UPDATED - EKOS of the Week: It's baaa-aaack ...

Now, there is, of course, lively debate amongst polling methodology geeks over the relative importance of sample size vs. freshness, so this seemed like the best way to present it without understating either of those factors. The plus/minus for the first week refers to the change from the previous sample -- June 16 to 21, which is available here. 

 First, the national numbers, which, due to the extended summer field cycle, I've broken down by week so you can get a better idea of exactly when (and where) the shifting sands of Liberal fortunes are slipping through the hourglass of electoral doom. Wait, did I give away the ending? Sorry about that. 

Anyway, onto the numbers!

     

	Week 1 (June 22-29) (MoE 2.18)	Week 2 (June 30-July 6) (MoE 3.08)
Conservatives	30.6 (-0.4)	34.4 (+3.8)
Liberals	26.2 (-1.5)	23.9 (-2.3)
NDP	18.3 (+1.8)	17.9 (-0.4)
Green	12.6 (-0.4)	11.2 (-1.4)
Bloc Quebecois	9.5 (+0.2)	10.0 (+0.5)
Other	2.8 (+0.3)	2.5 (-0.3)

As you can see, the Conservatives lost a sliver of support during the leadup to the summitacular, but managed to pick up nearly four points during the aftermath. (At the moment, we don't have access to the nightly numbers, which makes it difficult to tell whether there was it was a gradual slope or a sudden blip. Meanwhile, the Liberals sink to a new low of 23.9 percent -- at least, I think that's a new low; it's hard to keep track --  and the NDP seems to have settled into the high teens. Finally, the Bloc Quebecois head back above the 40 percent mark, the Greens are in a bit of a slump, and it all worked out even for The Others in the end, so that's alright. 

British Columbia	Week 1 (June 22-29) (MoE 6.86)	Week 2 (June 30-July 6) (MoE 12.25))
Conservative	33.8  (+6.1)	33.2 (-0.6)
NDP	26.3 (+4.0)	24.2 (-2.1)
Green	18.1 (-1.1)	11.7 (-6.4)
Liberal	17.1 (-12.0)	26.9 (+9.5)
Other	4.8 (+2.4)	4.0 (-0.8)

Alright, I'll admit it: I find it considerably trickier to even pretend to wax insightful over what the latest numbers mean when I'm confronted with two distinct datasets, one of which is a week old. (I think I've just outed my allegiance in the aforementioned sample size vs. age debate, haven't I?) 

Anyway, the Conservatives retake the lead, and the Liberals drop to a distant second in BC, although they were back ahead of the NDP this week after a rather dizzying 12 point plummet the week before. Or is the latter set of numbers a more accurate gage of West Coastian sentiment, which would put the Greens in third place? It's all so confusing, like watching sausages travel through time. Let's move onto Alberta, shall we? Just to get our bearings, what with it being the great unchanging landscape of electoral certainty?

 

Alberta	Week 1 (June 22-29) (MoE 7.39)	Week 2 (June 30-July 6) (MoE 12.98))
Conservative	56.2 (+1.3)	61.3 (+5.1)
Liberal	15.4 (-2.9)	16.0 (+0.6)
NDP	15.0 (+6.5)	3.5 (-11.5)
Green	10.9 (-1.6)	15.9 (+5.0)
Other	2.5 (-3.3)	3.4 (+0.9)

See? Sure, there are a few jostles and bounces -- the Conservatives are over 60 percent! while the NDP is less than half the margin of error! Still, compared to some of the other regional results, the Wild Rose province remains the soul of consistency.

 

Saskatchewan/ Manitoba	Week 1 (June 22-29) (MoE 9.39)	Week 2 (June 30-July 6) (MoE 11.88)
Conservative	32.2 (-12.9)	44.5 (+12.3)
NDP	27.9 (+9.9)	18.0 (-9.9)
Liberal	23.4 (+0.7)	25.2 (+1.8)
Green	14.4 (+0.9)	12.4 (-2.0)
Other	2.1 (+1.4)	0.0 (-2.1)

Poor Saskitoba, on the other hand, is likely going to get its nose seriously out of joint under this new EKOS regime; those double digit downfalls-turned-recoveries that both the Conservatives would be so much more dramatic if spread out over two weeks. Instead, the upshot is no change here. Well, that's not quite true -- there's plenty of fleeting, ephemeral, Obama stump speechesque change, but the overall result is pretty much exactly the same as the last weekly poll. It's a little like soccer, really: the action is on the field, not on the scoreboard.

 

Ontario	Week 1 (June 22-29) (MoE 3.37)	Week 2 (June 30-July 6) (MoE 4.58))
Liberal	34.4 (-0.9)	26.4 (-8.0)
Conservative	32.2 (-0.1)	39.5 (+7.3)
NDP	18.1 (-0.1)	18.9 (+0.8)
Green	13.6 (+2.1)	13.3 (-0.3)
Other	1.7 (1.0)	2.0 (+0.3)

Now, see, this is where I actually think we can point to a trend and see, "Hmmm," in a thoughtful way, possibly while stroking our chins and furrowing our brows; the Liberals, it seems, have once again run out of moxie in Ontario. 

 Yes, I know the margin of error is higher than usual, but look at the trend -- we've seen this exact same scenario unfold more than once over the last year: The Liberals start to creep up in Ontario, and Conservative support deflates, until the former has an entirely acceptable 7 or 8 point lead, and then it just seems to collapse. Not all of a sudden, mind you -- it's usually over the course of a few weeks -- but the end result is always the same, although to be fair, I'm not sure when we last saw the Liberals fall quite so far below 30 percent in Ontario. 

I don't have an explanation for it, mind you, I just thought it was worth pointing out. (I'd tell you to feel free to share your theories in the comments, but somehow I have a feeling that those of you who have them won't need an invitation to do so, but consider it offered anyway.)

 

Quebec	Week 1 (June 22-29) (MoE 4.24)	Week 2 (June 30-July 6) (MoE 5.73)
Bloc Quebecois	38.5 (-0.4)	40.8 (+2.3)
Liberal	20.5 (+1.7)	21.4 (+0.9)
Conservative	14.9 (+0.3)	16.4 (+1.5)
NDP	13.0 (-0.1)	12.0 (-1.0)
Green	9.5 (-3.7)	6.5 (-3.0)
Other	3.6 (+2.2)	2.9 (-0.7)

Ahh, Quebec, the bellest of the provinces. Both the Liberals and the Conservatives are beginning to show signs of life -- or, given how often that's happened in the past, perhaps "undeadness" is the correct term, although just like the Liberals' Sisyphean struggles in Ontario, the Conservatives seem perpetually on the verge of breaking out of the mid-teens, only to find themselves abruptly plunging down an open manhole once again. 

On the other hand, the NDP now commands twice the support of the Greens -- which would, of course, be more impressive if the Greens weren't mired below the cruel but useful watermark that is the margin of error. Finally, the Bloc Quebecois -- well, they're fine, aren't they? They're always fine. It's their thing.

 

Atlantic Canada	Week 1 (June 22-29) (MoE 8.06)	Week 2 (June 30-July 6) (MoE 11.71)
Liberal	36.4 (+3.9)	24.8 (-11.6)
Conservative	34.4 (-3.4)	23.5 (-10.9)
NDP	17.7 (-1.1)	42.0 (+24.3)
Green	8.7 (-0.3)	7.4 (-1.3)
Other	2.8 (-1.0)	2.3 (-0.5)

At last, we make it to Atlantic Canada, where I'm having some trouble believing that the NDP have gone up by nearly 25 points over the course of five days, and are now at nearly twice the support of their closest rival. I mean, it's possible, I guess, but it seems more likely to be a statistical fluke. I guess we'll find out next week -- or, I should say, the week after, given the summer cycle.

UPDATE: So it seems that even the Conservatives -- who, by all accounts, should be dancing in the streets over numbers like this -- are reluctant to pop the champagne corks and break out the party hats just yet. Or, at least, so saith the party's InfoAlerteBot, which sent out the following caveat via Talking Point Express earlier today:

" ... While we do hold a lead over the Liberals, Caucus members should be aware that a 10-point lead is inconsistent with the findings of other pollsters and our own, internal assessments. When subsequent polls show different findings, this will NOT mean that Michael Ignatieff has suddenly gained momentum ..."

 So, what should send our eyebrows skyward over the next few weeks? Well, I'm not a trained pollologist, but after nearly two years of painstakingly recrunching EKOS data, here's what I'd consider worthy of note: