Friday, June 6, 2014

PCs gain post-debate?

After showing a rather comfortable and sustained lead for the Liberals among all eligible voters throughout the campaign, the latest poll from EKOS Research suggests the Progressive Conservatives have made some significant gains post-debate. Is it a blip, or a sign of things to come?

The Liberals still lead among likely voters, however, so the projection still gives the Liberals the edge with 38.9% (or between 37% and 43%). That is a drop of 0.5 points from yesterday. The PCs are up 1.5 points to 35% (or between 34% and 38%), while the New Democrats were down 1.2 points to 18.4% (or between 17% and 20%). Support for the Greens stood at 6.6% (or between 5% and 8%).

The Liberals have moved off from a majority with a reduction of two seats to 53 (or between 48 and 65), with the PCs up three seats to 40 (or between 29 and 45). The NDP was down one seat to 14, or between 12 and 14.

New polls will be coming out soon to counter-balance what has been an EKOS-only polling period since May 31 (a week ago!). Ipsos Reid is supposed to report tonight and Abacus Data will have its regular poll on Monday, while we can probably expect to hear from Forum Research over the weekend. It will be very interesting to see what these polls show as EKOS has recorded a rather significant bump for the PCs.

The last independent sample of EKOS's rolling poll was done May 30 to June 2, so just before the debate. This gives us a good chance to see if the debate had any effect, since the newest poll was taken between June 3-5 (polling on the 3rd would have encompassed some of the post-debate period). Of course, we should also be careful not to draw any specific conclusions. While the debate could be the cause for any shift in support, the real reason could be anything else. But we can reasonably assume that the debate and post-debate coverage is a significant factor.

Among eligible voters, the PCs picked up 4.4 points since May 30-June 2 to move into the lead with 34.9%, the first lead that EKOS has given the PCs since April 2013. The Liberals fell 4.9 points to 33.9%. Both of these shifts are outside of the margin of error, and so worth noting.

The NDP was up 2.7 points to 20.5% (within the margin of error), while the Greens were unchanged at 8.4%.

Among likely voters, however, the Liberals continued to lead with 39.5% against 35.6% for the PCs and 16.7% for the NDP. This means the Liberal lead over the Tories has shrunk from 9.4 points to only 3.9 points since yesterday's poll. The NDP, despite gaining among all eligible voters, dropped among likely voters.

This PC gain could be significant, or it could be a blip. We will have to see what EKOS shows in subsequent polls, as well as if other firms record something similar.

At the regional level, the PCs made gains in every region except the southwest, though only the 10-point gain in Toronto was outside the margin of error. The NDP picked up 15.1 points in the southwest, a region that is now a three-way tie according to EKOS. The Liberals still led in Toronto and the GTA, while the PCs were ahead in eastern, southwestern, and north/central Ontario.

But let's take a moment to talk about Toronto, because something may be happening there. I already discussed this potential earlier in the campaign, when the Tories seemed to be making gains. They are still polling relatively highly, but it appears that the NDP may have lost a lot of support in the city to the Liberals - and it is costing them a handful of seats.

As the chart above shows (showing the regional projection for the city), the NDP has been slipping since May 29 (and if we remove that bump, perhaps since May 23). The latest EKOS survey has them at just 16.7% in the city. While EKOS has been under-scoring the NDP compared to other pollsters throughout the campaign, this Toronto phenomenon is not limited to EKOS.

The last poll by Forum (May 27), had the NDP at just 17% as well. The most recent poll by Ipsos Reid (May 26-29) had the party at 22%. And the last Abacus survey (May 28-31) pegged the party at 17% in the city. The rolling EKOS polls have averaged 16% for the NDP in Toronto, a region defined the same way by all four of these pollsters.

In 2011, the NDP took just under 27% of the vote in Toronto. It is the region where they have lost the most support. In the rest of the province, they are polling about as high as their performance in 2011.

The Liberals are not taking much more of the Toronto vote than they did in the last election, but the PCs seem to have hoovered up about seven-points' worth of new supporters. We can only speculate why this is happening. In other parts of the province, it is perhaps easier to envision more populist NDPers swinging over to the PCs, but in downtown Toronto it seems more likely that NDPers fearful of a PC victory are going over to the Liberals, replacing those disgruntled Liberals who are swinging over to the PCs. With the NDP polling so low in Toronto, the chances of a Liberal majority are significantly increased. Something to monitor.

53 comments:

  1. With the reported EKOS numbers, I get:

    44 OLP
    43 PC
    20 NDP

    With their likely-voters, I get:

    56 OLP
    37 PC
    14 NDP

    And the new 308-aggregate gets me:

    55 OLP
    37 PC
    15 NDP

    ReplyDelete
    Replies
    1. If u believe expected voters, dont. It is a rallying cry every time to the trailing party.calling a dead heat at the polls.

      Delete
    2. I'm not passing judgment on the quality of polls or likely-voter models, I'm simply reporting the numbers as they come out when I put them in my simulator. I have a feeling (without watching the debates or hearing anything about how it went, so that may be changing) that it will be a OLP majority, so numbers that seem to indicate this, I would trust. I don't believe people will be as divided come election time as the polls have been showing.

      Delete
    3. Hudak has somehow energized the Liberal voters who were tired and needed to start over. The PC's should have won this in a cake walk but decided to bring in American Tea Party political consultants from the states to help with the campaign, Along with Kim Campbell and the PQ recent campaign, this may end up being one of the worst run campaigns in Canadian history.

      Delete
    4. Don't worry 11matt11, the PCs will win on June 12th.

      Delete
    5. Bebe, it's looking increasingly less and less likely you'll be right. The PC's may actually get trounced on election day, and they'll have no one but themselves to blame.

      Delete
  2. So the big loser is really the NDP !!

    ReplyDelete
  3. you know, one thing that would be nice to see on this site would be the highest possible level of prediction, like that chance of each form of government
    Like Liberal minority = x%, PC minority = y%, Liberal majority = z%, PC Majority = w% ...etc....
    It might help with the view from 30,000 feet.

    ReplyDelete
  4. The overall numbers at Ekos show a PC bump in two consecutive polls in the rolling 3-day poll right after the debates and a slow NDP rise at the expense of the Liberals. It will be interesting to see if this is reflected in the Abacus numbers on swing voters.

    It looks like strategic voting is taking hold. The Liberals seem to be consolidating an anti PC vote in Toronto, which is what their campaign set out to do. The NDP is consolidating the anti-PC vote elsewhere. The PC's could end up winning the overall vote, but not have it reflected as much in the seat count because of the inefficiency of the vote.

    I'm all for strategic voting, so I hope this is the case.

    ReplyDelete
    Replies
    1. Re: slow NDP rise, those changes aren't statistically significant, so it may not show anything.

      Delete
    2. Oh you're referring to the longer term trend. Nevermind. My bad.

      Delete
  5. Re: the NDP's downslide in Toronto... Demonstrating a breathtaking combination of disastrous strategy with indefensible dismissal of progressive principle.

    ReplyDelete
    Replies
    1. The NDP's campaign seems to have been designed to drawn in SW Ontario soccer moms. They took for granted their traditional base, rust belt diehards, north of North Bay and downtown TO progressives. The last group seems to have given up on the party in big numbers.

      Delete
    2. It has been a very lousy campaign for the NDP, in what was their best chance at making a breakthrough since they lost power in 1995.

      Horwath's SW Ontario strategy is not paying off much either. The best they could do is hold the seats they currently have and pick up Windsor West. Hardly, what I call a breakthrough.

      Delete
  6. If I read the Ipsos numbers right, among likely voters (nothing short of the emergency), the PCs are down one point and the OLPs up three from their last poll. This doesn't sound like a PC rally to me. Now, when we all digest the Globe's "reasoning", surely Hudak's numbers will go through the roof! ;)

    ReplyDelete
  7. Ipsos now showing a tie between the Tories and Liberals among eligible voters but not much change with the Tories still 40 for likely voters.

    ReplyDelete
  8. Apparently there are new Ipsos number out:

    https://twitter.com/blissblogs/status/475037647321456641

    PC 35 Lib 35 NDP 26.

    As it is a twitter post I don't know if this is support or LV.

    ReplyDelete
    Replies
    1. With those numbers, I get:

      43 OLP
      43 PC
      21 NDP

      Delete
    2. The Ipsos poll found:

      Only 32% of Ontarians believe the Wynne-Horwath-McGuinty government should be re-elected.

      Signs are beginning to indicate Ontarians want change at Queen's Park.

      Delete
  9. So the real result Thierry is a Lib govt !!

    ReplyDelete
  10. For what it's worth, this looks to me like more of a convergence than a bounce. The Ekos results have had the Liberals implausibly high (esp. in E and SW) over the last several days and the NDP implausibly low (esp. in SW and probably N/C). This result seems much more in keeping with what other polls have been showing, save that the PCs seem implausibly high in the E - and their overall lead amongst eligible voters seems to depend principally upon that lead in the E.

    For clarity, let me stress that when I speak of "implausible" results, I am talking about the likely direction of sampling error, given the results of other polls. I am not just referring to my gut feelings or to local anecdote and I am, in particular, not making any suggestion of "partisan bias". Of course, I eagerly await new results from other firms!!

    ReplyDelete
  11. Lwt's prepare for whiplash the moment Eric folds the latest Ipsos poll into the mix since it has much higher PC and NDP numbers and lower Liberal numbers than the Ekos polling

    ReplyDelete
    Replies
    1. With the daily EKOS numbers, that isn't likely to happen.

      Delete
    2. Why do you give EKOS such weight? On your appearance on the Agenda the other night you said that one of the factors was the track record of the pollster. Frank Graves from EKOS is known to be a donor to the Liberal Party. Doesn't that affect how you interpret his results? He has consistently had higher Liberal numbers than the other pollsters.

      Delete
    3. Track record is defined by the performance of the polling firm in past elections, not any arbitrary considerations. Knowing Frank Graves a little, he'd saw off his leg to get his polls right.

      Delete
    4. I don't see how the consideration I named was arbitrary. Do you evaluate the pollster based on their projection the day before the election or their projection the day the campaign starts? Pollsters only have to be right the day before the election and the rest of the time they can say whatever they want and there is no way to know if they are accurate or not. The narratives of the press are built upon poll numbers. So as Allan Gregg says in this video,

      https://www.youtube.com/watch?v=pyin16pcN0k

      polls don't directly influence voters but they indirectly influence voters by having an influence on the way the media reports about the campaign. If a poll frames the contest as a 2-way race, as in this Ontario election, people are less likely to vote for the 3rd option in a case of strategic voting.

      Delete
    5. Only problem though is his East numbers are swinging very wildly compared to the areas further west and south. (not long ago was there actually a Liberal lead). Considering, that Hudak will do absolutely nothing for transportation (except maybe regarding the Highway 174 in Rockland), the PC will be doing nothing regarding transportation and Hudak has publicly said he is against LRT expansion. People in the suburbs should remember this especially since Hudak would fund massively for the GTA but nothing for Ottawa. His second big snafu of the campaing

      Delete
  12. Eric, one can assume from a non technical perspective that ipsos may be disclosing the most recent truth. I'm sure ull say, since ur a numbers guy, that wait and see, but my non statistician gut says more PC in the next while.

    ReplyDelete
  13. New Forum poll LIB 39 PC 37 NDP 17 Green 6.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. Sorry, slight mistake! The numbers I get are actually:

      51 OLP
      41 PC
      15 NDP

      Delete
  14. This might be my political bias creeping in but I've never trusted poll number less then I do now. There is something very wrong with what's being reported.

    As I've stated I am and ONDP candidate this election, so I know what some people might say about my opinion... but with no two pollster getting similar results it tells me there is something wrong. I honestly don't know what the result will be on June 12th and the polls are making the situation more confusing for many.

    What I do know and see is there seems to be a narrative about this election being driven by main stream media outlets, which is then conveniently being supported soon after by polls. Problem is each outlet has a different narrative they are pushing and each one has polls to back them up.

    So forgive the quote but 'Something is rotten in the state of Denmark'

    ReplyDelete
    Replies
    1. I've thought that too. However, polling firms risk their reputation if they get it wrong before the election.

      I will say that polls converging to the same number at the end of the election seems to suggest that there are political shenanigans going on early in the campaign (when numbers cannot be confirmed) and less so as the vote gets near.

      Delete
    2. Don't Count 100% agree. From these polls very little if anything can be taken as reality.

      Delete
    3. I agree, the polls are up and down far too much, I am beginning to think this fluctuation is best explained by bias of polling firms. The Ekos poll above is a case in point; the OLP gains 5.5% when likely voters are counted but the PCs gain only 1.5? That just does not make sense since, we know Liberals derive more support from young people than the Tories and young people don't vote!

      Delete
    4. I believe it is, indeed, your political bias creeping in - but it isn't necessary wrong. These polls could certainly be off the mark, there will be no way to tell for sure until e-day. I would honestly reserve judgement until then.

      Delete
  15. Eric, you seem to high very high regard for EKOS but they consistently give the Greens unrealistically high numbers.

    Why is that? And why do they not correct it?

    ReplyDelete
    Replies
    1. I wouldn't say I hold EKOS in any higher regard than other firms, just that I know Graves wants to get it right, not help a party people assume he supports.

      EKOS would argue that the Green support is accurate, just that Green voters don't turn out to the polls (that is why the party is lower in their likely voter tally).

      Delete
  16. What we are really seeing here is NOTHING !! No public commitment to any specific party. My guess is the result will be probably a Liberal minority govt. The ultra radical right hasn't managed to really swing the view. The ultra left has more or less collapsed so it will be through the centre.

    ReplyDelete
    Replies
    1. Peter,

      Don't worry I hold Ontarians in high enough regard that they will toss the ultra-self-serving-tired-mega-corrupt government out!

      Delete
    2. bede I doubt that very much. Will be as it was before I think.

      Delete
    3. All it takes is a handful of seats to shift and the PCs have a minority. True the Liberals may pick up two seats from the NDP in Toronto, but the PCs may pick up a few NDP seats outside of Toronto (e.g. Niagara Falls) which cancels them out, so we're back to that handful needing to shift from Liberal to PC. I suppose we'll probably get a better picture into next week as we close in on election day, but with what has happened in BC and Alberta I'm not even sure I would be willing to predict anything the very night before. I personally think we're just as likely to see a Hudak minority as we are a Wynne minority, with *maybe* a slight edge to Hudak if only because the pollsters seem to be coalescing around what looks like a post-debate bump. When EKOS has been showing a consistent 5-7 point lead for the Liberals during this entire election campaign only for it to become a statistical tie with the PCs that has been holding after the debate, you something is up. We just have to wait to find out what it is I guess.

      Delete
    4. Nick this election has turned into a "can't predict, can't poll" event.

      I watch a lot of TV and the political ads are interesting right now. What I think we will get is what we have right now, a Lib minority with NDP support. Sure a little adjustment in seat numbers but basically no real change. Which just may indicate the public actually doesn't want change ??

      Delete
  17. These polls! We got Forum giving the NDP 17% and we got Ipsos giving them 26%. Pollsters can't seem to agree if they are on a surge or if they are on a decline!

    On another note, finally Forum had some individual riding polls done. Some interesting numbers. Most of them seem just about right, except that the Liberals are leading Cambridge by 12 points and the PCs are at 30% in Thunder Bay Atikokan (probably an error).

    ReplyDelete
    Replies
    1. I don't think those were individual riding polls, rather I think they just applied a weighted swing model to estimate what they perceived to be the level of support in those ridings.

      Delete
    2. They were individual riding polls. Details here: http://poll.forumresearch.com/post/59/18-on--swing-ridings-polled/

      Delete
  18. Once again an oversample of rural voters in the East. 56% in today's poll. Average should be around 40-42 to 30-32. Also if the Kanata area would have it's own riding like it will be (or almost) in the next federal election, that riding would have been a toss-up not a PC stronghold. Because of all the rural areas they will be stuck with a Tea Party/radical right wing MPP

    ReplyDelete
  19. Ontario sounds like will be condemned with a minority government. I'm not sure if I've said this, but the PC would have to be ahead by a couple of points (maybe up to 5 points) if they would get a minority, because of the rural ridings they would win with 50% or even in some cases 60% to 70% of the votes (think the Muskoka, Haliberton, Seaway Region and Upper Ottawa Valley), while the Liberals will win many ridings (GTA area) with much closer margins.

    ReplyDelete
  20. I'm wondering if a discussion here about what went wrong with the parties might not be worthwhile??

    Did Hudak's "fire 100,000" hurt him? I think the link to Mike Harris may have indeed.

    Did Wynne's admittedly poor debate performance hurt the Libs. Got a hunch that it actually hasn't to any measurable degree. Gas plants and other scandals haven't either as the public knew about them well before the election.

    How about Horvath's shift to the right and essentially lack of any platform hurt her ?

    Questions worth discussing IMO

    ReplyDelete
    Replies
    1. No it didn't. Politicians should be honest. If Hudak had said nothing then got elected and eliminated 100,000 positions voters would have felt betrayed. One need only look at Gordon Campbell and the HST fiasco to see how damaging such a statement or lack thereof can be on a political career.

      We are still seeing the effects of Wynne's poor debate performance, everytime the scandals are brought up it hurts the Liberals both in terms of potential voters and energizing their base. This combined with the "business as usual attitude" has probably condemned the Grits to a reduced minority at best. The Liberals needed a better plan and narrative on why they should be re-elected and how they would make Ontario stronger-spending more money is not a compelling narrative especially when juxtaposed against a Government that has been wasteful with tax dollars.

      Horwath had the most difficult job this election. People such as chirumenga have criticised her for trying to be mainstream and appeal to the "middle voter" but, the evidence is fairly conclusive that most Ontarians and Canadians will not vote for a NDP government based solely on their principles and policies. As Einstien famously quipped: "Insanity is doing the same thing over and over again and expecting different results". Therefore, I can't blame her for trying a new approach-it does not appear to be successful but, we know almost with metaphysical certainty that a "traditional" NDP campaign would also not be successful.

      Delete
  21. No update today Eric with all the new polls out? It is, but a pity!

    ReplyDelete

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.