i think maybe you were undersampling them. that ndp bump in saskatchewan, especially, makes sense, in terms of strategic voting. and, the fact that you’re broadly in line with nanos is nice in terms of developing a little consistency.
www.ekospolitics.com/index.php/2015/10/liberals-maintain-lead-over-conservatives/
Johnsmith
Have to laugh at those wildly optemistic Lieberal share of the vote in Ontario. I see the NDP is building momentum again taking vote share from the Liberals.
graham
Well, it’s actually in line with the other companies.
deathtokoalas
yeah, and the liberals are up over the last ekos poll.
the last provincial election was
liberals – 38.65
conservatives – 31.25
ndp – 23.75
…and the polls consistently got the liberals and ndp about right, but overestimated conservative support.
that was enough for a liberal majority in ontario.
the ridings are not identical, but they’re close enough to ballpark it. that would convert to 64 federal seats at 38.65%. with the conservatives and ndp both running lower, and the liberals up almost ten points on it, i don’t think 75-80 seats is an unreasonable projection.
that said, that brick wall of rural ridings is a real thing and there may be a kind of ceiling, where it no longer matters much if they get 43% or 48% or 52% even in terms of seat totals. they’d have to really push higher to start thinking about winning those seats. they may be nearing that ceiling, or may be past it.
these uniform swing models are going to be brutal at correctly distributing a 30+% swing. you’re going to see 50% swings in some ridings, and 5% swings in others. it matters less with a 6% swing or something. but, this is huge and they will get a lot of it wrong.
i saw some polling in larry miller’s riding that suggests the liberals are in striking distance, but there’s some extra circumstances there that are probably not transferable to other rural ridings.
but, if they’re running that high, and the ndp are down, then you need to look at them maybe winning a few long time ndp seats. maybe. don’t anybody get their hopes up. maybe even in the north. and, if the conservatives are running under 30 then the entire ring around toronto swings or is at lest very close – brantford, kitchener, newmarket, barrie, peterborough…i think that’s where you start seeing those 40-50% swings….
it would be useful for grenier or whoever else is doing this to come out and say this, so strategic voters understand to be extra careful with the model this time around.
with a 30% swing in ontario? you can’t use a uniform swing. it’s just too big.
and, a uniform swing in alberta makes little sense in the smallest of changes.
and quebec? and bc? yeah. good luck modelling that…
graham
Actually, the Conservative vote federally tends to be underestimated by the pollsters.
It was underestimated 2% to 6% in 2011
deathtokoalas
i agree that this was an anomaly. but, the last provincial election in ontario consistently had the conservatives at or above 35. grenier had them at 35.8, with a minimum of 32.8. that led to a prediction of a liberal minority.
my best guess on that is low turnout. see, here’s a twist pollsters haven’t yet considered: they claim the over 65 is more likely to vote. i don’t disagree. but is the over 75 more likely to vote? can they get to a polling station?
my guess is that what happened in ontario was they oversampled the really, really old voters, and they just couldn’t get there to do it. and i might suggest that’s something to be aware of.
now, does that mean i’m suggesting 28 is exaggerated? probably not. but i wouldn’t expect harper to outperform hudak, either.
it’s a valid question, though. the demographics are 65+. but, your average 65 year old is mobile. your average 80 year old is not. how much of the sample is 65-80, and how much is 80-100?
there’s a whole host of issues with the oldest voters, which are a substantial and increasing percentage of the population.
can someone drive them?
will they remember what day voting day is?
will they be too sick or sore that day?
might their kids decide to not let them vote, due to disagreements?
this is a provincial map, with the liberals at 39% and the conservatives at 31%. you can see what i’m saying.
http://cnews.canoe.com/CNEWS/Canada/2014/06/13/ElxnResultMapTAB.jpg
grenier’s map (with a uniform swing with the liberals at 42 and conservatives at 42) has the conservatives winning a lot of those ridings, and the liberals overperfoming in places like haldimand.
that’s because he’s applying the swing evenly, which is probably underestimating liberal support in the places they won provincially and overestimating it in the places they didn’t.
http://stephenmcmurtry.org/election_map
i would be more likely to bet on the provincial map being closer, which is going to bet liberals seats in brantford, cambridge, kitchener, thornhill, milton, flamborough (maybe; the riding splits), burlington, newmarket, barrie, durham, peterborough, northumberland – it’s a nearly clean sweep all the way around toronto.
and, you add it up, and it gets close to 80.
sorry, the projection map is with the liberals at 42 and the conservatives at 32, clearly. but i’m really checking to see if everything is getting moderated, or just the posts with links.