Canadian federal election - 2015

[Boston Bread]

Abacus weighs in
31 (-5 from last poll)
28 (nc)
28 (+4)

http://abacusdata.ca/wp-content/uploads/2015/06/Abacus-Release-Headline-Political-Data_May2015.pdf

Interestingly NDP is taking evenly from LPC and CPC voters. Further 8% of 2011 CPC voters support LPC, 9% NDP, so another even split. It looks like whole LPC->CPC voter to prevent an NDP win phenomenon is overblown. In that case maybe we should start talking less about blue liberals and more about orange tories.

[Boston Bread]

Interesting graphic:



The NDP actually has the largest base (16%) and is the most inoffensive to voters (more would consider them than any other party). It's just that the NDP are doing poorly with voters who are still deciding between them and LPC/CPC. It suggests that the NDP both has the highest floor and highest ceiling, and that the NDP are merely closer to their current floor. Would be fantastic if true.

[Boston Bread]

If I understand that graphic correctly - given the phrase "federal political parties" and the percentages adding to 100 - Bloc-NDP swingers would be included in "NDP only", since the NDP is the only one of the federal parties asked that they would consider voting for.

Most of the Bloc probably selected "none". Considering that Green+Bloc+Other is about 9% in that poll it's the most likely possibility.

[Boston Bread]

I could see an accord like Ontario 1985 - defeat the Conservatives on a confidence vote, then let the second place party govern as a minority without a formal coalition, although the third party would get some concessions.

[Boston Bread]

Latest ekos numbers have the liberals fading fast.



http://www.ipolitics.ca/2015/06/05/the-ekos-poll-mulcair-is-picking-up-all-the-marbles-as-trudeau-fades

[Boston Bread]

Ekos' second preference poll matches Abacus' findings. Not only does NDP has the highest ceiling, but Conservative's second preferences are evenly matched between NDP and LPC. And there are a lot of voters wafting between LPC and NDP.

[Boston Bread]

Hopefully all these greens in BC, Saskatchewan and the Atlantic do the right thing and switch to orange (I find it surprising that the Green vote is nearly double the last election, despite the fact that the NDP and Liberals are both up as well).

Ekos always overestimates the Greens, and on top of that the Green #s usually decline by election day. If you average the polls Greens are only at 6% so I'd say a 4-5% Green result is more realistic on election day.

[Boston Bread]

My model now produces a CPC-NDP seat tie with the new numbers:

[Boston Bread]

Time to start pondering who will replace Trudeau as the leader of the Liberal Party.   Scott Brison?  Dominic Le Blanc?  Joyce Murray?  Someone from the Ontario Liberal caucus?

It would be a mistake for the LPC to continuously purge leaders like that. If they are a distant 3rd, they need to accept their position and start giving their leaders more room for failure as the NDP usually does. The biggest criticism of Trudeau is inexperience, so letting him stay on might eventually result in better leadership.

[Boston Bread]

Forum's seat model is hilariously bad. A 50 seat lead while trailing by 1%, really?

This poll also polled respondents on how they voted in 2011. The results were 41% CPC, 28% Liberal, 23% NDP. lol.

[Boston Bread]

The NDP only needs about 40% of the vote in Quebec to get 57 seats (they're at that in half the polls), and the conditions in Quebec now aren't that different from 2011, so it's still a reasonable outcome.

[Boston Bread]

Do the Bloc exclusively steal off left-leaning votes, or does it also get CAQ conservative Francophones who might vote Tory as well?

A good chunk of CAQ voters voted NDP federally, so it's not like the NDP live off exclusively on left-wing votes either.

[Boston Bread]

My poll average in Ontario comes to 32.8% CPC, 32.4% LPC, 28.8 NDP. My model has the NDP gaining these seats, all from the Conservatives:

Brampton East
Brantford-Brant
Essex
Hamilton West-Ancaster-Dundas
Kenora
Oshawa
Sarnia-Lambton
Sault Ste. Marie

Losses:
Scarborough Southwest

They actually hold or win back all of their 2011 seats in the GTA, albeit by a close margin against the Liberals in several of them. Impressive considering they were looking at a near wipeout in the GTA before the Notley/C-51 bump.

An interesting bit: A bit earlier we were talking about the possibility of a Liberal pickup in MacKay's seat. But I think the NDP, rather than the Liberals would place second there based on current polls, even with the large Liberal lead province wide.

[Boston Bread]

You must be logged in to read this quote.

Since the BQ averaged 16.5% in QC in the latest 308 projection, 3 times more support equates to ~50%.

"Potentially" speaking... "Houston... we have a problem".

Actually, the BQ internal poll had them at 29% with Duceppe. Probably the biggest bump they can manage.
http://www.lapresse.ca/le-soleil/actualites/politique/201506/09/01-4876726-retour-de-duceppe-beaulieu-a-lui-meme-entrepris-les-demarches.php

[Boston Bread]

As long as Trudeau doesn't approach Iggy levels he will stay on because at that point the Liberals will realize have nothing to gain by recycling leaders, and let Trudeau get some experience.

I don't know if this amazes me or makes me laugh. Those ON numbers...

[Boston Bread]

On those ON numbers, the NDP would win rural southern Ontario. My model shows them gaining seats like Haldimand-Norfolk, Oxford, Haliburton-Kwartha Lakes-Brock, Barrie-Springwater-Oro-Medonte, Lambton-Kent-Middlesex, Oxford, and Parry Sound-Muskoka.

[Boston Bread]

I would love for the NDP to take Parry Sound-Muskoka. How does Nippissing look?

My model still shows Liberals taking Nippissing under those numbers. Which sorta makes sense, considering they won it notionally while finishing in 3rd in 2011.

[Boston Bread]

This is my model's overall seat projection, using an average of the recent polls (so it includes Forum, Ipsos, CRA, and Abacus numbers which are less optimistic for the NDP)

With the NDP stagnant in AB and rising in the Maritimes, the NDP's best seat is no longer in Edmonton (though they still manage a respectable 69% in Strathcona) but in St.John's East (70%). You can see the EKOS Green bump in SK rather visibly.  In fact SK is the best province for the Greens in my average now. But what's most significant is the strong rise for the NDP in Ontario, and when compounded with the CPC falling to Hudak levels there they wind up 3rd in the overall seat count.

[Boston Bread]

The NDP got 32.6% of the vote in NL in 2011. They're at 29% in my average. So there isn't a big drop for the NDP there, more so for the CPC.

Also, directly inputting the EKOS results give me
152 NDP
86 Con
95 Lib
3 Bloc
1 Green
And 1 FetD - due to other polling at 5% in Quebec.
I believe the EKOS poll is underestimating the CPC, however.

[Boston Bread]

Hatman, the model is assuming a large portion of the "other" vote in Quebec is for FetD. In my weighted poll average the other vote in Quebec is 2% which is far to weak for FetD to win a seat. This is also after I apply a penalty that reduces the "other" vote.

But directly inputting EKOS results which show other at 5% in QC, FetD increases a lot with it. And since my model believes their vote is concentrated in the 2 incumbent's ridings, the proportional swing is large enough to let one win, albeit barely. Roughly 4% "other" is the threshold where my model thinks the FetD starts to be competitive in a riding. (Although only 2-3% of that would actually be for FetD, since 1-2% go to other minor parties).

[Boston Bread]

When you assume...

If "other" parties in QC get 5%, where exactly would they be going? It's just an artifact of the polling and mathematical magic.

I don't believe F&D has a chance, given that I doubt 5% in QC will vote for a party other than the "big 5". But if 5% did it is a possibility.

I already account for this in my main model where I weigh several polls and penalize "other" votes. Don't take my 1 F&D seat projection as a prediction.

[Boston Bread]

Again: below 25%, I *cannot* see the Liberals winning 80 seats, much less 100.

I can see it if:
1) Much of the vote goes to Green/Other. In the EKOS poll the two get 11%, while in 2011 they got 5% combined. That's 6% less for the big parties.
2) The CPC vote collapses. Suburban Ontario is EXTREMELY key for the Liberals. It's the difference between 60 and 100 seats. And what wins suburban Ontario for the Liberals is not a low NDP vote, but a low CPC vote due to how poorly the NDP fare there. A 36-26-26 split allows Liberals to win a whole lot of GTA ridings with about 40% of the vote, simply because the Conservative vote falls through the floor. Tip the Liberals just a couple points lower and the bottom starts falling out for them as well.

The EKOS poll is suggesting 1) and 2). But I doubt either would occur an actual election.

[Boston Bread]

So far Duceppe's leadership bump is no bigger than Paillé, who polled in the high 20's for a month or so after becoming Bloc leader. Disaster averted?

[Boston Bread]

Although the Tories have never won a majority in the years in which the Bloc won Quebec (1993-2008)

[Boston Bread]

48% in Quebec omg! Post-Duceppe or not? If so, RIP Bloc.

You must be logged in to read this quote.

NDP actually tails CPC in ON 36-34 in the poll, not leading