Monday, June 29, 2015

NDP now favoured in new polls and seat projection

In the first update since the site went on hiatus two weeks ago, the New Democrats are now leading in both the vote and seat projections for the first time since 2012.

It was quite a two weeks to be away, as it featured some of the most dramatic swings in voting intentions we've seen since the immediate aftermath of Justin Trudeau's Liberal leadership victory. Coupled with the NDP's surge into first place in every poll conducted by a gaggle of pollsters using every methodology under the sun was the return of Gilles Duceppe as leader of the Bloc Québécois, jarring the race in Quebec as well.

The NDP now leads in the poll average with 32.4%, an increase of over three points since the pre-hiatus projection update. The Conservatives have dropped a little more than one point to 28.9%, while the Liberals are down a little less than one point to 27.4%. The Bloc has moved ahead of the Greens with 5.2% to 4.9%.

In terms of the seat count, the NDP is now projected to win between 113 and 140 seats, up significantly from the pre-hiatus update which did have the NDP overlapping with the Conservatives but still firmly in second place. The Tories have fallen to between 99 and 141 seats, while the Liberals have dropped to between 71 and 106 seats.

The Bloc is now projected to take between one and five seats, and the Greens are projected to win just one.

While the NDP and Conservatives have a similar high range, the Conservative's low range is 14 seats below that of the New Democrats - and the precise projection puts the NDP at 127 to 114 for the Conservatives. This is the first time that the Conservatives have trailed in the seat projection since the official model was launched at the beginning of the year.

There were quite a few polls released over the two weeks I was away, but their field dates influenced the last five weeks of projections in the model. For instance, the Angus Reid poll that was released while I was away was actually older than the EKOS Research poll that was out on the Friday just before my departure.

So, I've compiled all the numbers in the chart below to give you all a full accounting of how the projection would have looked throughout the month of June. As a reference point, I've included the projection just before the NDP's provincial victory in Alberta, which seems to have been the catalyst for the recent shift in voting intentions.

As you can see, the NDP's numbers were relatively stable over the last week of May and first two weeks of June, after initially surging from the pre-Alberta numbers. But the party has seen been edging up over the last two weeks.

And with the addition of the Angus Reid poll, we can see that the NDP has been leading in the poll average now since the week ending June 8.

This is because the Conservatives have been slipping. The Liberals seem to have contributed to the NDP's surge in May, but since the beginning of the month the NDP has primarily been taking from the Conservatives and Greens. The Tories were at 30.8% and leading in the week ending on June 1, but have since been dropping in every subsequent week. The Greens have taken a hit over the last two weeks, though that might be because of the unusually poor results for the party in the most recent polls by Forum Research and Ipsos Reid.

In terms of the seat ranges, we can see that the Conservatives were comfortably ahead at the beginning of May. They overlapped only slightly with the Liberals, while the NDP was solidly in third.

The NDP did move into second place by the end of May, and were overlapping with the Conservatives more than they were the Liberals. By June 8, the NDP had moved clear of the Liberals and was seriously challenging the Conservatives. By June 22, the NDP and Conservatives were effectively tied in the seat count, and they have since pulled into a far superior position. The Liberals' high range has inched upwards, while the Bloc has gone from end-times-disaster levels to 2011-disaster levels.

Going just by the averages, we see the same sort of story being played out. After dropping, the Liberals are now holding steady as the NDP eats into the Conservatives' seat numbers. Sooner or later, the Tories will need to turn more of their attention on Thomas Mulcair.

The shift over the last two months has been nothing short of incredible. The Conservatives have dropped 29 seats and the Liberals have dropped 20 seats in the projection since the beginning of May. The NDP, meanwhile, has picked up 48 to move into first place.

At these levels of support, the NDP is on track for a plurality of seats. It is interesting to note, though, that the maximum projected total for the NDP is 179 seats, which puts them over the majority mark. Granted, that assumes the NDP has hit about 41% support in voting intentions. But the path to a majority can at least be laid out. The same goes for the Conservatives but not, at this stage, the Liberals.

There are two questions that will be answered over the next few weeks. The first is, of course, whether or not the New Democrats will continue to lead in the polls through to the start of the campaign (official or not, we have to consider August 6 to be the effective start of the campaign as that is when the Maclean's debate will be held).

The second question (a set of them, really) revolves around the Liberals and the Conservatives. Look at the recent set of polls that have been out:

The Conservatives seem to have dropped quite a bit, after routinely polling over 30%. They are now polling in the high-20s, which has put them in a tie with or behind the Liberals in four of the last polls. This is in contrast to the EKOS and Angus polls from earlier in June which suggested that the Liberals were collapsing.

Are the Conservatives going to continue losing support? Did the Liberals hit a rough patch in early June, only to recover over the last few weeks, or were those polls by EKOS and Angus Reid slightly anomalous? Can the Liberals sustain support in the high-20s when the NDP is polling in the mid-30s?

We're in the midst of a period of transition in voting intentions, so it will be very interesting to see where the numbers go from here.

Duceppe: 2011 results in a 2015 context

I would be remiss not to address the shifting landscape in Quebec, which has two factors currently at play. The first being the surge of the NDP, which was felt in Quebec before the change of leadership at the Bloc, and the second being that return by Gilles Duceppe.

There definitely has been some movement that can only be attributed to Duceppe. In the week ending June 8, just before Duceppe returned to lead the Bloc, the party was polling at 17.4%. It is now projected to take 21.4%, its best score since the official model was launched at the beginning of the year and better than any monthly average Mario Beaulieu ever managed. Even two large-sample polls by Léger and CROP put the Bloc at 26% and 25%, respectively.

Still, in terms of the weighted average it is an increase of just four points that can be attributed to Duceppe. And this is still two points below Duceppe's performance in 2011. But despite the poorer showing than that election the model gives the party between one and five seats, potentially matching or surpassing that 2011 performance. The reason for that is simple: the NDP remains the Bloc's main opponent in most ridings in Quebec, and the NDP is averaging 35.1%. That is below the 42.9% of 2011, and so this gives the Bloc a better shot. Every extra point pays out-sized dividends to the Bloc when the NDP is below 40%.

But has Duceppe's return hurt the NDP? That is more difficult to say. The NDP was in the midst of a surge in Quebec when Duceppe made his announcement. In the week ending on June 1, the party was averaging 37.8% in the polls and was well on its way to repeating its 2011 scores. At first glance it appears that Duceppe may have reversed that NDP surge somewhat from the high-30s to the mid-30s, but it is impossible to know for certain. The average for the week ending June 8, before Duceppe's return, put the NDP back down to 35.9% in Quebec.

What can be said with certainty is that both the Liberals and Conservatives are doing worse in Quebec than they were just a few weeks ago. The Liberals were slowly leaking support through to the end of May, but fell steeply over the last few weeks to 23.5%, down from 26% or 27%. The Conservatives, meanwhile, have dropped to 15.8% from 21% at the beginning of May, and from 18% just before Duceppe's return.

The trend lines tell the story that the NDP was making gains at the expense of the Liberals and Conservatives in Quebec after their Alberta victory, and then subsequently the Bloc made gains at the expense of the NDP when Duceppe came back onto the scene (though not nearly enough to erase the gains the NDP had made). Undoubtedly, though, the truth is somewhat more complicated than that, with voters crossing the political spectrum in less direct ways.

These are interesting times, with multiple front campaigns taking place everywhere. The NDP appears to have made its gains at the expense of the Liberals in most parts of the country, and are now starting to eat into Conservative support. The rejuvenated (re-oldinated?) Bloc in Quebec makes for a different dynamic there. Every party needs to start reviewing its strategy with just a few months to go.

Friday, June 12, 2015

ThreeHundredEight.com on hiatus until June 29

In order to take some time off for a much-needed vacation and to re-charge the batteries for the final sprint towards Election Day in October, I will not be posting any new articles or projection updates to ThreeHundredEight.com or writing for or appearing on the CBC for the next two weeks.

I will be very infrequently checking Twitter and moderating comments here, but I will otherwise be incommunicado. 

But it isn't all bad news! In fact, shortly after I get back to work I will have something very, very cool to show all of you. I'm really excited about it, and I think you will all love it.

In the meantime, there is plenty right here on 308 to keep you busy. Here are a few of my favourite pieces I've written over the last few months that you might have missed:

Introducing the 2015 federal election projection model. Goes over how the model did in 2011 and what changes have been made since then. Probably a good thing to read as we head into the final stretch.

- The Federal Election of 2003 that never was. Paul Martin vs. Stephen Harper vs. Peter MacKay. Go! (Also Jack Layton and, of course, Gilles Duceppe)

- Without Wildrose or a divided right, the Alberta NDP would have still won. A somewhat controversial post. Really, it is about how the NDP's win was not the product of a divided right, and not an alternate history in which I neglected to consider the possibility of Zombie Lougheed winning it all.

- Are conservative parties under-estimated in the polls? Spoiler alert: they are, but not as much as you might think.

The quest for official party status. Suddenly more relevant with the return of Duceppe. Looks at the Greens, too.

- Redemption for the pollsters, revolution for Albertans. My relieved look-back at Alberta's election results.




You can also take a look at the archive of my weekly columns for the CBC here, and my monthly columns for The Hill Times here (paywalled, but nothing's free in this world, except this site).

I'm sure I'll have a lot to catch-up on when I get back to work, so if any of you see a poll you think I should know about please do tweet me or email me. It will help!

See you in a few weeks! 

Wednesday, June 10, 2015

NDP gains, Liberals drop in Atlantic Canada

The latest polling from the Corporate Research Associates shows that the New Democrats have made impressive gains in Atlantic Canada over the last few months, with big increases in each of the region's four provinces. The gains have come primarily at the expense of the Liberals, but that party is nevertheless on track to win a majority of the region's seats.

The Liberals led in the region-wide poll with 43% support, a steep drop from the 56% that CRA pegged the Liberals to be at in February.

The New Democrats were up 15 points to 29%, while the Conservatives were down two points to 24%. The Greens were unchanged at 4%.

Undecideds, would-not-votes, and non-responses totaled 41%. While that might seem very high, note that CRA does not ask a follow-up 'leaning' question outside of an election campaign, contrary to what most other pollsters do.

Justin Trudeau scored highest on who would make the best prime minister with 36%, followed by Thomas Mulcair at 22% and Stephen Harper at 19%. After taking out the undecideds, these numbers are virtually identical to party support, and the shift for each of the leaders mirrors that for their respective parties.

What is most interesting about CRA's polling is that it allows us to look at the party standings in each of the region's four provinces.

The results were generally uniform, with the Liberals leading in all four provinces with between 40% and 47%. The NDP placed second in Newfoundland and Labrador, Nova Scotia (where the margin was smallest), and Prince Edward Island. The Conservatives were second only in New Brunswick.

Compared to CRA's last poll, and taking into account the margin of error, the shift in voting intentions was generally uniform throughout the region. New Brunswick bucked the trend, though, with the Liberals sliding only a little and the Conservatives dropping quite a bit more than they did elsewhere in Atlantic Canada.

The drama comes when we compare these numbers to the 2011 election. The NDP is almost back to where it was in 2011, up two points from that score in Nova Scotia (to 32% in CRA's polling) and 11 points in PEI (to 26%), and down three points in both Newfoundland and Labrador (to 30%) and New Brunswick (to 26%).

The Liberals, however, have taken a huge chunk of the vote away from the Conservatives, even in light of this more recent drop. The party is up four points over its 2011 performance in PEI (to 45%), nine point in Newfoundland and Labrador (to 47%), 12 points in Nova Scotia (to 41%), and 17 points in New Brunswick (to 40%).

The Conservatives have dropped six points in Newfoundland and Labrador since 2011 (to 22%), 14 points in Nova Scotia (to 23%), 16 points in New Brunswick (to 28%), and 21 points in Prince Edward Island (to 20%). Those are some huge shifts in support in four years.

In terms of seats, the Liberals would take the lion's share. They'd win five in New Brunswick, six in Nova Scotia, all four in PEI, and five in Newfoundland and Labrador for a total of 20. The New Democrats and Conservatives would each win six seats. Three of them would come for the NDP in Nova Scotia, two in Newfoundland and Labrador, and one in New Brunswick. The Conservatives would win four seats in New Brunswick and two in Nova Scotia.

The shift in voting intentions since February is too large to be a blip (click on CRA's PDF for the comically enormous swing in their tracking chart). And other polls have shown the NDP on the uptick in Atlantic Canada and the Liberals slipping, as shown by the polling averages chart below (before the inclusion of the CRA poll).

Polling averages before inclusion of CRA poll
That the Conservatives are on track for a drubbing in Atlantic Canada seems clear, as they have been struggling in the region for most of the last four years. They will put up a tough fight in a limited number of ridings, however.

The real question is whether the movement between the Liberals and NDP is part of a trend or just a new reality. Will the NDP continue to climb until the two parties pass each other, or will the Liberals retain their double-digit lead?