked" their supposed internal numbers that supposedly showed that it was highly unlikely that the the CPC could get a majority. Yet, Guy Giorno said on election night that their range of expected seats was 151-163 – majority much more likely.

    I wonder whether this lulled the anti-Harper voters into a false sense of security – not to worry if you don't vote against Harper, he won't win anyway.

    I also wonder if that was exactly what the CPC intended.

    • Yes, your first paragraph makes a key point and that is the flaw in Alice Funke's speculation. In the US they do exit polls and you can ask who someone voted for and whether they switched their vote from one party to another. Without that kind of information, this is speculation and I would say it looks like speculation mean to reinforce a certain bias. If Alice Funke wants to push that she will somehow have to find a way to survey people's past and current votes.

  4. Alice Funke is speculating. She has no way of knowing which party lost voters to which other party. All she knows is what the final vote counts are. One can look at the close races where the Liberals lost to the Conservatives and see that in most cases the NDP vote increased significantly from 2008 and that if a fraction of that vote had gone to the Liberals a Conservative would not have been elected. That is what is called vote-splitting. The rest is speculation.

    In particular look at the very close races, a couple were lost by less than 30 votes – one I think by 14. In that case, had 15 voters voted Liberal rather than NDP, the CPC would not have won. Same thing happened in 2008 in Kitchener Waterloo. No one disputed it was vote splitting then. I think the NDP and their supporters and Alice Funke are more sensitive this time because they are concerned about how voters view the Harper majority in connection to the rise in the NDP.

  5. That piece by Pundit's Guide is pretty awesome. The numbers are pretty tidy, so the conclusions drawn (and quoted here) are reasonable. Even if you disagree with this particular narrative about where votes went, this is at least as plausible (way more plausible, really) as any argument made that people moving to the NDP somehow delivered enough seats to Harper that he got a majority.

    Anyone making (or buying into) vote splitting arguments needs to realize that they're doing two things:

    1. You are second guessing (and insulting) voters
    In stating that you think some batch of voters managed to deliver the seat to the wrong person by voting in a certain way, you're suggesting that people voted against their interest. This is flat out insulting to them, and entirely disrespectful of the democratic process. They voted how they wanted, and that is pretty much the end of the story.

    2. You are questioning the legitimacy of our electoral system
    Implicit in assertions of vote splitting is a criticism of either FPTP or a criticism of multi-party democracy. With multiple parties in a FPTP system, results that return candidates who received significantly less than a majority of votes in their riding aren't exactly uncommon. Indeed, this has been the reality since at least '93. If you have an issue with either FPTP or our multi-party system, stand up and say so rather than wishing for something magical result that doesn't exist in reality.

    The bottom line?
    The results are the results; respect what voters have said.

    • The piece is not awesome because Pundit's Guide has no information on voters switching from one party to another. She has no information on how many who voted CPC voted CPC previously, didn't vote, voted Liberal, Green, NDP or something else. Without that kind of data her whole thesis is speculation. One can see from her full writeup that she thinks others who take a vote-splitting analysis make the NDP look bad and she is trying to counter that.

      Vote-splitting has been used to describe the phenomena that voters who really don't want a Harper majority and would prefer any of the other parties in power, split their vote between several parties. The existence of vote-splitting has not been disproven and to the extent it occurs, it helps the CPC more than those voters would like.

      • Just to remind you, in case you have forgotten:
        You can't vote against any particular party or candidate; you have to vote for someone.

        Your state that this piece is unsupported, but your assertions about the facts of vote splitting are even more unsupported. You ask about numbers about vote switchers, where are your surveys of voters who've said, more than anything, that they didn't want Harper?

        This piece is awesome because, as a well argued and pretty well supported piece using the numbers available, it allows statements like yours to be shown for what they really are: arrogant, baseless assertion of your own belief.

        • Where did I say there was vote-splitting? I said vote-splitting had not been disproven. And it hasn't. Actually the support of strategic voting sites shows that there is at least some vote-splitting. As to how many people would actually prefer either a Liberal or NDP government over a Harper government, I have no idea. I am one of them and I know some others, see quite a few on blogging sites, but I have no information as to exactly how many voters this adds up to. I'm sure there are voters who would prefer either an NDP or Harper government to a Liberal government too.

          As to voting against/for, I'm not sure what you think I have forgotten. I know voters who voted strategically in the hopes that their vote would prevent a CPC MP from being elected – some feel they were successful, some feel they weren't, although all are disappointed in the overall outcome of a Harper majority. All the ones I know, did vote for the candidate who ended up being first (the ones who consider themselves successful) or second. So, it follows that if more people had followed their lead, we would have fewer CPC MPs. These people don't seem to figure into Alice Funke's analysis. I have no idea how large a group they are, but they do exist.

          If you have arguments, please stick to them and leave out the slurs against those who disagree with you. Discussion works better that way.

          • This whole discussion in general, about vote-splitting, frustrates me deeply. I've read you disagreement with the linked post and your description of vote splitting as you supporting a statement like "vote splitting exists and was a factor in this and previous" or something along those lines. I apologize if I am in error, but you must admit that it isn't exactly a far reach on my part.

            As far as "slurs" go, I believe you are misreading me if you think I use that kind of language because we disagree:
            I use strong language because I believe that talk of vote splitting delivering a seat one way or another is extremely arrogant and insulting to voters and to our democracy and because I believe that people talking of vote splits deserve to be confronted the implications of the statements they make.

            While the existence of strategic voting sites does affirm that there are people who either view the NDP and Liberals as interchangeable and/or that there are people who view keeping the Conservatives from gaining a seat as their primary goal, it is a far reach to say that seats were delivered or no because of their efforts or that they, in other words, split their vote.

            Again, your evidence is comprised of anecdotes and assertion, and I find it to be significantly less credible than the linked article which is logically constructed, uses actual data, and is quite plausible. Don't get me wrong, I don't mean to say that the article is some kind of certainty, but it is more than enough to counter what you offer up.

          • If one had a two party system and the voters lists for two elections AND if the shifts were large compared to changes in the two voters lists (voters dropping out between the two elections and new voters being added), THEN one could make the claims Alice Funke tries to make about voters switching parties. We don't have a 2 party system, so this is not possible without surveying people on how their votes changed.

            The concept of vote-splitting is sometimes used to mean nothing more than how could the smallest number of votes have been distributed differently between non-CPC parties in order to get fewer CPC MPs. Usually people asking the question of how one might get fewer CPC MPs, want fewer CPC MPs. These people are often inclined to vote strategically and may try to avoid so-called vote-splitting. This is an understandable analysis and is useful for those people. For others, I think tedbetts summed up the situation accurately and succinctly below.

          • I'm at a bit of a loss for how to reply to this. We've been having a discussion I don't just want to leave it hanging, but I truly don't know what to say in reply as you appear to be claiming the smallest amount of ground possible in your last comment: the first paragraph states that we can't know everything definitively without asking directly, and the second tends towards viewing this whole thing as something along the lines of "some explanations work better for different people" which avoids the entire discussion.

            I'm a little confused. You're all over this blog posting (even more than I am), and that's all you're saying?

          • I. Somewhere on the web there is an analysis which shows that if about 6000 voters in 14 ridings had switched their votes between Liberal and NDP, Harper would not have a majority. That kind of analysis interests those who would have liked to prevent a Harper majority. I would put that in the vote-splitting category of analysis.

            II. An analysis based on comparing which parties increased or decreased their vote share in each riding can also be interesting, but since these changes can be due to new voters, voters staying home, straight switches from party A to B where A goes down and B goes up, or more complicated switches, from A to C and C to B, where A goes down and B goes up and C may go either up or down, I don't see how one can get much out of it. One can make assumptions, but they may or may not be correct.

            So, I don't get much out of analysis II, but find analysis I interesting and useful. I understand you think the opposite, but I don't agree.

          • Your line about assumptions works equally well for both analysis.

            It isn't that I find value in one and not the other (I primarily like a hypothetical analysis II because it puts analysis I in question), it is that your hypothetical analysis I has important implications that are very unkind to voters and democracy.

            I've outlined these previously; at this point, I'm repeating myself.

    • In this era of Rovian analysis for FPTP electoral systems, the results are never attain the finality you are calling for and respect is well out the window. Regrettable.

      • While what you describe is indeed regrettable, it isn't what bothers me in this instance:

        If people think the results are somehow unfair or unfortunate, I have absolutely no issues with that. What bothers me is that people don't see the inherent assumptions they make about the illegitimacy of multi-party FPTP systems and that they don't try to resolve these issues by either coming out and stating that they either support a two party system, or perhaps that they think FPTP is outdated.

        Blaming vote splitting is much like throwing your hands in the air and blaming an act of God; this is the system we have, this is the way the election was run, everyone agreed on the rules at the beginning. If you don't like the outcome, either come to terms with the way the system works or consider reform.

        • I am assuming your sentence "What bothers me…" is meant to read:
          What bothers me is that people won't acknowledge their tacit assumption that multi-party FPTP systems are illegitimate or outmoded and would be better replaced by a two-party or PR system. Am I reading that wrong?

          I would "but" your second paragraph with a third option, which is the Rove/von Nuemann game theory that seeks the small differences that tip the balance to one's favour in relatively stable selection groups by pushing the reading of the rules in the worst John Yoo spirit. Such results are fleeting, but then so are the lives that lust for power in this life above anything else. As much as you or I might be tempted to dismiss such attitudes as unsporting, their glamour is far too tempting to the broadly distributed amoral personality.

          • Sorry for the lack of clarity, that is a perfectly fine reading.

            First, I'd hope that no one in Canada feels the need to go that direction. Second, I'd hope that the political situation in Canada always remains fluid enough that tactics like that are unnecessary.

  6. In the 2008 election the CPC vote was higher than the latest polls showed and people didn't attribute it to people switching votes from Liberal to CPC. They attributed it to Liberal voters staying home and the CPC doing a good job of GOTV, along with CPC support continuing to rise over the weekend. I see nothing in any data that shows that these three factors did not play a role again in 2011. On top of that, of course, we have the NDP surge also playing a role. The four factors would be deadly to the Liberals and advantageous to the Conservatives and NDP – which is what we seem to have got.

  7. Let me expound on her thoughts.

    Where the Conservatives won the most votes, the Conservatives won seats.

    Where the NDP won the most votes, the NDP won seats.

    Though more rare a circumstance, where the Liberals won the most votes, the Liberals won seats.

    Where the Liberals did not win the most votes, they lost seats to the Conservatives, the NDP or in four cases the Bloc. This happened in every single riding the Liberals did not win.

    There was not one single riding, in the entire country!, where the Liberals won the seat with fewer votes than the winner.

    • +1

    • Yeah it's amazing isn't it? How the party with the most votes wins? Crazy stuff that. LOL

      Too bad that has nothing to do with vote splitting, or she may have actually had a point or somethin'.

  8. Yes, but wouldn't one united left-of-centre party have won all those seats regardless?

    When the conservatives have been united over the course of the past thirty years, they have dominated. It's something the progressives of this country might want to think about for a while. Just saying.

    • A two-party system would be a horrible outcome for the country. Either we'd end up with a situation where both parties collapse to the center searching for swing votes, which would alienate huge chunks of their bases, or we'd end up with a situation where they deliberately polarized, to try and neutralize the center, depress voter turnout, and bring out their base.

      Neither of these is an appealing option and to be completely honest, I'm disappointed that conservatives in this country picked electoral success as their main priority and settled on a single Conservative party.

      • How can you enact any principles you stand for without having power? And the progressives in this country won't have it if they keep letting Harper get his way, right? I think it's just political reality.

        • How can you expect to have principles if searching for power is your only real goal? As evidence, see the CPC (who have traded on nearly everything they previously stood for these past few years) and LPC (who lost their focus well over a decade ago now).

          Yes, I did vote NDP this past election. No, I don't expect they'll be any better if/when they get near the levers of power. However, I expect to be able to turf and humble them (as happened to the PCs in '93, and the LPC now) if/when they screw up.

          It is way harder to do that in a two party system.

  9. Where the final polls didn't reflect the daily polls, it was the electorate that switched.

    • The daily polls showed about 10% of voters undecided right up to the end. I even talked to voters who were still undecided as late as 3pm on May 2nd, but who nevertheless did plan to vote that evening and hoped to make up their mind by the time they had their ballot in hand. 10% plus some voters deciding to stay home could easily account for any difference between polls and final results, although all polls have error bars as well.

  10. The real significance of the federal election–a Globe and Mail story, "Canada's new electoral divide: It's about the money" (May 4), states that "The true divide, the new reality of Canadian politics, is between the economic heartlands that the Conservatives now dominate throughout the country and the economic hinterlands won by the NDP." I differ.

    The true divide, as it was in the 2008 election but ever more so now, is between Québec and the Rest of Canada (RoC, once quaintly known as English Canada). The Conservatives in Québec this year won 16.5 per cent of the popular vote and only six seats out of 75, that is eight per cent of them.

    In the RoC the Conservatives won 48 per cent of the vote (almost a majority, in a contest with three other serious parties) and 167 of 233 seats, that is a whopping 72 per cent of them. The difference with Québec could hardly be more pronounced.

    The clear fact is that the Conservatives are dominant at this point in the RoC while barely a force in la belle province. Moreover Ontario, Alberta and British Columbia are in line to receive significant numbers of new seats to reflect their increase in population. Most of those seats will be suburban ones, just the sort of seat very likely to be picked up by the Conservatives. So it seem probable that their dominance in the RoC will increase; meanwhile it is hard to see any great breakthrough for them in Québec in light of the three most recent federal election results there.

    So the true great Canadian political divide looks well set only to widen further.


    • Interesting points. 48% in RoC with 3 other parties is very significant. This after what most described as a lacklustre leader's campaign and lots of no-show candidates.

  11. I am soooooo tired of the comment "Harper only got 40% of the vote" Actually Harper got 40,NDP 30, and Lib 20. If you want it ANY other way then MERGE!!!!!!!!!!!!!!

  12. This is something I noticed, too, and I think it has the potential to be extremely harmful to the country. What makes the present situation even worse is that Quebec's economic interests are pretty dissimilar from the West's:

    The high dollar is working out great for Canadians who aren't employed by manufacturers that export their products to the US, but Bombardier is having a lot of trouble competing with Brazillian company Embraer and pharmateutical companies like Phizer are thinking about moving pill production to India. Manufacturing centres in Ontario are also hurting under the high dollar (and also voted for the NDP).

    A lot of Quebec's demands over the next four years will be the usual bs, but some will be entirely reasonable. One trenchant way to judge Harper's performance will be how he decides which is which.

    PS — One small nitpick: the Tories took 161 (69%) of the RoC's seats.

  13. While that is true. Trawling for votes among the MC in and of itself is no way to provide leadership.

  14. You're right that I don't fully know which way voters switched. And I said I didn't know in the cited article.

    I merely said that there were many hypotheses worth testing that fit the data, and that we couldn't test any of them until we have (a) the poll-by-poll results which don't come out for another three months, and (b) the Canadian Election Study data which won't be made public for a year.

    Also, you'll note that I limited my observations to the GTA/905 ridings I looked at and no others.

  15. But, if I may, you still haven't absorbed the possibility that Liberal votes may have switched to the Conservatives, and that the NDP may have simultaneously gained from the ranks of 2008 Green voters and 2008 non-voters.

    If that is what occurred, it's most definitely not vote-splitting, it's vote-shifting or -swinging.

  16. Sorry, you haven't understood the post, then, because I clearly say there is more than one explanation for the data, that needs to be tested when we have more detailed data available.

    The Canadian Election Study is the only "exit poll" we're probably going to get in Canada. To catherine's point, I'm not going to conduct an exit poll, because I'm just a volunteer blogger, so let's get real.

  17. Turnout went up in 2011, unlike 2008, in every one of the ridings studied.

    So, if Liberals stayed home in 2011, then logically & mathametically it was even more previous voters who came back to the polls to increase the vote of the NDP and Conservatives.

    That part is zero-sum game.

    If you argue that 2008 Liberal voters switched to the NDP in 2011, then you have to explain where the increase in Conservative support came from, and the only remaining answer is 2008 Green voters and 2008 non-voters. It could be true, we have to test it. And we won't have the data to test it with until later on, as I've explained several times.

  18. Not quite what I said either.

    I said that where the Liberal votes seemed to switch primarily to the Conservatives, the Conservatives gained seats.

    And where the increase in the NDP was greater than could be explained from merely the drop in Green votes or increase in turnout, therefore it must have come from out of the 2008 Liberal support. And those happened to be the areas where the NDP gained seats.

    And I didn't assert these propositions as facts. I just said that if the earlier hypotheses were found to be true, those would be the conclusions that followed.

    Are people having trouble with the subjective tense or the scientific method or something? Sheesh.