Thursday, April 4, 2013

On the Leadnow Co-op Poll...

Democratic reform advocacy group Leadnow, one of a couple of organizations to have endorsed Joyce Murray over the co-operation issue, decided to pay for a poll to be conducted on the issue of co-operation and Canadian's attitudes towards electoral reform. This poll, done by Environics, checks out more or less, and the company is a reputable one, so there should be no issues with this like there was with the CARP voodoo polls of a time ago.

You can read the full poll results here, but we're just focusing on the co-operation aspect for this post. Environics found these numbers when asking respondents how they would vote with a single progressive candidate in their riding, and this was Canada-wide:

The results show that between 60-75% of Greens, Dippers, and Liberals would vote for a co-op candidate in their riding. The runner-up choice is them deciding not to vote at all, followed by voting for the Conservatives then don't knows.

Now, I've decided to apply these numbers to 21 of the closest ridings that were won by the Conservatives in the 2011 federal election - a list which you can see here on Pundit's Guide. I did all all ridings on that list, except for the Yukon. I decided to apply these broad numbers to all ridings, the best I can do given that I have zero regional numbers. I decided to not include won't votes or don't knows, you get the following numbers voting for a co-op candidate: 93.5% of NDP, 97% of Greens, 90.9% of Liberals, 3.6% of Conservatives. Keep those numbers in mind.

Here's what I got for those 21 ridings, based on the 2011 results:

So, with 90%+ support going to the co-op candidate from the other parties, its pretty obvious that it works relatively effectively. Except for two ridings in Saskatchewan, 19 of 21 ridings where I put in those numbers flipped to the "progressive" candidate. In the case of the two exceptions, the numbers for the Liberals and Greens were so low that they barely helped out the NDP candidate - though the ridings were certainly more closely fought. This will be the case in any riding where a close race was fought in the 40% range or so. The remaining "progressive" candidates numbers aren't really big enough to make a difference, and it shows in all of the Western NDP ridings where they came close to knocking off the NDP and had their vote totals in the high range - even in ridings where they win in a co-op situation, the race is close. See Elmwood-Transcona, Palliser, and Van Island North.

Okay, so that's great! you say. Sure, but I already did this once before, I parsed out the numbers of some university professor and did a 1:1 vote transfer. At 90%+ transfer, we're practically at 1:1. And as I said before, 1:1 is just not a thing.

What happens when you apply the actual numbers of the polls to these ridings? This time I'll include those that won't vote, and half-and-halfed the don't knows. That gives the progressive candidate 73.5% of Liberals, 75% of NDP, 68.5% of Greens, and 7.5% of Conservatives. The Conservative candidate would get 84.5% of Cons, 10.5% of Liberals, 8% of NDP, and 6.5% of Greens. The rest have become non-voters.

Here's what we get under that scenario:

What happened!?

Well, very simply, the Conservatives retain more of their vote than the united progressive option does. And this is what I keep telling people: some of our voters will not vote if they're only given these two options. The poll itself is pretty clear in saying that - 18% of Liberals and NDP said they wouldn't vote, as did 24% of Greens. Only 8% of Conservatives said they wouldn't vote. That is a huge difference, one that cannot be ignored and shunted off to the side, as I did with the first numbers.

Basically, this LeadNow/Environics poll is saying that at least 1 out of 5 Liberals, NDPers, and Greens would not vote for a progressive unity candidate.

I already know where I stand, but that statement must make you think. In ridings across the country where this is co-op idea is implemented, this poll is saying that generally 1 out of 5 wouldn't go along with it. Just, think about that for a second. That's 20% of our support, poof gone, in 57 or 100 or however many ridings. That is a pretty big effect, one that people would need to account for.

Moving on, I mentioned it before briefly but there's an obvious issue going on. Its clear that the lower the third party numbers are, the less effective co-operation would work; we see this in the cases of the ridings the Conservatives win in the above models. This is because that every Liberal or Green in a close NDP/Con fight willing to go over to the NDP probably already had by 2011. It would give you an interesting metric to play around with, if so interested. Just figure out what the best ratio between second and third party support is, and you could find the best ridings where co-operation would work.

If Joyce Murray and Elizabeth May went insane - always a possibility - and said that they would tell everyone to vote for their riding's NDP candidate, based on these numbers and on the 2011 election results, the NDP would win 168 ridings, the Cons 137, and the Bloc 3. However, there are a large number of marginal ridings - 37 for the NDP and 32 for the Cons - that could flip it either way. Our own special polarized Parliament. Yay!

Anyways, this is just one poll and while you can play around with the numbers, you can't take them at face value either. It'd be nice if we had more numbers like this coming out. It would definitely be a plus for Murray's whole idea, and its interesting for someone like me to look at and parse out the results.

That being said, I can't get over the fact that these co-op pushers would be perfectly fine with alienating 1 out of 5 Liberals, or just overall non-Conservative voters. I mean, really? I thought this was party was about offering opportunity to everyone, not just the four people who are willing to follow them. My Liberal Party extends the same opportunity to vote for a Liberal alternative even to that one guy down the street that nobody likes.

It makes me wonder what I'd do if Burlington had to run a progressive co-op candidate that wasn't a Liberal (very unlikely, because we aren't exactly a riding that co-op advocates find would be the right fit for their schemes). There are very few NDPers I'd vote for, and most of them don't live here. I'm also not interested in voting Conservative. Hm, dilemma, and a false one I've been forced into at that. Lovely.


  1. the poll is actually questionable accuracy, for a number of reasons, but the fact that 80%of the respondents claim to have voted in 2011 when only 61% actually did means that this poll is not an accurate representative sample (out by a wing 19%), or 19% of respondents lied. Neither scenario is good news for the polls accuracy.

    The best data that I have seen on this is the Liberal Party's predictive data in BC. It shows that 2/5 Liberals would never vote NDP, and 1/3 NDP would never vote Liberal. I think that push comes to shove these numbers are likely more reflective of what would actually happen. Your scenarios above are much less positive with these numbers.

    I don't think we can afford to take a risk on cooperation, especially when other polls done in the same time frame suggest that the LPC could form a minority, or in some cases a majority, without having to cooperate at all.

  2. Quick question Kyle - did you include the 3% of Conservatives they show flipping to vote for the co-operation candidate?

    That number didn't make sense to me at first, but it was pointed out to me that some Conservatives who support electoral reform (and polls show a fair number do), they may be willing to "lend their votes" to see it accomplished before reverting to their previous allegiances. Either way, the poll does have 3% flipping to the cooperation side.

    1. I did Ryan, thanks for pointing it out. In each case its nktae that there is some exchange both ways.

    2. Er, important* not "nktae"... Stupid phone.

  3. Thank Kyle, you did a lot of work here to suport what ought to be an obvious conclusion. I am continuously surprised at just how simplistic the proponents of 'Unite the Left' are. It is tempting to label them 'simpletons' and move on, but they do represent an awful lot of very angry and motivated people.