Friday, August 17, 2012

CAQ gains put PQ majority in question

Le Journal de Montréal released the details of a monster Léger Marketing survey at midnight to great anticipation, though the results were less shocking than they were in-depth. But they do show that the Coalition Avenir Québec is continuing to make modest, but steady, gains. Now that they are pushing the Liberals for second place, they are also starting to push the Parti Québécois out of some seats.

But the PQ is still in the lead, though the party slipped one point to 35% in the projection. They are also down five seats to 65, putting them just over the bar of the 63 seats needed to form a majority government.

The Liberal slide continues, as they have dropped 1.2 points to 29.9%. Because of better numbers in and around Montreal, however, the party has managed to increase its seat haul by one to 38.

The CAQ is up 1.4 points to 24.6% and five seats to 21, while Québec Solidaire is down 0.2 points and one seat to 5.7% and Amir Khadir.

Option Nationale is up 1.2 points to 2.1%, putting them ahead of the Greens. They are down 0.2 points to 1.6%.

The seat ranges have moved significantly, with the Parti Québécois capable of winning between 53 and 79 seats. As the Liberal range has dropped to between 24 and 51 seats, a Liberal plurality is no longer considered possible. Instead, they are now threatened by the CAQ for the role of Official Opposition, as the party could win between 14 and 32 seats. That is a bit of an overlap, and with the CAQ having the momentum and running much closer in the straight poll average, the odds are turning against the Liberals.

Québec Solidaire dropped significantly on the island of Montreal, falling 3.9 points to only 8.1%. That puts them only a few ticks above their 2008 performance, and that is why the party is only projected to win one seat, and as few as none. The Liberals dropped 2.1 points to 36.4% and the PQ two points to 29.1%, while the CAQ made a big five point gain. They now sit at 18.7% and actually have three seats in range. That is primarily due to the divergent poll results increasing the volatility on the island, but it is something to keep an eye on.

In the suburbs around Montreal, the PQ dropped two points to 36.4% and six seats. The lion's share of the gains went to the Liberals, up 1.3 points to 28.7% and four seats. The CAQ was also up, gaining a point to reach 26.6%.

Quebec City saw a fair amount of change, as the Léger poll gave the CAQ a wide lead, rather than the Liberal lead that CROP identified. As a result, the CAQ picked up 3.6 points to reach 34.1%, and is up three seats to six. The Liberals dropped 5.6 points to 29.6% and five seats to only three, opening an opportunity for the PQ. The party only increased its support by 0.4 points to 26.4%, but gained a seat in the process.

In central Quebec, the PQ continues to lead with 33.6% (-0.3) but the Liberals fell 0.3 points to 28.4%, costing them one seat. The CAQ won it despite dropping 1.7 points to 27.5%. Notable is the 3.3-point gain by Option Nationale in the region. They currently have 4% support, just behind Québec Solidaire. It is no coincidence that their best region of the province is also where Jean-Martin Aussant is running.

And in western Quebec, the CAQ picked up 3.6 points and one seat. They are still in third at 25.2%, but the gap has narrowed with the Liberals as they were down 2.8 points to 26.4%. The PQ, at 38.3%, still leads in the region.

Note that the projection model gives the Liberals (and the PQ) an organisational/turnout bonus, and this is why the party is still in second place in the Montreal suburbs and in eastern, central, and western Quebec. If we go just by the polls, however, the Liberals are in third in those regions.
Léger was last in the field on Aug. 6-8, and since then the PQ has gained one point and has extended its lead over the Liberals. The PQ sits at 33%, while the PLQ has dropped three points to only 28%. At 27%, the CAQ is steady.

Québec Solidaire is also unchanged at 6%, while Option Nationale is up one point to 3%, their best polling result on record. The Greens were unchanged at 2%.

Only the Liberals' three-point drop appears to be statistically significant. Their drop of 19 points among non-francophones is also notable, as they are now at 62%. The CAQ made a big 11-point gain to 20%, the second consecutive poll to give the party this level of support among these voters. It could be a coincidence, or it could be that the CAQ's attempts to woo the anglophone electorate are paying off.

The CAQ's increase on the island of Montreal (from 14% to 22%) is accordingly outside the margin of error, while Québec Solidaire's drop of five points to only 7% is also more than a wobble. It is quite unusual for QS to be polling so low in Montreal, their one region of concentrated support. 

Riding polls

Segma Recherche completed its tour of the Saguenay-Lac-Saint-Jean region for Le Quotidien today. In Roberval, Segma gives the PQ a wide lead with 53% to 26% for the Liberals. The CAQ, at 14%, trail in third by some distance. In Jonquière, the PQ is also very secure with 51% support to 20% for the Liberals and 17% for the CAQ. Both of these results were very close to what the projection had, but again we're seeing the Liberals being slightly over-estimated at the expense of the CAQ. 

But there is a bit of a variation between the results from Léger's poll (in the field Aug. 13-16) for the Saguenay-Lac-Saint-Jean and the combined results of Segma's polling (the five surveys spanned Aug. 11-16 altogether). Léger gives the PQ the lead with 38%, but puts the CAQ at 26% and the Liberals at 22%. By contrast, the average of Segma's five riding polls gives the PQ 52% to 23% for the Liberals and only 17% for the CAQ.

Granted, Léger polled only 262 people in the region while Segma's five polls surveyed almost 3,000, so the margin of error can account for much of the discrepancy. Segma's more in-depth polling also has to be given the benefit of the doubt. But it does suggest that the PQ's lead in the area might not be as solid as Segma has found it to be. And if that is the case in the Saguenay-Lac-Saint-Jean region, could it be the case elsewhere?


  1. I wonder if we might see something of a repeat of the 2011 federal election with CAQ playing the role of the NDP. The traditional Liberal strategy is to present themselves as the only credible alternative to their main competitor (the Tories, federally, the PQ, provincially) and to say "Apres nous, le deluge". But toss in a competitive third party, and all of a sudden that strategy blows up, because that logic applies equally to voting for the third party.

    Moreover, in an scenario where the third party has little hope of forming a majority government, voting for them becomes a risk-free strategy for people who might otherwise be Liberal voters (especially if such a government would rely on a Liberal rump for support - you get to both spank the Liberals for being naugthy why letting them retain some influence over the government).

    Assuming Legault doesn't shoot his mouth off and say something stupid (and so far he's run the best campaign), I wouldn't be surprised if we see as big surge in CAQ support in the final weeks of the election (and wonder if the apparent increase in support among anglophones might be the first sign of that).

    1. This is what I'd bet on happening too. The "strategic voting" argument has always been a weak one.

  2. The CAQ got 27% in last the last Léger-QMI poll, but your "high" projection is only 25.6%...
    Is there a lag in your projection? Is it an average of the X last surveys?

    1. The projection assumes that a party like the CAQ (also QS and ON) will be over-estimated in the polls, that is why they are registered at 24.6% in the projection.

      If you see in the left-hand column of the chart, their poll average (unadjusted) is 26.3%.

      All of the polls included in the projection are at the bottom of the page. It is a weighted average that takes into account sample size, date, and track record.

    2. Why do you this? The ADQ was underestimated by 1-4 % in the polls leading to both the 2008 and 2007 election.

    3. Hi

      Can you answer the above question please.


      Anonymous17 August, 2012 19:10

      Why do you this? The ADQ was underestimated by 1-4 % in the polls leading to both the 2008 and 2007 election.

    4. The adjustment is based on where the party sits in the legislature and is based on the results of 16 federal and provincial elections.

      Here is the relevant section in the Vote Projection Methodology explanation:

      While the polls are generally accurate in almost every election, differences between voting intentions and voting behaviour do exist. In some cases, the polls are on either side of the result and there is no consistent under-estimation or over-estimation of support. However, in many cases there are important differences, and an analysis of 16 recent provincial and federal elections indicates that support for a party is usually under-estimated in polls when that party is the governing (83% of the time) or main opposition (63% of the time) party in the House of Commons or provincial legislature. Third and other parties are over-estimated 75% of the time, while parties not in the legislature are often over-estimated in the polls by a significant degree (and 94% of the time).

      This adjustment acts as a proxy for organization, fundraising, and enthusiasm. In short, larger parties tend to be better funded and organized, and have an easier time getting their vote out. This intentions vs. behaviour adjustment takes this into account.

      Accordingly, the vote projection is adjusted based on where a party sits in the legislature.

  3. Regarding the riding polls, I think they can be a bit of a lagging indicator. I remember how in April 2011 the province wide polls in Quebec showed the NDP gaining a lot of ground, but then polls in individual ridings tended to show much less of an NDP "bounce". A riding poll in Sherbrooke 10 days before the election gave the BQ incumbent a commanding lead - and he ended up losing to the NDP's 18 year old Pierre Luc Dussault!

    1. Why would they lag in particular, though?

    2. Perhaps because they focus more on local issues and local preferences, which are slower to change? You get to see Marois, Charest and Legault on TV all the time so your opinion on them fluctuates a lot. Your local candidate you will see far less often, so that preference may remain stable. If it's one's opinion of Marois, charest or Legault that ultimately rules the day for a lot of people...

      Just throwing a hypothesis out here, but it's an interesting point by DL.

  4. I might be partly because the other parties have incumbents running for re-election and that early in the campaign when the names of candidates are read it might favour the better known sitting MNA over an unknown challenger...but then if a wave builds - all bets are off.

  5. Now that Mulcair has announced a QNPD, I wonder if we will get some hypothetical polls? Who would lead the party, Pierre Ducasse?

    1. I highly doubt that before the 5th September. After, they could, but NPD support is an odd thing at federal level and too allow it on the provincial level, you'll need to severly damage the PQ and the CAQ who hold between them the "left vote" and the "not so inclined toward Liberals but still federalist vote".

      As tvtrope sum it up somewhere, Canada's politics is plain weird and Quebec's is even stranger.

    2. It'll be interesting to see how this all pans out, because I have a feeling this isn't a very good idea for the NDP. Given the ties to the federal party, and what happened last time, an NPDQ would have to be as resolutely federalist as the Liberals lest the federal party be harmed, which would probably turn off a majority of PQ voters.

      Moreover, Quebec politics is firmly to the Left of Canadian politics in general – both PLQ or CAQ (who deny being centre-right even in provincial politics) would probably count as left-of-centre on the federal spectrum. So an NPDQ would have a very thin line to take economically, to distinguish itself from the PLQ and avoid being seen as a far-Left embarrassment to the federal party. Of course, were the NPDQ to poll badly, it'd be an embarrassment in itself and could drag down the federal party.

      Then again, it would have novelty, its name, and Quebec politics' volatility on its side, and we might see some kind of realignment after this election, so who knows?

    3. It's a clever idea on Mulclair's part. Establishing a provincial NDP in Quebec would be a way of capitalizing on the NDP's current strength in Quebec to establish a permanent presence there federally.

      True, as Jo points out, the downside of having a provincial party is that the federal party tends to catch flak for the actions of their provincial cousins (which has been a real problem for the Liberals and Tories over the years). On the other hand, a provincial party gives you a presence on the ground, and serves a potential training ground for federal politicians. Moreover, in the Quebec context, having a provincial wing might avoid the problem of, say, having an acting party leader who, less than a year previously, was a member of a separatist party, since provincial parties often serve as a breeding ground for federal politicians (especially when the federal party is out of office).

      In fact, this could be a stroke of genius on Mulclair's part if there is an appetite for a left-wing federalist (or, at least, not explicitly separatist) party in Quebec. That is at least one conclusion (though not the only one) that one could draw from the results of the 2011 federal election. If, as polls suggest, the sovereigntist movement is at its weakest point in 40 years, now might be an ideal time to try to carve-out space for a non-separatist left-wing party in Quebec. If he pulls that off, well, more power to him.

  6. The most surprising thing is, who lead in Quebec region? Liberals are solid there and have a good chance to retain their seat, but Leger seem to put them second with somewhat of a gap. Charest might not be so popular, but he did deliver the goods for the grand region of Quebec. And the «mystère Québec» is not so much of a mystery that the voter there do take in account what the different party did for them.

    Well, at first - when the elections started - I though that Liberals had the best shot to from a minority government, with some support from the CAQ. Now it look like the situation could reverse. PQ need a majority, they won't be able to govern long without it.

    Their is momentum for Legault and he is a good debater, if Charest doesn't make a great performance, thing could get quit interesting. Once you get thing going your way, it's easier keeping them like that.

  7. Chantal Hebert has a very interesting piece in yesterdays Toronto Star. Well worth the read.

  8. Do we expect polls today to gage the impact of the debate last night?


      Charest won.


COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.

Note: Only a member of this blog may post a comment.