What do Trump, Clinton, Pallister, Couillard, and Trudeau have in common?

They are all leading in the polls!

- We'll start with the next election around the corner, the Democratic and Republican caucuses being held in Iowa on Monday. Hillary Clinton narrowly leads Bernie Sanders in the polls for the Democrats, while Donald Trump is narrowly ahead of Ted Cruz for the GOP. I wrote about the state of the race for the CBC here.

- In Manitoba, which votes in April, a new poll shows the Liberals are dropping in support, with the Progressive Conservatives taking advantage. The NDP still trails at a distance, tied with the Liberals. The poll is from Mainstreet Research and I also wrote about it for the CBC here.

- A poll by CROP conducted in Quebec flew under the radar, as the polling firm posted the results quietly to its website. Provincially, it shows the Liberals leading with the Parti Québécois dropping back. Both the CAQ and Québec Solidaire were up. Federally, the Liberals were way ahead of the other parties.

- And the latest federal numbers from EKOS show the Liberals enjoying a wide lead nationwide, with 47% support against 25% for the Conservatives and 16% for the NDP.

How this NDP surge is different from the last one

The projection has been updated, including the two most recent polls from EKOS Research (for iPolitics) and Ipsos Reid (for Global News). Both polls show a remarkable three-way race, with EKOS gauging the numbers as 30% for the Conservatives, 29% for the NDP, and 27% for the Liberals, and Ipsos as 31% for the Conservatives, 31% for the Liberals, and 30% for the NDP.

The Ipsos poll serves as a much-needed confirmation of the EKOS numbers, which had been virtually alone since the Alberta election of May 5.

The projection still lags the polls a little, in part because May has been a quiet month (10 polls had been done in April, while we're up to seven for May). Polls that show a three-way race currently take up 61% of the projection. But the NDP has averaged 29.5% in those polls, whereas they have averaged just around 23% in the remaining polls taking up 39% of the weight.

Nevertheless, the NDP's growth has been remarkable. This is especially so when we consider that the model, at this stage, is designed to react slowly to what could be blips in voting intentions. Since the week ending on May 4, the New Democrats have jumped from 22.9% to 27.1%. Both the Conservatives and the Liberals have dropped 1.3 points over that time.

The Conservatives were leading in the seat projection then with between 125 to 164 seats against 93-129 for the Liberals and 65-91 for the NDP. That count currently stands at 118-153 for the Conservatives, 77-120 for the Liberals, and 87-118 for the NDP.

The New Democrats have picked up 5.5 points over that time in British Columbia, 2.5 points in Alberta, 4.5 points in the Prairies, four points in Ontario, 5.8 points in Quebec, and 0.8 points in Atlantic Canada. Interestingly, though, neither the Conservatives nor the Liberals have been the sole relinquisher of votes. The Liberals have suffered the bulk of the losses in Alberta, the Prairies, Ontario, and Atlantic Canada, but the Conservatives have slid most in British Columbia and Quebec.

A recent piece by the Canadian Press suggested that this latest surge was the first to happen for the NDP since the late 1980s, making it difficult to make comparisons. Of course, as readers of this site know, the NDP has surged much more recently than that. The party was averaging about 35% in May and June 2012, putting it in first place nationwide. So, we do have something to compare the current uptick to. Let's do that.

What I've done below is taken the two most recent polls from EKOS and Ipsos and compared them to the polls from these two firms in late June 2012. At the time, Ipsos had the NDP ahead with 38% to 35% for the Conservatives and 18% for the Liberals. EKOS put the NDP at 32% to 29% for the Conservatives and 19% for the Liberals.

For both Ipsos and EKOS, the NDP was doing better in 2012 than they are now. We should expect their numbers to be better at the regional level as well. But that is not so. The NDP's electorate has shifted since the last time the party was popular.

First, the good news for the NDP. It has made some major gains in Alberta. Both Ipsos and EKOS put the NDP at 13% in the province at a time when the party was doing better nationwide. Now, they have the NDP at 25% to 31%, meaning the New Democrats are roughly doubling their support from 2012. That extra support means a boost of around two points nationwide. Without it, the race would not be as close.

The NDP is doing slightly better in Quebec now than it was in 2012 meaning that, proportionately speaking, its Quebec support is far more significant to the party than it was three years ago.

Two areas have mixed news. According to Ipsos, the NDP is doing about as well in British Columbia as it was doing in 2012, while EKOS has the NDP doing worse. The opposite is the case in the Prairies. These two regions, then, are inconclusive.

But the party is doing significantly worse now in Atlantic Canada than it was doing in 2012. And more importantly, the New Democrats have not made the same inroads into Ontario that they had during Thomas Mulcair's honeymoon period.

This is important, because the gains the New Democrats have made in Alberta will not translate into a lot of new seats. They were much better positioned, with their strong numbers in Ontario, in 2012. The New Democrats have put together a bit of a West + Quebec coalition that aligns with their 2011 breakthrough in Quebec and their traditional strength in B.C. and the Prairies. In 2012, however, they had added Atlantic Canada and Ontario to that coalition.

This is why we're looking at a three-way race instead of a two-headed contest. The Liberals are doing much better in Atlantic Canada and Ontario than they were in 2012, putting them in contention. The Conservatives have their western bases as well as a good chunk of Ontario, giving them the edge in seats. The NDP can now threaten to win a plurality thanks to inroads in B.C., as well as a handful of new seats in the Prairies, in addition to their retention of Quebec.

It makes for an interesting regionalized race between three parties. The contest is between the Conservatives and the Liberals in Manitoba, Ontario, and parts of British Columbia (Vancouver suburbs) and Atlantic Canada (the Maritimes). It is between the Conservatives and the New Democrats in Alberta, Saskatchewan, and parts of British Columbia (the Interior) and Ontario (the southwest and north). And it is between the Liberals and the NDP in Quebec and parts of British Columbia (Vancouver), Ontario (Toronto), and Atlantic Canada (Halifax and Newfoundland). There are plenty of smaller local battles that include other match-ups on Vancouver Island, in Calgary, in Winnipeg, in Quebec City, and in central Quebec.

A swirling mess of a campaign that could produce virtually any result. Will the summer clarify things?

Three-way federal race?

This week's EKOS poll, with the preliminary results leaked to the Ottawa Citizen before EKOS was forced to publish their final results a day early yesterday, has a lot of tongues wagging in Ottawa. Why? The poll put the New Democrats right in the middle of a three-way race, with the Liberals bringing up the rear. Are we in the midst of a re-alignment, or is this a blip?

This far out from the election, the projection still turns like a big ship (as it should at this stage). Nevertheless, it has recorded a sharp uptick in NDP fortunes with the addition of this new EKOS poll (as well as an Insights West poll for British Columbia, which had similarly good news for the NDP, and the previous week's EKOS poll).

The Conservatives still lead with just under 32% support, enough to give them between 121 and 159 seats. The Liberals are in second with 30% and between 87 and 125 seats, while the NDP is in third at 24% and between 74 and 101 seats.

Both the Tories and Liberals have dropped since last week's update. The Conservative low range has fallen by eight seats from 129, and its high range has dropped five seats from 164. The majority mark is now 11 seats away at the high end, rather than just six last week.

The Liberals' seat range has moved less, falling from 94-128 last week. But most significantly is that the NDP's likely range now overlaps with the Liberals. Last week, the NDP topped out at 88 seats, six below the Liberals' low range. Now, the NDP tops out at 101 seats, 14 seats above the Liberals' low range. This is a big shift, and it could be the start of something even bigger.

There was word from Twitter that a Forum poll also showing a three-way race is forthcoming, and I have heard of an internal poll showing the same thing (take this with caution, though, since there could be plenty more internal polls showing no such thing that I have not heard about). Coupled with the Insights West poll in British Columbia, a Forum poll in Ontario showing gains for the provincial NDP (due to Patrick Brown winning the PC leadership, Forum implausibly explains), and the party's big win in Alberta last week, things are looking up for the New Democrats.

We'll only know if this is a momentary blip or something enduring in the coming weeks or months. If it is a blip, we'll know sooner rather than later.

But let's take a look at this EKOS poll in detail.

The Conservatives led with 30% support, virtually unchanged from where the party stood in EKOS's polling the previous week.

The New Democrats were up a remarkable 4.9 points, vaulting them into second place with 29.1% support. The Liberals were down 2.6 points to 27%, though that shift was just inside the margin of error.

The Greens were at 8% and the Bloc Québécois at 4.1%, while support for other parties was at 1.8%.

That last number is noteworthy, since EKOS's previous poll from April 29-May 5 had support for the Others at 4.5%. That was an unreasonably high number, and for EKOS to now have them below 2% is certainly out of character. I'm not sure why that happened this week, but it freed up 2.7 points that seems to have gone to partly to the Greens and mostly to the NDP.

This poll sets off alarm bells because of the dramatic shift it is highlighting - not in the sense that it makes the poll unreliable, but in the sense that we need to sit up and pay attention to what is happening. Is it an outlier or the first poll to catch a new trend?

As it marks the first time the Liberals have placed in third in any poll since March 2013, over two years ago, it definitely stands out. And this is indeed a bad poll for the Liberals, as they lead only in Atlantic Canada. Their scores were low in Quebec and Alberta. They were also low in Ontario, though so were the Conservatives.

The Tories certainly aren't coming anywhere close to a majority government with 33% support in Ontario and 29% in British Columbia. EKOS also seems to be pointing towards a deflating Tory balloon in Quebec. Something to watch as well.

The NDP numbers were only unusually good in Ontario and Quebec. The party has not held such a wide lead in Quebec for some time, and being in a close three-way race in Ontario is something new. The party was already competitive in B.C. and doing well in Alberta, so those are well within the norm. It is the surge in Ontario and Quebec that is behind this second place finish for the NDP, so we need to keep an eye on the NDP's numbers in these two provinces over the next few weeks.

But if these numbers became reality on election day, the House of Commons would look like nothing we've ever seen before. The Conservatives would win about 131 seats, the NDP would take 110, and the Liberals would win 94, with the Greens and Bloc winning the remaining three seats.

It would be close enough that, if these were what the polls were showing the day before the election, the model would project 113 to 153 seats for the Tories, 98 to 131 for the NDP, and 71 to 108 for the Liberals. That means the NDP would, potentially, be able to win a plurality of seats.

Is this where we are heading? This is such a large change in the political landscape that we need to stand back and exercise a little patience. Maybe these numbers are completely real, and we're now in a three-way race. Maybe these numbers are just a bout of post-Alberta zeal that will fade away in a matter of weeks. Maybe no polls aside from EKOS and the (apparently) forthcoming Forum will show these kinds of results, and we'll return to the status quo. For now, we just don't know. Watch this space!

Alberta NDP maintaining lead

21h10 Update - Final (?) poll of the campaign confirms NDP lead, Notley advantage

What could be the last poll of the campaign was out this evening, from Insights West. Like the other polls, it gives the NDP a very comfortable lead over Wildrose and the PCs, though it does suggest the race in Calgary could be close enough to force some tight contests. But what is most revealing may be the leadership numbers.

Insights West was last in the field between March 27-30, so before the campaign began. That doesn't make it useful from a trends perspective.

The poll gave the NDP 42% support, with Wildrose trailing at 27% and the PCs at 23%. This falls well into line with the other polls we have seen (below).

the poll awarded 4% to the Liberals, 3% to the Alberta Party, and 2% to other parties. A total of 9% of the sample was undecided.

As the poll was conducted online, no margin of error applies. See the link above for full details on the questions asked.

The NDP led in all three regions of the province. In Edmonton, it had 58% support to 16% for both the PCs and Wildrose.

In Calgary, the NDP had 35% support, against 30% for Wildrose and 27% for the PCs. Of the three polls released since yesterday, this is the best one for both the Tories and Wildrose here.

In the rest of the province, the NDP had 35% with Wildrose at 32% and the PCs at 25%.

Fully 82% of respondents agreed it was time for a change of government, while just 35% agreed that they would be upset if the NDP won.

This makes the current climate very different from the 2012 Alberta or 2013 B.C. elections. Rachel Notley is much more popular than either Danielle Smith or Adrian Dix were in those campaigns.

Insights West found that Notley's approval rating was an astounding 62%, with just 21% of Albertans disapproving of her. Apart from Jim Prentice, she scored the lowest on the number of 'unsures'.

Prentice had an approval rating of just 25%, and he was the only leader with a majority disapproval rating (63%).

The other leaders had mixed ratings, with Brian Jean posting a 35% to 40% split, David Swann having an approval rating of 33% to 36% disapproval, and Greg Clark splitting the question 20% to 23%. A majority of Albertans did not have an opinion about the Alberta Party leader.

And just like other polls have shown, opinion has deeply soured on Prentice. Just 6% of Albertans said their opinion of him had improved during the campaign, compared to 52% who said it worsened. By contrast, 52% of Albertans said their opinion of Notley improved, while just 11% said it worsened. These are very consistent numbers with what we've seen from ThinkHQ and Léger last week.

Notley beat Prentice by a wide margin on being the better leader on health care, the environment, education, crime, housing and poverty, and government accountability. She was narrowly ahead, 28% to 24%, on the economy and jobs, and narrowly behind, 23% to 27%, on energy and pipelines. This is a little asterisk to keep in mind tomorrow night.

But on who would make the best premier, Notley scored 37% to just 18% for Prentice and 13% for Jean. Christy Clark and Alison Redford did not trail so far, nor were their approval ratings as low as Prentice's, the day before their victories.

16h18 Update - EKOS poll still shows NDP in front, lead growing

EKOS, in the field as recently as yesterday, adds to the trend of the Alberta New Democrats holding their own and, seemingly, in little danger of a last minute slide.

EKOS's previous poll was last in the field between April 25 and 29.

Since that survey, the NDP has picked up 2.1 points to move further into the lead with 44.3% support. Wildrose was up 2.7 points to 24%, while the PCs were down 0.6 points to 22.5%.

All margin of error shifts, but certainly arguing against a softening of NDP support and a rebound by the Tories. They are almost 22 points behind the NDP. How plausible is it really to still consider this a three-way race? We wouldn't in any other election.

The Liberals were down 0.7 points to 5.6%, while the Alberta Party was down 2.4 points to 2.2%. Interestingly for them, their support dwindled to almost nothing everywhere but Calgary, where their leader Greg Clark does have a shot (the projection currently awards his party a seat).

In battleground Calgary (sarcasm), the New Democrats were up to 40.5%, followed by the PCs, down to 22.3%, and Wildrose, down to 22.2%.

The NDP slipped slightly in Edmonton to 58.6%, followed by the PCs at 18.8% and Wildrose, which jumped to 17.7%.

In the rest of the province, the NDP was up to 35.5%, followed by Wildrose at 31.1% and the PCs at 25.9%. Both parties were up slightly as well.

These regional numbers are very similar to what Forum found in its May 2 poll below.

09h53 Update - Alberta NDP maintaing lead

One of the biggest polling issues with the 2012 Alberta election campaign is that too many pollsters dropped out of the field too early. We'll never know if they would have captured anything different had firms like Léger, ThinkHQ, or Abacus Data stayed in the field over that final weekend.

Luckily, it seems that some pollsters are not going to make the same mistake this time. The last round of polls dropped out of the field on Wednesday, leaving Albertans almost week to mull things over. The new round of polls has continued in the field throughout the weekend, and might even include some interviews from today.

Two post-April 29 poll have already been published or are on their way. One, conducted by Forum Research and published yesterday on the Edmonton Journal website, was in the field on Saturday. The other is forthcoming from EKOS Research. Frank Graves hinted at his numbers yesterday on Twitter, and they showed the NDP's lead growing - not shrinking.

Forum was last in the field on April 23, when it was the first mainstream pollster to see the NDP moving into majority territory.

The NDP led in this new poll with 42%, up four points from the previous survey. Wildrose was down one point to 24%, while the Progressive Conservatives were up one point to 21%.

The Liberals were down two points to 5%, while the Alberta Party was down one, also to 5%.

Another 3% said they would vote for another party (down two points).

None of these shifts were outside the margin of error, though the results look a lot like the polls we saw last week.

The numbers argue forcefully that the NDP's lead is going nowhere. There was much speculation that the round of polls published on Thursday and Friday showing the NDP in majority territory would have an effect on people's voting intentions. Instead, it seems to have had no effect. If the PCs were going to stage a comeback, we'd see it somewhere in the trends. Trailing by 21 points is not something that can be overcome in a matter of days. Even Wildrose's lead in 2012 was just eight points, on average, in the final week.

The New Democrats led in the poll in Edmonton with 55%, followed by the Tories at 19% and the Wildrose at 13%. These numbers were mostly unchanged.

The NDP was also ahead in Calgary, moving up to 37%. They were followed by Wildrose at 26% and the PCs at 21%, both of which were stagnant. Even Calgary might not turn out to be the three-way race many people are expecting.

And the NDP also led in northern Alberta, with 38% to 30% for Wildrose and 24% for the PCs, and central Alberta, with 32% to 29% for Wildrose and 27% for the PCs.

Wildrose led only in southern Alberta, with 42% to 37% for the NDP and 16% for the PCs.

Again, we'll have to wait and see what happens on Tuesday. But a lot of the prognosticators saying that the NDP doesn't have the map to win a majority government are ignoring these sorts of numbers. In this poll (which admittedly has small regional samples, but which is in line with others) the NDP has at least 32% outside of the two main cities and is ahead by double-digits in Calgary. If those numbers hold, then they most certainly have the math to win a majority government. These things can happen - would anyone really have considered the NDP a lock to win all of conservative Quebec City's ridings in the last federal election? If the NDP can win Quebec City, it can win Calgary.

An NDP majority in Alberta?

Yesterday was poll day in Alberta. First Return on Insight and Léger reported in the morning, then ThinkHQ surprised everyone with some new numbers. EKOS put their poll out in the early afternoon, and finally Ipsos Reid weighed in. Add this morning's Mainstreet poll, and we've seen six surveys in the last 24 hours.

While that would normally be unusual enough, as each poll landed the skepticism people had that the New Democrats were leading in Alberta was chipped away. Of those six polls, not a single one gave the NDP less than 37% support, and only one put the gap between the New Democrats and the second place party in single digits. The others showed a lead for the NDP of between 11 and 19 points. In Alberta.

The projection has now lurched from a largely minority situation between either the NDP or Wildrose to a likely majority victory for the New Democrats. The party leads with a projected 42% of the vote, enough to win them between 43 and 58 seats. Their minimum seat haul is projected to be 25 seats - which would count as a historic best.

Wildrose is narrowly in second with 26% and between 16 and 28 seats, while the Progressive Conservatives stand at 25% and between five and 23 seats.

The Liberals, with two to three seats and 5% of the vote, bring up the rear.

Yes, these are shocking numbers. But a few things should be kept in mind.

First, a lot of the seats the NDP needs to win to get to majority status are marginal. Seven of the 52 seats projected to go to the NDP are projected with less than 60% confidence - a toss-up. We cannot discount the importance of incumbency, and the potential for a snap-back as Albertans consider the prospects of an NDP majority (they could, of course, also like the idea).

Second, the NDP is projected to take slightly more than the average of the most recent six polls because of the projected over-estimation of the Liberal vote (as well as for the Alberta Party). This is due to the number of candidates these parties are fielding. The polls are averaging 7.6% support for the Liberals, which the projection reduces to 4.7%. They are averaging 4.7% for the other parties, including the Alberta Party, which the projection reduces to 2.7%.

As the New Democrats are topping the polls, and those extra votes are distributed proportionally, they get the lion's share. But the polls back this up. The second choice preference for Liberals and supporters of the Alberta Party is the NDP by a wide margin.

We'll see what the weekend and Monday has in store for final polls. It will be very interesting to see if these trends continue, or if they shift. If they do not shift, the NDP could still have problems overcoming their incumbency disadvantage and a lack of a ground game outside Edmonton and a handful of other ridings. However, the NDP didn't have any trouble with that in Quebec in 2011. Is this a wave election?

The tide is definitely shifting in the four polls released since yesterday morning.

Only Mainstreet was a returnee to the field, and showed a remarkable swing to the New Democrats. From 31% in the immediate aftermath of the debate, the NDP jumped to 44%. Wildrose fell six points to 26%, while the Tories were down five points to 21%.

Overall, the polls were very consistent. The NDP was between 37% and 42% in the three polls done over several days between April 25 and April 29. Three of the four polls had Wildrose at either 26% to 27%, while the PCs were in a tight range of between 20% and 24%.

That is a great deal of agreement for such an oddball election. We also have different methodologies at play here, with EKOS and Mainstreet using IVR and ThinkHQ and Ipsos Reid using both live-callers and the internet. Return on Insight used just live-callers, while Léger used their online panel.

And when we take into account the margins of error (assuming a probabilistic sample for argument's sake), we see there is even more agreement.

With the exception of the Mainstreet poll, which is the newest, a range of between 38% and 41% fits within the margin of error of all of the polls. They all top out at 41% or higher.

There is less agreement for Wildrose and the PCs, but the numbers are nevertheless consistent: mid-20s for Wildrose, low-to-mid-20s for the Tories.

The polls also saw eye-to-eye at the regional level, capturing some startling results.

The New Democrats, of course, led in every poll in Edmonton, with the PCs running second and Wildrose third. Looking at all six, the NDP ranged between 53% and 73%, the Tories between 16% and 22%, and Wildrose between 7% and 17%.

Most surprising is Calgary, though. I have read a great deal of skepticism surrounding the NDP's chances in the city, but the polls really can't back that up. The NDP finished in first place in the city in four of the six polls, and in second in the remaining two. They ranged between 25% and 35% (only one poll had them lower than 29%), compared to 23% to 33% for the Tories and 24% to 28% for Wildrose.

The New Democrats also led in four of the six polls in the rest of Alberta, with the PCs finishing third in five of the six. The NDP ranged between 32% and 39%, Wildrose between 28% and 33%, and the PCs between 22% and 35% (and only one poll had the PCs higher than 23%, which is disastrous for the party).

ThinkHQ, which had a more detailed breakdown for the area outside of Calgary and Edmonton, put the NDP ahead in northern Alberta, Wildrose ahead in central Alberta, and the two parties neck-and-neck in southern Alberta. The PCs weren't competitive anywhere.

Despite all of this, Ipsos found that 50% of Albertans still think the Tories will win, with just 21% thinking the NDP will win. Mainstreet, slightly newer, found it more of a split: 32% thinking the PCs will win and 29% giving the nod to the NDP. It would be interesting to see how that might change in the coming days.

And what effect might it have? It could be exaggerated. ThinkHQ found more Albertans concerned about a PC victory than either a Wildrose or NDP win.

By the election, there will have been five full days of campaigning between the most recent poll and the vote. That makes for a lot of time for opinion to shift. It shifted in that amount of time in 2012, so hopefully we won't be left in the dark again. Because something is happening in Alberta - and missing a whole week of it could be a big mistake for the pollsters.

Uptick in Alberta for Mulcair's NDP?

The federal projection has been updated, incorporating the latest polling data from EKOS Research, Léger, and CROP. The overall results have not shifted much, though the Conservative seat count has dipped to the benefit of the Liberals. You can see all the details here.

An interesting development does seem to be underway in Alberta, however, where the New Democrats are polling at their highest level in a year. Is the NDP on the upswing in the province for real?

First, let's briefly go over the new polls added to the model.

In EKOS's poll for iPolitics, the Liberals moved ahead with a gain of 1.6 points since the previous week to 32.1%. The Conservatives dropped 2.7 points to 30.2%, while the NDP was up 1.5 points to 21.2%.

Only the Conservative decrease was outside the margin of error, though the Liberal bump does end a losing streak for the party. Overall, however, the poll is well within EKOS's usual range.

The Léger poll reported by CTV Montreal showed a similarly close race, with the Liberals at 35%, the Conservatives at 34%, and the NDP at 20%. Compared to Léger's previous poll of Jan. 30 to Feb. 2, the Liberals held firm, the Tories were up two, and the NDP was unchanged. Suffice to say, those shifts would not be outside the margin of error of similarly sized probabilistic samples.

Finally, the poll by CROP for La Presse gave the New Democrats the lead in Quebec with 30%, unchanged from CROP's previous two surveys in the province. The Liberals dropped four points to 29%, while the Conservatives were up two points to 18% and the Bloc Québécois was up one point to 18%.

None of these shifts would be statistically significant. Noteworthy, however, is that the Liberals have dropped in two consecutive surveys (from 37%) while the Conservatives have increased their support over that time (from 13%). Also note that the results from CROP are broadly in agreement with those from Léger for the province.

Alberta turning a little orange

The longest and largest streak (both in real terms and proportionately) in the country currently belongs to the New Democrats in Alberta. Their numbers have been increasing over five consecutive weeks in the province, from 11.9% at the beginning of February to 16.3% in today's update. That is a gain worth 4.4 points, or an increase of more than one-third since that low ebb.

This upswing for the NDP in Alberta has coincided with a period of Conservative decline, as the party has slipped 4.2 points over the last five weeks from 54.7% to 50.5%.

(The third-longest streak at the moment is in Quebec, where the Liberals have fallen over four consecutive weeks from 30.3% to 27.7%.)

This jump for the NDP has pushed the party to two seats in the projection (Edmonton Strathcona and Edmonton Griesbach, though the latter is by a very slim margin). It has also put the New Democrats in play in Lethbridge.

The shift has been registered in most polls. In the two newest surveys, Léger has the NDP up five points in the province and EKOS has them up almost 10 points since the poll it conducted at the same time as Léger's last outing. Over the last eight polls, the NDP has managed between 13% and 19%. In the previous eight, that range was from 10% to 17%.

Whether or not the increase has come directly from the Conservatives is harder to determine. In those same most recent eight polls, the Conservatives have ranged between 49% and 54%. In the previous eight, that range was 47% to 61%. That suggests the Tories' numbers have just become more volatile, rather than necessarily dropping in the face of NDP gains.

The reality is rarely ever linear, and in this instance we may be seeing voters heading in every direction, the net result being an NDP uptick.

But why would this be taking place? We cannot definitively chalk it up to an improved performance by Thomas Mulcair, whose personal numbers in Alberta show no such clear trend line as those of his party.

We instead might be seeing a bit of cross-pollination between the provincial and federal scenes. Talk in Alberta is about the possibility of the New Democrats, under new leader Rachel Notley, taking over from Wildrose as the Official Opposition after the next election. The better press for the provincial NDP may be having a positive impact on the federal party.

The real question, as is always the case, is whether or not these numbers will prove to be sustainable or a mere flash in the pan. In the grand scheme of things, it does not have much impact on the national portrait. But with the race looking as close as it is, even losing an extra seat or two to the NDP in Alberta can have an important effect on the Conservatives' electoral calculations.

Stability in the polls, or Liberal decline?

At the moment, the only pollster giving us regular updates on the political situation in Canada is EKOS Research. The firm is reporting every week on Fridays via iPolitics. This far out from the election, the week-to-week changes have been usually quite marginal. But this may be masking a wider trend that is not a good one for the Liberals.

The latest numbers from EKOS put the Conservatives ahead with 32.9% support, an increase of 0.7 points since the previous week of polling.

The Liberals were down 0.8 points to 30.5%, while the New Democrats slipped 0.2 points to 19.7%.

The Greens were down 0.2 points to 8.8% and the Bloc was up 0.2 points to 4.8%, while 3.2% of respondents said they would vote for another party.

None of these variations were outside the margin of error. Also note that this poll has not yet been added to the projection.

EKOS's reports contain tracking data stretching back to the beginning of 2012. That gives the numbers good context, but makes it difficult to look at the week-to-week trends the pollster is recording. So let's focus on that.

When we look at the polls over the last few weeks, it is difficult to discern any serious trends.

The Conservatives seemed to be trending downwards slightly, before up-ticking over the last two polls. EKOS's method of reporting to the first decimal point makes it look like there is more movement than there really is - if we round off those numbers as other pollsters do, the Conservatives have polled at 35%, 32%, 32%, 31%, 32%, 33%. Really not much movement there, and we can't even find a statistically significant shift between their current 32.9% and their high and low marks since EKOS began reporting weekly.

Not so for the Liberals, however. Their trend line is slightly negative starting with the poll ending on February 10, when the party was at 33.8% support. The drop of 3.3 points since then is outside of the margin of error. We could indeed be looking at an actual decrease in Liberal support over the last few weeks.

The issue, though, is that it is difficult to corroborate this with other polls as there haven't been many that have been conducted over the same time period as EKOS. What we do see is that Abacus Data recorded an uptick in Liberal support worth two points between its polls of January 26-28 and February 12-16, while Nanos Research had the Liberals down two points between its four-week rolling polls ending February 20 and February 27.

That the Conservatives are stable seems to be the consensus opinion of the polls. The question appears to be whether or not the Liberals are on the downswing.

In terms of seats, the Conservatives would win about 148 with these levels of support, with the Liberals taking 120 and the New Democrats capturing 54. The Bloc Québécois, benefiting from low NDP support in Quebec, could win about 14 seats despite having dropped more than three points in support since 2011. The Greens would win two seats.

Outlier British Columbia?

Back to the EKOS poll itself, the only regional result worth looking at is British Columbia. Last week's EKOS poll had the Greens at 21%, and solicited some excitement on the part of Greens in the province. It also happened to have the Conservatives at just 23%.

Now, we see the Conservatives jumping 12 points to 35% and the Greens falling eight points to 13%. Both of these shifts are outside the margin of error, but it looks to me like a reset (and maybe an over-correction) rather than an actual shift in voting intentions.

Why? Take a look at the Conservatives' results in B.C. since the beginning of EKOS's weekly polling: 

Feb. 3: 27%

Feb. 10: 29%

Feb. 17: 28%

Feb. 24: 31%

Mar. 3: 23%

Mar. 10: 35% 

The result of March 3 certainly stands out from the pack. The March 10 result also stands out, but seems much more intuitive and matches some of the results other pollsters have recorded in British Columbia. It does correspond with weaker than usual numbers for the Liberals (23%, rather than the 28% to 33% recorded in earlier polls), so we could see another reset next week with the Liberals up.

Now let's look at the Greens in B.C.:

Feb. 3: 16%

Feb. 10: 16%

Feb. 17: 15%

Feb. 24: 12%

Mar. 3: 21%

Mar. 10: 13%

Here again, the March 3 result stands out rather dramatically. The conclusion we should reach is that, as far as B.C. was concerned, the March 3 poll was a bit of an outlier. Nothing wrong with that as it should happen from time to time, but it was a lot of fuss for not much at all.

Party divisions over ISIS mission

Moving on, I though it would be interesting to note EKOS's results on support for the mission against ISIS. Overall, 60% of Canadians support the mission while 33% oppose it. But the party breakdowns shed some light on the politics of it all.

The Conservatives are safest in their robust support for the mission, as 80% of their voters are in favour of the mission. Only 16% aren't.

The New Democrats are better positioned to be the voice of opposition, as 45% of their voters do not support the ISIS mission. Nevertheless, 54% do. It is a damned if you do, damned if you don't situation for Thomas Mulcair, but his opposition to the mission fits within the historical position of the New Democrats. 

And the Liberals off-again-on-again relationship with the mission can perhaps be best explained by these numbers: 38% of their voters oppose the mission, while 60% support it. The party has not seemed quite comfortable with the position it took in September, and their supporters seem to agree. It will be interesting to see whether the Liberals will walk it back when the vote for renewal comes up. 

Liberal recovery or blip?

The projection was updated yesterday, reversing the first instance of the Conservatives leading in both the vote and seat projection. The Liberals are now back on top in the polling average, and have closed the gap in the seat count to just five. Does this mean the Liberal slide has come to a halt, or is it just a blip?

The Liberals lead again with 34.2%, up a point from the last update incorporating polls done to Feb. 3. The Conservatives dropped 0.6 points to 32.9%, while the NDP was down 0.7 points to 19.3%.

In terms of seats, the Conservatives have dropped three to 140, and their range has gone from 128-162 to 125-158. The Liberals have increased their seat tally by 13 to 135, and their range from 106-140 to 114-152.

The New Democrats have dropped 12 in the seat count, and their range has fallen from 53-81 to 43-77. This was due primarily to a drop in Quebec, where the party is down seven seats from the Feb. 3 update.

The bulk of the Liberals' gain, however, came in Ontario, where the gap between them and the Conservatives widened from 0.5 points to 2.9.

Let's look at the two polls that prompted the update. They were by EKOS Research for iPolitics and Forum Research for the Toronto Star. Both left the field on Feb. 10, but the EKOS poll had a longer tail.

Forum was last in the field from Jan. 27-28, and recorded a relatively large shift in voting intentions since then. The Liberals were up five points to 39%, re-taking the lead from the Conservatives, who were down three points to 32%. The NDP was down three points as well to just 17%.

The increase in Liberal support is outside the margin of error, whereas the drop for the Conservatives and NDP is not. But before Liberals get too excited, consider their recent trend in Forum polling: 36% - 41% - 37% - 34% - 39%. On average, Forum has had the Liberals wobbling by about four points from one poll to the next. This would suggest to me that this bump in Liberal support is more house effect than anything real.

That is, at least in terms of its size. EKOS also recorded a modest rebound by the Liberals, which argues that perhaps the party is indeed recovering from their recent slide.

EKOS was last in the field Jan. 28-Feb. 3, and put the Liberals up 1.6 points to 33.8%. The Conservatives were down three points to 32%, while the NDP was up one point to 18.9%.

Only the Conservative drop was outside the margin of error, but both Forum and EKOS recorded similar trends for the Tories and Liberals. For EKOS, this poll halts a series of surveys showing Liberal/NDP decline and Conservative growth.

Can we reconcile the two polls? One wonders whether we should bother, considering Forum's history of volatility. But if we extend the polls' results to their respective margins of error, we get the following:

As you can see, the ranges for the Conservatives and NDP overlap quite comfortably. The ranges for the Greens do not, and the Liberals only meet at 36%. I do not think that this is coincidental. The Greens are generally higher in EKOS's polls than in any others, while the Liberals are generally higher in Forum's.

The two polls suggest that the shift in voting intentions may have occurred among men. Forum recorded a jump of six points for the Liberals among men, outside the margin of error, while EKOS also showed a statistically significant jump of 4.6 points for the party. Both showed the Conservatives sliding by four to six points. Together, it puts the Liberals at either 34% or 40% and the Conservatives between 34% and 35% among this demographic. There was no similarly consistent shift among female voters.

One anomaly of the Forum poll was in British Columbia, where the Liberals were up 15 points to 47%, the Conservatives were down 11 points to 24%, the NDP was up six to 23%, and the Greens were down nine to 5%. That is a lot of movement, particularly in the context of EKOS finding no party shifting by more than a handful of points. EKOS put the parties at 31%, 29%, 21%, and 16%, respectively.

In Quebec, the province everyone is keeping an eye on, both Forum and EKOS showed a bit of a reversal for the Conservatives. Forum had the party down to 21% and EKOS had them down to 23%, though both of those shifts were inside the margin of error. They are both still, however, relatively high numbers for the party in Quebec. The Liberals appeared to take advantage in both polls, Forum recording a significant jump of eight points (but EKOS just two).

The Liberals increase nationwide coincided with a drop in Justin Trudeau's disapproval rating, from 44% to 38%. His approval rating was up slightly to 48%.

The Conservatives' decrease, meanwhile, coincided with worsening numbers for Stephen Harper. His approval rating was down four points to 36%, while his disapproval rating was up five points to 56%, the worst numbers EKOS has recorded since November. This may not be too worrying for the Prime Minister, though, as his approval rating was up in Ontario and down in the West, where he is at little risk.

Thomas Mulcair's numbers were fairly steady, at 51% approval to 36% disapproval. Regionally, his numbers improved west of Quebec and worsened east of Ontario.

It would be helpful to have a third (and fourth?) poll weigh-in on this, preferably one which uses a different mode of contact. Has the Liberal slide been reversed? Is the Conservatives' momentum coming to a halt? Or will this prove to be just a pause in the trends that have been building for the last few months? We'll see.

Polls show Liberals still leading, but by how much?

Two polls published over the last two days show the Liberals still leading in national voting intentions. But the surveys by Léger and EKOS Research differ significantly over the size of that lead - as well as whether it is shrinking or growing.

First, the numbers themselves.

Léger, in the field between December 8 and 11 and reporting for Le Devoir, found the Liberals to have 38% support, up two points from where the party stood in Léger's previous survey of November 3-6.

The Conservatives were down one point to 32%, while the New Democrats were unchanged at 19%.

Support for the Greens was down one point to 6%, the Bloc Québécois was up one point to 4%, and support for other parties was unchanged at 1%.

None of these shifts would be outside the margin of error of similarly sized probabilistic samples.

EKOS, in the field between December 4 and 12 for their new poll published by iPolitics, gave the Liberals 31.8% support, down 1.7 points from the previous survey of November 4-6.

The Conservatives were up 0.6 points to 30.8% and the New Democrats were down 0.5 points to 20%.

The Greens were down 1.9 points to 7.8%, the Bloc was up 1.5 points to 5%, and support for other parties was up two points to 4.6%.

Only the shifts for the Greens and others would be outside the margin of error.

The differences between these two polls are well within the margin of error for the Conservatives, New Democrats, and Bloc, and just about within the margin of error for the Greens. So those discrepancies can be easily explained away by standard sampling error.

But the differences between the Liberals and support for other parties is harder to explain. About three points of the discrepancy between EKOS and Léger for the Liberals cannot be explained by the margin of error, and about 2.6 points for 'other' parties. That those numbers are very close may not be coincidental, as EKOS routinely has support for other parties to be significantly higher than other firms. Throughout 2014, for example, 'others' have averaged 3.3% support in EKOS's IVR polling, compared to 1.1% for everyone else.

Discrepancies like these, however, can often be chalked up to methodological differences. EKOS is conducting its poll via interactive voice response, whereas Léger did its poll online. The two modes of contact may inherently produce slightly different results. The focus, then, should be on the trends instead.

Here there is a little more clarity. Over the last three Léger polls, all three major parties have been very steady. The Liberals went from 37% in October to 36% in November, before ending up at 38% here. That is a statistically insignificant wobble. The Conservative numbers have been 32%, 33%, and 32% - very stable. And the NDP has registered 21%, 19%, and 19% support over that time.

It is a little more complicated with EKOS. In the time span that Léger did those three polls, EKOS has done four. Three of them were on virtually the same dates, but the first was done online. The last three were done via IVR. It wouldn't be appropriate to look at the trends by mashing together those two very different modes of contact.

So if we just look at the IVR polls by EKOS conducted since the end of October, we see similar stability to Léger: 33%, 34%, and 32% for the Liberals, 28%, 30%, and 31% for the Conservatives, and 21%, 21%, and 20% for the NDP. Trend-wise, at least, Léger and EKOS are not in complete disagreement.

But Léger is showing a very different sort of margin between the Conservatives and Liberals, growing from three points to six points in this poll. Considering the stability in the numbers, we could say that the margin is simply moving back and forth with the wobble. But EKOS's trend is more towards convergence, with the gap being about five points at the end of October, four points at the beginning of November, and now one point. If those trends continue, the Liberals will be overtaken. On the other hand, if Léger's trend continues the Liberals will continue to lead comfortably.

Adding to the confusion this morning is a new poll by Forum Research, which I only noticed after I had updated the averages (it will be added tomorrow). The poll gives the Liberals 41% to 33% for the Conservatives and just 17% for the NDP. This would suggest Conservative stability, a Liberal uptick, and consistent NDP decrease. But Forum's one-day flash polls are perhaps more vulnerable to showing wonky results.

Returning to Léger and EKOS, however, the two polls do tell us something. The parties' support appears to be relatively stable, solidifying the newer, more competitive race that has taken shape since the summer. And the NDP remains stuck in third place without any sort of momentum.

But the differences in the numbers themselves have enormous consequences. If these were the final polls of an election campaign, we'd have a long night ahead of us.

With EKOS, the Conservatives would manage to win a plurality of seats, taking 138 to 125 for the Liberals. The NDP would be reduced to 60 seats, while the Bloc would benefit from vote splits to take 13 seats in Quebec and obtain official party status again.

With Léger, however, the Liberals would win the plurality of seats with 146, the Conservatives taking 134 and the NDP just 55. The Bloc was not high enough in Léger's poll to be in a position to win more than a seat.

These seat numbers tell different stories. In the EKOS scenario, Stephen Harper might try to stay in office and govern with a minority. The opposition might then topple the government, but would need to work together to convince the Governor-General to give them a chance to govern rather than sending Canadians back to the polls (which could be the other outcome). The role of the Bloc would be marginal, but the party would be 'back' and regain much-needed funds and staff.

In the Léger scenario, convention would allow Justin Trudeau to head-up a minority Liberal government that would likely rely on the NDP to survive. Unlike the EKOS scenario, the Liberals and NDP would not necessarily have to formalize any sort of agreement, which might be more palatable to their bases. Harper likely steps down and a leadership race ensues, delaying any threat of an election, while the Bloc is marginalized further and might just cease to exist.

And all of this is with a few percentage points of difference, showing just how significant minor shifts can be when the race gets close. If things continue this way, particularly if the NDP can regain some ground, 2015 is setting up to be one of the most exciting elections in recent memory.

Could Quebec keep the Liberals from a majority?

A new poll from EKOS Research for iPolitics and Radio-Canada suggests that, despite the fallout from the Liberals' decision not to support the government's mission in Iraq, there has been little change in voting intentions nationwide. But the poll also suggests that Quebec could be the obstacle blocking the path to a Liberal majority government.

EKOS was last in the field September 21-25, and since then has recorded no significant shift in support. The Liberals were up 0.2 points to 38.5%, followed by the Conservatives at 26.4% (+1.5) and the New Democrats at 25% (+0.6). The Greens were down two points to 5.7%, while the Bloc Québécois was down 0.9 points to 3.1%.

Suffice to say, none of the shifts experienced by the major parties were outside of the margin of error. However, the New Democrats have made gains in three consecutive EKOS polls, worth 3.6 points since the spring.

Regionally, the Liberals led in Ontario with 48.8%, followed by the Conservatives at 31.5% and the NDP at 16.4%.

In Atlantic Canada, the Liberals had 43% to 24.9% for the NDP and 23% for the Conservatives.

And in British Columbia, the Liberals were up 13 points to 36.1%, with the NDP down to 30% and the Conservatives at 19.5%.

The Conservatives led in Alberta with 42.7%. The Liberals were down to 27.9% in the province, while the NDP was at 12.3%.

The New Democrats led in two regions. They were ahead in the Prairies with about 39% (my estimate from EKOS's tiny samples in Saskatchewan and Manitoba), with the Liberals at 31% and the Conservatives at 28%.

The NDP was also in front in Quebec, with 38.8% support to 29% for the Liberals, 14.3% for the Conservatives, and just 12.6% for the Bloc. That was a drop of almost five points for Mario Beaulieu's party. Beaulieu's approval rating in Quebec was just 11.7%, with his disapproval standing at 47.6%.

Now, these numbers in Quebec are not unusual for EKOS. At the end of September, the NDP had 36% to the Liberals' 32%, while in July the New Democrats had 37% to the Liberals' 29%. Over that time period, other polls have averaged 38% for the Liberals and 27% for the NDP, so in this regard EKOS seems to be out of step with consensus opinion (with the exception of one CROP poll where the gap was two points, the NDP has not led in any of the last 15 polls done by other firms since mid-June).

But polling by CROP and Léger have suggested that the NDP holds the edge among francophones, a finding corroborated by EKOS's latest survey as well. This poses a problem for the Liberals. Though the party is doing respectably well among this demographic, they are at a distinct disadvantage. Francophones decide the results of the vast majority of Quebec's ridings and polls suggest that francophone support for both the NDP and Liberals is generally uniform. That means the Liberals could find themselves losing a large number of seats outside of Montreal by slim margins - but losing nevertheless.

The seat projection using EKOS's numbers shows what kind of role Quebec could end up playing in the next election.

Outside of Quebec, the projection model would give the Liberals 131 of 260 seats - a majority. But in Quebec, with the New Democrats a handful of points below their 2011 result, the Liberals win just 17 seats, with 57 being retained by the NDP. The end result is that the Liberals find themselves 21 seats short of a majority - and it is hard to imagine the Liberals doing much better in the rest of the country than what EKOS awarded them.

Even if we put the gap in Quebec at what the other pollsters think it is, we still get the Liberals falling short of a majority. This suggests that unless the Liberals can make the same kind of breakthrough that the NDP did in the province in 2011, their hopes for a majority government are likely to be dashed.

Overall, EKOS pegs Justin Trudeau's approval rating to be 46.1%, with a disapproval rating of 37.1%. That compares quite well to Stephen Harper's 29.9% to 63.3% spread, but is worse than Thomas Mulcair's 58% approval to 21.7% disapproval rating. And in Quebec, Mulcair's approval rating increases to 70.1% against 43.2% for Trudeau.

That will make it difficult for Trudeau to gain ground among francophones, though this weekend's appearance on Tout le monde en parle could help in that regard. But the prospect of forming government may not help him. The NDP led in Quebec in voting intentions, despite just 8% of Quebecers thinking the NDP would form government in 2015. Almost half of Quebecers think the Liberals will win.

For the Conservatives, this poll is a disaster. With under 20% support in B.C., the party takes just three seats there, and ties the Liberals for second in the seat count in the Prairies. With 79 seats (three-quarters of them in Alberta and Ontario), the party would find itself reduced to third-party status, with the NDP remaining as the Official Opposition. How this setup would work in practice is a little difficult to determine.

Of course, the Conservatives have not been as low as 26.4% in other polls, so these results are likely on the lower end of what is plausible. But we can still compare trend lines, and it seems that, so far, the Liberals are holding steady despite what has been widely considered to be a bad couple of weeks for Trudeau. But let's see what others have to say.

Guest post: Frank Graves of EKOS Research discusses his likely voter model

The final poll of EKOS Research, whether it be the tally of all eligible voters or those of likely voters, told much of the story: the Ontario Liberals were on the cusp of a majority government. But the poll suggested the NDP would take a significant step backwards from its performance in the 2011 election, which turned out not to be the case. Why?

In this guest post, Frank Graves of EKOS Research investigates.

“LIKELY” VOTERS AND THE NDP VOTE IN THE ONTARIO ELECTION: A CAUTIONARY TALE

Frank Graves

Looking back at the Ontario election, we were very pleased to have predicted not only the victor, but also the majority. Only one other firm made this correct forecast. While we are very happy with the nearness of our final poll to the actual vote tallies for the two frontrunners, we were considerably less pleased with the performance of our Likely Voter (LV) model and our underestimate of NDP support. We were a bit over four points low on the NDP, but in a majority government this underestimate is of little practical consequence. The far more important challenges were identifying the winner, the leader of the opposition, and whether it was a majority or not; all of which we accomplished clearly and accurately. Yet in a closer race, the underestimate of NDP support could have been far more consequential and the challenge of forecasting who will vote and who won’t remains an important and unsolved puzzle.

As it turns out, the failure of the LV (LV) model, and the underestimate of the NDP are closely linked. The LV model significantly lowered NDP vote from a modest to large error and it moved the Liberal and Conservative votes from spot on to too high. Clearly, this wasn’t the intention and this error is somewhat baffling as the ingredients of the model are based on established historical patterns which show significant connections between these terms and likelihood of voting. To recap, we know that generally speaking (ceteris paribus), individuals who don’t know where there polling station is, who are younger and display lower socioeconomic status, who haven’t voted in recent elections and who don’t tell us that they are absolutely certain to vote, are less likely to vote. It is therefore extremely puzzling to see that applying these factors to estimate who will show up to vote not only wasn’t helpful, but magnified the error.

Further analysis of our final polls shows why this failure occurred and it is directly linked to the underestimate of NDP support. These could be quite separate issues, but an analysis shows they are interdependent. First, let’s look at the connection between our LV model and the final result.

The basic premise of our LV model involves assigning each respondent a “likely voter” score (maximum of eight points), based on a number of factors, such as age, past vote behaviour, knowledge of voting station, etc. Since we know that voter turnout is likely to be in the range of 50%, we isolate the (roughly) 50% of respondents who received the highest scores (giving us a “cut-off” of six points). Whether we apply the model weakly or strongly makes little difference. We find that different cut-off points make some impact but basically no model decreases error; they all magnify it. This is perplexing as we and others have shown clear linkages to these terms and voter turnout.

We then looked at the individual impacts of each of the terms of the model and none of the individual terms improved things. In part, the poor performance of the LV model is rooted in the paradoxical reverse linkages between turnout and the NDP vote. NDP vote appeared to be modestly rising in the later stages of the campaign, but we didn't make much of it because it was focused in parts of the voter spectrum that are not linked to high turnout: lower education, younger ages. 

Worse, the NDP vote was much more concentrated in non-voters from 2011, they were less likely to know where their polling station was located, they were considerably less certain to vote, and they were less likely to have voted in the advance polls. A few other notable findings: the NDP were performing more strongly with households with children at home and they were more likely to be union members. The link to the deep cuts to public sector workers may have caught the attention of these groups as the campaign matured.

One other interesting feature was evident in the data: NDP supporters were much more likely to only use a cellphone. We do sample cellphones, but we were under the population values and didn't weight this group up. This was because we were burned in the 2011 Federal Election for capturing cell only respondents who were much less likely to vote and less supportive of the Conservative government. In this instance, cellphone-only households were less likely to be Liberal or Progressive Conservative supporters, but they were not less likely to vote. If we had sampled more cellphone-only households, we would have probably been closer to the actual NDP vote.

In the end, we're left conclusion that even the most cautious and empirically informed attempts at creating LV models can be dangerous because things change; sometimes quite significantly. This harkens back to the classic problem of induction and the fact that the future will often not resemble the past. We are quite certain that presidential hopeful Mitt Romney was very surprised that he didn't win, as was Frank Newport of Gallup for precisely this reason. We therefore propose a more modest and less influential LV strategy that avoids the more ambitious approach we tried here. We also think that greater effort needs to be put into ensuring full coverage of all groups (e.g., cellphone-only households and younger voters), even if these groups hadn't been critical in past exercises.

We now have a reasonable handle on the reasons behind the underrepresentation of NDP supporters and the failure of the LV model. Our solution is a lighter less ambitious LV adjustment focusing on likelihood of voting. We also think that greater efforts to represent younger voters and cellphone-only households would be a prudent strategy.

We are, however, left with one big unanswered question. What is it in the latter stages of the campaign that seemed to engage what should have been a relatively disengaged NDP base to actually show up? That question is a very interesting one to which we have no real answer at this stage. A post-election survey would help to clarify this critical question.

What we do know is that this improbable appearance on Election Day was in defiance of historical patterns of likely voting and reinforces our conclusion to put more effort into creating the best random samples and less into seeking an elusive, unified field theory of voter turnout which we increasingly believe is a chimera.

Ontario race remains tight with days to go

It is coming down to the wire in one of the closest elections this site has covered. The latest polls added to the model all suggest an extremely close race, with the margin among all eligible voters being three points or less and the margin among likely voters being less than two points. It will make for a nail-biting finish, particularly for your correspondent.

The Liberals are now narrowly in front with 37.3%, or between 36% and 41%, followed by the Progressive Conservatives at 36.5%, or between 35% and 40%. The New Democrats have up-ticked to 19.8%, or between 18% and 21%, while the Greens stand at 5.2%, or between 4% and 7% support.

The Liberals have dropped away from majority territory and are pegged to take 51 seats, or between 42 and 60 seats, while the PCs sit at 37 seats (or between 32 and 45). The NDP is now back up to 19 seats, or between 13 and 22.

This will be the last update before the final projection update, which will be posted either late Wednesday night or early Thursday morning. The final analysis, either way, will be up on Thursday morning.

It really is very close, but it is one thing to have a close race and another to have a race that might not end up being very close at all. The polls suggest a very tight margin between the Liberals and PCs, but the likely voter model from Ipsos Reid gives it to the PCs in a near landslide. And support for the NDP is either well above their 2011 performance, or well below it, which has a profound effect on the seat numbers for the Liberals and PCs. Unless the final polls show a move all in one direction, we should be going into Election Day with little comfort.

The three latest polls, from Abacus Data for the Sun News Network, EKOS Research for iPolitics, and Oracle for ECO, used a mix of methodologies. They all came to the same general conclusion for the two leading parties: either a tie or a statistically insignificant lead. The Liberals get either 34% or 35% in these polls, the PCs between 31% and 36%. That is rather tight.

For the NDP, though, the range extends from 20% to 28%. That is a very wide range, and spans everything from a waste-of-time election for the NDP or a major breakthrough.

It is interesting to compare the trend lines, however. Abacus was last in the field May 28-31, just before the debate on June 3. Since then, the Liberals dropped three points, the Tories picked up one, and the NDP gained four.

EKOS has been conducting a rolling poll since before the debate, but if we look at its May 29-June 1 numbers, done at mostly the same time as Abacus, we see similar movement: a drop of four points for the Liberals, a gain of one for the Tories, and a gain of three for the NDP. It will be interesting to see if, in the last numbers put out by these polls, we will see things continuing to move in the same direction.

Adam Radwanski in the Globe and Mail today writes about the internal polls being done by the Liberals and Tories. Both now seem to expect a minority government in which they will hold sway, but more interestingly perhaps is that both internal polls have shown an uptick for the PCs since the debate. That jives with the public polls. The projection, for example, had the Tories at just 33.8% in polls done up to June 2.

Abacus had a series of other questions included in their poll that may shed some more light on these horserace numbers.

32% of Ontarians expect the Liberals to win, while 25% think the Tories will prevail. That is rather close, and suggests that voters may go into the polling booths expecting a very tight result. That might encourage strategic voting, which could hurt the NDP and Greens.

On the other hand, just 24% think the Wynne government deserves re-election, a drop of three points since last week, while 54% think it is time for a change, a jump of six points.

Interest in the campaign is growing, as 42% told Abacus they had given quite a lot of thought to the campaign, up from 32% when Abacus first polled in the early stages of the race. Interestingly, and perhaps depressingly for those in the media, the proportion of respondents saying they are following the news concerning the campaign very closely has hardly changed (24%).

Counter-acting the benefits that Tim Hudak has probably gained as a result of the debate (which Abacus's respondents said he won) is that the percentage of Ontarians saying they have a negative opinion of the PC leader has jumped to its highest of the campaign so far, at 52%. No wonder, then, that the Liberals are going after NDP voters with the Hudak boogeyman. Will it work? According to Abacus, NDP/OLP swing voters favour Kathleen Wynne as premier over Andrea Horwath by a margin of 48% to 26%. Oracle, meanwhile, showed that 46% of non-PC voters were casting a ballot against Hudak, rather than for another leader or party.

It is a dizzying array of pros and cons for each leader. Plausible, defensible, and ultimately accurate arguments as to why Wynne or Hudak will win (or that the NDP will out-perform expectations) on Thursday could be written up, and proponents of either could only reasonably agree to disagree. A terrific situation for democracy and election-watchers, a horrible one for pollsters.

Decisive final week begins

A number of polls have been released in the last few days, and they only emphasize just how important the next few days might be in deciding who will form the next government. The numbers are very close between the Liberals and Progressive Conservatives among all voters, but it is among those who will actually vote that the election will be won.

The projection has ticked back towards the Liberals since yesterday, with the party at 39.2% support, or between 38% and 43%, a gain of 1.9 points. The Tories are down 0.5 points to 36.8% (or between 35% and 40%), while the NDP is down two points to 17.6%, or between 16% and 19% support. The Greens are at 5.3%, or between 4% and 7%.

The Liberals are back up to the bare minimum for a majority government, with a four-seat gain since yesterday to 54, or between 46 and 64. The PCs are down two seats to 38, or between 30 and 50 seats. So there is a bit of overlap between the two parties. The NDP is down two seats to 15, or between 8 and 16 seats.

Support for the New Democrats is, perhaps, the wildcard of this election. It is less consequential for the NDP than it is for the other parties, however. If the NDP ends up below 20%, the Liberals will likely prevail even with a vote deficit against the Tories. But if the NDP is over 25%, the Liberals stand little chance. But where will the party end up?

The chart below shows the latest tally among all eligible voters for Ipsos Reid and EKOS Research. I've included Forum's numbers, but the company uses a turnout model to adjust their numbers.

When you look at these three polls, the latest from these three firms, there is little real divergence on the two main parties. The margins are very close - a tie according to Ipsos and EKOS, a two-point lead for the Liberals according to Forum. Support for the Liberals stands at between 35% and 39%, the PCs between 35% and 37%. Those are pretty clear numbers.

But support for the NDP is another thing entirely. The IVR pollsters (EKOS and Forum) give the party just 17% or 18% support, while Ipsos (online) gives them 26%. That is a huge difference, and one that is extremely important for the seat outcome. The Liberals have a much better chance of winning with the numbers laid out by Forum and EKOS than they do with those of Ipsos.

It should be noted that throughout this campaign Ipsos has been on the high side for the NDP and EKOS/Forum on the low side, so we may be seeing some methodological bias at play here.

Can the results of the past few elections provide any guide for us? Every pollster over-estimated the New Democrats in both the 2007 and 2011 elections, by between 0.2 and 3.3 points. In 2011, two of the final polls were done online (Angus Reid and Abacus), two were done via the telephone with live-callers (Ipsos and Nanos), and two were down via IVR (Forum and EKOS).

The online polls over-estimated the NDP the most, by an average of 2.3 points. The telephone polls over-estimated the New Democrats by 1.4 points, while the IVR polls over-estimated the party by just 0.5 points. If something similar happens again in 2014 (and past performance is by no means something that remains constant going forward) we could expect Forum and EKOS to have the NDP's support closer to the mark. But is that realistic at 17% or 18%?

In terms of trends, all of the change between these new polls and the last time these firms were in the field at around the same period shows general stability.

The estimates by EKOS and Ipsos of likely voters is even more confusing, since they portray two starkly different outcomes. Again, I've included Forum for comparison.

Ipsos gives the Tories a significant lead among likely voters with 40% to 32% support, with the NDP still holding its own at 24%. EKOS, however, gives the Liberals the edge with 42% to 35%, with the NDP at only 14% support. That is extremely low, and would be the worst performance by the party since 1999.

Oddly, Forum's turnout-adjusted numbers seem to split the difference (Abacus Data's numbers have generally done the same, we'll hear from them later today). If this pattern holds through to election day, either Forum and Abacus will land close to the mark while both EKOS and Ipsos miss, or one of the last two will nail it and put everyone else to shame (the middle-of-the-road pollsters to a lesser extent).

But we should hear from every pollster again before Election Day, so perhaps we will see some convergence. Or, someone will be in for a rough night on Thursday.

Riding polls

The Star published the numbers from a series of 18 riding polls conducted by Forum in some swing ridings (a few did not make it on the last that would have been interesting). I've broken them down by region below to see if we can discern any patterns - but we should note that all of these riding polls were conducted before the debate.

In the Toronto/905 region, we can see that the Liberals are looking rather strong. They looked good against the PCs in Etobicoke-Lakeshore, Mississauga-Erindale, Pickering-Scarborough East, and St. Catharines, while they were narrowly edging out the Tories in Brampton-Springdale and Burlington. This generally backs-up the decent polling numbers the Liberals have been putting up in the GTA.

In the downtown core where the Liberals are facing off against the NDP, the races were close but they nevertheless held the advantage: six points in Davenport and one point in Trinity-Spadina. The NDP, however, looked to be in a good position to retain Niagara Falls.

The polling numbers in the four southwest ridings seem to give some credence to the better-than-expected support levels of the Liberals in the southwest. They were ahead in Cambridge, narrowly led the NDP in Kitchener-Waterloo, and were competitive in London West and Windsor West. The PCs were not doing as well as they could have, suggesting that the Tories' support in the southwest may be concentrated outside of the urban centres more than the party needs it to be.

In the north, the NDP was leading in Sudbury while the Liberals were ahead in Thunder Bay-Atikokan. In eastern Ontario, the races were very, very close between the PCs and Liberals, with the NDP a non-factor.

Did the debate shift any of these? It is hard to tell since the polls have been inconclusive on the effects of the debate, apart from a possible uptick for the PCs. This means the party may have moved ahead in ridings like Ottawa South, London West, Brampton-Springdale, and Burlington. But local candidates can mean a lot as well, particularly in ridings like Ottawa South and London West that held by-elections recently. There could be many very close races on election night.

Programming note: the next update on Tuesday will be a little later than usual. On Wednesday, I will probably save my update until the very end of the day, after all of the last polls of the campaign are published.