Friday, August 21, 2015

2015 Federal Election Link Round-up, Week 3

With the trial continuing to unfold in Ottawa, one might be forgiven for forgetting there is a national campaign happening! But the pollsters have started to weigh-in, so we're in it now.

Friday, August 21, 2015

- My analysis of the Duffy polls and the latest Poll Tracker update.

- The riding projections have now been updated to align with the latest Poll Tracker.

- I spoke on Maritime Noon about the latest Duffy polls.

- The Poll Tracker has been updated with the new Forum poll. Boosts the NDP in the seat count, the Liberals in the vote.

- With all the talk about kids lately on the parties' campaigns, I took a detailed look at what parents are thinking of the race.

- A couple polls from Abacus Data and Angus Reid on the Duffy trail. Potentially bad news for the Conservatives.

Thursday, August 20, 2015

- Some interesting riding polls conducted by Environics for LeadNow. I'm not sure how to handle them, since they are from an interest group. At this stage of the campaign, though, it is not too important to take riding polls into account. If these are still coming out in October, then I'll have some serious thinking to do. The vast majority of them line-up with the projection pretty closely.

- A CROP poll with some surprising results makes waves in Quebec. No, this isn't a rerun from 2011. Full regionals are here.

- I was on Power and Politics last night, talking about the latest numbers.

Wednesday, August 19, 2015

- Discussion sur le sujet des sondages: Format libre, Radio-Canada à Moncton (37:30)

- Exclusive, original content alert! I wrote a post here on ThreeHundredEight about the latest Saskatchewan poll, with regional breakdowns.

- To discuss the findings of his latest poll, Abacus Data CEO David Coletto joined me on the Pollcast. Don't forget to subscribe on iTunes so you don't miss an episode - we're currently recording two per week!

- Here's my deeper analysis of the state of the polls.

Tuesday, August 18, 2015

- The riding projections are now up-to-date with the Poll Tracker.

- Two updates today! The Poll Tracker has been updated again, now with the just-released Abacus Data poll. It has a heavy weighting because, when it comes down to it, the data is the only new data since the Léger poll that was out of the field on August 12 (almost a week ago). And look out for the new Pollcast episode coming soon - David Coletto of Abacus will be joining me to talk about his latest numbers.

- The Poll Tracker has been updated with the newest Nanos poll. These four-week polls pose a bit of a problem for the weighting model. But to let you know how I am handling them, I am treating each Nanos poll as if it is a new poll with 1/4 of the total sample. So, from the model's perspective, it is as if Nanos was releasing a weekly poll of 250 people.

- A few new polls were out today. Nanos has its weekly tracking out, though again we're still talking about a poll that includes almost two weeks' worth of pre-writ information. Forum has a poll for Spadina-Fort York, showing Olivia Chow leading Adam Vaughan by a wide margin. It matches the current projection almost perfectly, which is a good sign that the new by-election methodology is on the right track. And finally, a big-sample poll for Saskatchewan showing a close race between the Conservatives and the NDP.

Monday, August 17, 2015

- I looked in detail at the race in Quebec, thanks to the regional breakdowns of Léger's new poll.

- The riding projections have been updated to match the latest numbers from the Poll Tracker.

- The Poll Tracker has been updated, incorporating the latest numbers from the Léger poll. The Conservatives have taken a hit.

- This latest update also includes a change in how by-elections are handled. After doing some testing, I have settled on a better method than one that relied entirely on the results of the by-election, where the polls were in that region at the time, and how things have swung since then. That is still a consideration, but now the previous general election is also taken into account. The weighting applied to the swing from the general and the by-election is now determined by the turnout in each. One immediate impact of this change is that the Greens' chances in Victoria and the old Calgary Centre riding have been reduced.

- On The House, we discussed battleground Saskatchewan. That shows how weird this campaign is - Saskatchewan is a battleground!

77 comments:

  1. The new by-election approach makes a lot of sense. Hopefully it works out well post election when you test your accuracy. (I'd be interested for your comment on this change when you do your post election analysis).

    ReplyDelete
  2. Why are the NDP put so low in Atlantic Canada, the Leger and EKOS polling had them practically tied and they have been constantly higher then the Conservatives in every poll.

    ReplyDelete
    Replies
    1. They are still higher than the CPC, but Forum and Mainstreet had the NDP in AC at 24% and 28%.

      Delete
    2. The Nova Scotia NDP did a lot of ancillary damage to the brand. They came to office with overly high expectations and failed to live up to their own rhetoric. The economy stumbled and the NDP was not able to generate renewed immigration, opportunity or business or keep graduates in Province. The NDP will win win Halifax and Stoffer's riding but, Bobby Chisholm in Dartmouth is in serious trouble and it doesn't look as though they'll be able to pry a Cape Breton seat away from the Liberals.

      As an aside I was in Victoria last week and had the chance to speak with friends who all thought Murray Rankin may be in trouble as well. Apparently, they have not recruited as many volunteers as they expected. it proves nothing of course but is an ominous sign if nationally the NDP is on a bit of a tear yet, in perhaps their safest seat they are unable to attract expected volunteer numbers. Those who I spoke with had the general consensus, which I think is verified by some polls, that at this juncture NDP support is wider than deep and a lot of those who at the moment tell pollsters they'll vote NDP may change their mind.

      Delete
  3. What will be interesting is to see how things change if the Conservatives drop into 3rd and stay there for a while. Will more voters feel 'safe' to vote however they want thus helping the Green Party and causing more despair in the Conservative camp which could push them lower and maybe help get more votes to any alternative to them out there (Libs & Greens might be boosted, Libertarians too).

    ReplyDelete
    Replies
    1. It would be nice to see the Libertarians manage more than a few hundred votes.

      Delete
  4. The real question is how much damage the ongoing Duffy-Wright trial will bring ?? Pretty heavy right now.

    ReplyDelete
    Replies
    1. It will bring a fair bit of damage. I expect the Tories to hit a low in the coming days but, perhaps, that is the genius of a long campaign?? The juicy bits of the trial will be over by Labour Day and the Tories will be able to focus on the economy and Mulcair and Trudeau's lack of experience with economics, leadership, running a large organisation and competence.

      Delete
    2. Yet the Liberals "I'm ready now" ads are resonating very well with the public.

      Delete
    3. Which is surprising, given his substanceless public statements. His closing debate remarks were all about how much he loves Canada. He wants to grow the economy from the heart outwards? What the hell is the practical application of this?

      I want numbers. I want a political candidate to give me a plan with numbers, and not dress it up with all this touchy-feely crap.

      If one (just one) of the party leaders would release a detailed platform that involved real spending cuts and undo the damage of Harper's stimulus package (which created a ton of new apparently permanent spending), I'd jump up and down and sing his praises.

      And none of them are doing that.

      We need Paul Martin from 1995. We need someone who will commit to balancing the budget in a single year.

      The government of Canada has a spending problem. It needs to reduce expenditures in a variety of ways. My first choice would be subsidies and corporate welfare and supply management - those should all go away.

      Delete
    4. At the moment they may be playing well with voters but, you can bet your bottom dollar when voters get to the polling both those currently undecided will break in the Tories favour. Frankly, Trudeau did himself a fair bit of damage and contradicted his own "I'm ready now message" with his promise to grow the economy from the "heart out". Nobody knows what that statement means and it sounds like something an 8 year old girl, without much experience in economics, would come up with.

      Delete
    5. I'm glad, though, that he seems to have caused the invention of the term "care bear economics".

      Delete
  5. Hi Eric,

    I know it's a likely gargantuan pain but do you happen to have the a graph available of riding projections over time? I know some, like mine in Courtenay-Alberni have changed or are new, but it would still be very nice to see how things are potentially swinging in the ridings themselves for party support from the last election and through this one. If your projections are to be believed, formerly stalwart Conservative ridings on Vancouver Island have been cleaved and lost to the NDP. Interested to know how much of that is due to redistribution and how much to voters minds changing.

    Thank you so much for your tireless work here.

    Chris

    ReplyDelete
    Replies
    1. Sorry, I don't have any tracking charts like that. But the trends, at least, would mimic those for BC as a whole.

      Delete
  6. These Mainstreet/Post Media polls seem to be slanted way towards the Conservatives. Eric, is there a reason for that? Is their polling science or way of collecting different from every other poll?

    It doesn't look good on them considering they represent the right side of the political spectrum media in Canada. It makes their polls seemed biased. This seemed also to happen during the Alberta election. Even the Fox News polls in the States seem to be in line with everyone else.

    What's going on with the Polls at Mainstreet? It makes their company look bad and untrustworthy.

    ReplyDelete
    Replies
    1. I don't know what is behind it, but it is hard to say it is a consistent trend after just two polls.

      Delete
    2. Do you know if any pollsters have switched to a likely voter model now that the writ is dropped?

      Delete
    3. A model that is explicitly reported - no. But some of them may have a model intrinsic in their weighting formula.

      Delete
    4. 11Matt11,

      I do not understand why you are of the belief that somehow Mainstreet's numbers are skewed or biased towards the Tories? From what I can see they fit broadly with a trend that originated over a year ago. When ever multiple polls are conducted for a singular event it is expected that some polls will be at the high end of the range and some at the low end. A 31/30/29 split in favour of the Tories corresponds with the poll tracker very well. I simple do not see the reason for your skepticism of Mainstreet. Please explain if you have a moment.

      Delete
    5. If I can jump in, Pete. One ought to be skeptical of a firm that shows the Cons at 38% (on 21 July, when all other polls had them around 30) and then on Mainstreet's poll of 11 August, has the Libs in Quebec (decided only) at 30% - 2% behind the NDP - when all other polls have them well back. In fact, check out some more regional and other oddities reported by Mainstreet : Alberta - Cons 51, NDP 19 (!!), Libs 23; Atlantic - Cons 28, NDP 23, Libs 43; and in the 18-34 age group (includes undecided, no figures for decided or leaning and decided) - Cons 27, NDP 23 (!!), Libs 28. The 11 August poll doesn’t seem strange until you look at it this closely, and then so many numbers, like these, seem “off” (another example, are women really polling Cons 22, NDP 23, Libs 24? That seems very tight compared to other polls).

      Delete
    6. The other way to view this is that Mainstreet is doing a good job of sharing all of their results. It would be worse if they stopped sharing outliers.

      Delete
    7. Chirumenga,

      You are really comparing apples and oranges and failing to put the data into proper context. Should one also be skeptical of Abacus who have the NDP at 35? nationally?

      Remember the Tories at 38% or 30% comes with a margin of error and if I recall correctly the 38% was of decided voters so, if undecideds were included the Tory number would go down. 51%-19% in Alberta for the Tories and Dippers respectively sounds perfectly normal! In fact 19% would be the NDP's best ever showing in an election in Alberta but, yet you somehow manage to twist it around in order to discredit the pollster! Shame on you! You've unfortunately taken the numbers out of context. The Atlantic numbers look broadly in the range of what we have seen over at least the last 2 months if not longer-once again you seem amazed with the NDP numbers but, the Atlantic is not where NDP strength lies and with the hash the Newfoundland NDP created among themselves and the former NDP government in Nova Scotia still leaving a bad taste in people's mouths it is hardly any wonder why NDP numbers in the Atlantic seem to be stagnant. I agree with you that the "women numbers" do seem off to me but, is that bias or the 1/20? outlier? In short, I think you are too hasty to condemn and I do not think your conclusion justified.

      Delete
    8. ap4413m In principle, you're right, but if Mainstreet stopped sharing their federal outliers, they'd have nothing left.

      Delete
    9. Pete, I didn't mention "bias" - it's quite reasonable to assume Mainstreet is just incompetent. But I thought it was fair to provide the grounding for 11matt11’s post. My remarks were based on the trends, not on any poll in isolation - in fact, that was explicit in my comment. I compared the Mainstreet poll with the other election polls, all of which show more in common with each other than with the Mainstreet poll. So, take your suggestion that the 35% for the NDP in the most recent Abacus poll is just as worthy of skepticism. But of the nine polls conducted during the election, two have the NDP at 33, one at 34, one at 39 and finally the Abacus figure of 35. Outlier? Hardly.

      As for decided/leaning vs. undecided. All the poll numbers I gave were for decided/leaning (except where those figures were not supplied). So, I'm comparing apples with other apples. The point, again, is that the trends do not support Mainstreet's results, which in fact, diverge strongly from the trends, whether nationally or regionally. The divergence is less obvious in Mainstreet’s most recent poll, but in that case, it’s concealed in the regional and demographic figures, which are just as odd as the 38% for the Cons was earlier.

      Finally, I’m amused that you think I should be ashamed of my criticism of a poll/polling firm. Since when is fair criticism a reason for shame? (Admittedly, maybe in the Harper era that has changed, what with his response to inconvenient scientists, et al).

      Delete
  7. And now we have Nanos out with a poll that has the CPC in the lead?? What is going on ?

    ReplyDelete
    Replies
    1. I like Nik's methodology (live calls - no prompts), but he's not polling enough people in a short enough time.

      Though, what that means is he's showing an extremely close race dating back to before the writ drop, which other pollsters were not showing.

      Delete
    2. The Nanos poll is like looking in the rear-view mirror (including some of that convex distortion, I reckon - I think they're overrating the Cons)

      Delete
    3. What's weird is that he's showing Con strength and Liberal strength, and those two things didn't happen at the same time.

      Delete
  8. Eric why are you counting the Leger online panel numbers as a poll. An online panel is really not a poll by any standard. It's an online panel and its self selected. it doesn't matter if the numbers are weighted to census data. It's just not a poll at all.

    ReplyDelete
    Replies
    1. Online polls are not the same as some poll on a website. Online polls from companies like Leger are completely legitimate and reliable.

      Delete
    2. I know its not the same as a web site poll but it is a panel of people who answer polls for Leger. It's not random. I am aware of how they work.

      Delete
    3. The weighting ostensibly counter-acts much of the selection bias.

      Delete
  9. Is anyone doing a handful of swing riding polls this time? I know there were a handful done last time and it is always interesting to see what those riding polls say vs national as we all know it will be a handful that will shift things drastically again.

    Up here in Thunder Bay we have a 3 way Green-Liberal-NDP race with a Conservative hoping for a perfect split but the riding projections don't show that as no public polls of this riding have happened. There are probably a dozen of those ridings in BC and I'm sure there are some interesting ones in Quebec this time for NDP/Bloc/Liberal/Conservative (some mix of the 4). I think the Globe had a handful they watched closely last time.

    ReplyDelete
  10. Re your comment on CPC gaining seats in Quebec: When you've hit bottom, the only way is up.

    Same applies for LPC in Ontario.

    Will the LPC recover more seats than the CPC in Quebec?

    The recent movement in the numbers looks a lot like statistical noise.

    Maybe for the CPC we're seeing a UCCB uptick followed by Duffy trial drip, drip.

    ReplyDelete
  11. Hey Éric,

    I understand it may be low on your priorities, but I've been dying to read an analysis of Bruce Hyer and to a lesser-extent Jose Nunez-Melo's ridings. I just cannot believe that the NDP would be trending as high as they are there. Any plans on that? Fingers crossed we get some polls to help give a better look at the situation. Love following your work every day, keep at it! :)

    ReplyDelete
    Replies
    1. Really need some riding polls.

      Delete
  12. Given what Eric said on tonight's Power and Politics the polls are at this point essentially tied.

    Where things go from here is hard to say ??

    ReplyDelete
    Replies
    1. With the Liberals showing strength for the first time a year, and the Tories clearly not going up any time soon, and the NDP support still fairly new for them, this could go any direction.

      September should be informative.

      Delete
    2. Again though we are getting polls showing significant strength for the NDP. At least based on what Eric said on today's CBC news broadcast at noon.

      Delete
  13. A new CROP poll in today's La Presse has the NDP way out in front in Quebec: 47% NDP, 20% Liberal, Bloc 16% Conservatives 13%. With those results, the NDP would take seats from all three parties, the Bloc would disappear and all Conservative seats in Quebec would be vulnerable.

    ReplyDelete
    Replies
    1. Trudeau's seat might even be vulnerable. His attacks on Mulcair at the Macleans Debate may have backfired on him. He only got 37% in 2011 and the NDP got 29%. It all depends who the NDP nominates there. There last candidate had to step down due to sovereigntist comments. If they can find a star candidate they could conceivably take down Trudeau,

      Delete
  14. with that new Environomics(sp?) poll & Quebec poll, its starting to look like a NDP government in our future

    ReplyDelete
  15. Not sure of the "special interest" behind the Environics poll, but what stands out is that the Conservatives are down significantly in every one of the 13 Swing Ridings, Looks like a "throw the bums out" tide taking shape.

    ReplyDelete
  16. On the balance of probabilities there will be a new government elected October 19th.

    One question is, of course, will it be minority or majority in nature?

    Other key questions remain: Will the Conservatives and Liberals co-operate to defeat an NDP minority government? How long
    should they wait- 18/24/30 months(if at all)?

    Change is upon us. This much is certain.

    ReplyDelete
    Replies
    1. I'm betting that if Harper gets the most seats but not a majority he will force an election ASAP as he will have enough cash for a full campaign from the rebate alone while the Libs & NDP will both be in debt most likely. Ugly but I am trying to make sure everyone up here in Thunder Bay is ready just in case - ie: hold onto signs and office as a minority situation could easily result in a fast election if the NDP & Liberals aren't smart about it (and sadly history shows they won't be).

      Delete
    2. Actually if you are dealing with probabilities the Globe and Mail's algorithm, aka a probability model of 1000 election simulations in each riding, the same way Nate Silver projects winners, shows a 58% chance of the Conservatives winning.

      Eric's aggregate is not a probability model. It is a rough average with some weighting thrown in, and also counts "polls" that would not be considered valid in US models.

      Not to disrespect Eric, but I don't believe he is a statistician. He created a popular poll aggregate. It's not statistics.

      Delete
    3. Which polls would not be considered valid in the US?

      Correct, I am not a statistician (and I've never claimed to be one!).

      Delete
    4. I don't think the Americans count online panels.

      Delete
    5. JOhn Northey - I would expect the Governor General to refuse such a request, and instead offer the leaeder of the second party (probably Mulcair in that instance) a chance to form government.

      Only if no one was able to gain the confidence of the House would we get another election, and I'm pretty sure Mulcair and Trudeau would back each other up enough to avoid that.

      Delete
    6. "I don't think the Americans count online panels."

      538 and Silver do (or did in 2012).

      Delete
    7. I'm surprised you didn't read the article in Aug 20th Globe globe by Eric Andrew-Gee about the state of polling and its decreasing reliability.

      " No less an authority than Nate Silver – the writer and statistician whose nearly perfect predictions of the past two U.S. elections made him a polling superstar – has warned of a sustained dip in quality.

      “Polls, in the U.K. and in other places around the world, appear to be getting worse, as it becomes more challenging to contact a representative sample of voters,” he (Nate Silver) wrote in a post on his website, FiveThirtyEight.com,"

      Mr Reid from Angus Reid talks about how online panels can be good for gauging response to something like a reading or product advertisement, which makes sense. Or for asking questions about sensitive past experiences like sexual harassment. Does it work for future actions such as voting intentions of the general populace? Probably not. It's an incorrect use of the tool.

      Critics warn that polls conducted through these panels can yield warped results, since respondents have volunteered to participate and might be more opinionated than the general population, or motivated by money.

      "Nik Nanos, CEO of Nanos Research, which does live-agent phone polling for The Globe, said two-thirds of the polls conducted in Canada wouldn’t pass muster for publication in The New York Times, mainly because they aren’t random enough."

      Abacus is not a random poll, nor was Leger's latest. They just aren't representative polls Eric. They are junk numbers. You get enough junk numbers in an average and they cancel each other out, but it still doesn't make that aggregate valid.




      Delete
    8. I did read the article, thanks.

      Nothing new seeing pollsters that use one methodology saying the methodology of their competitors is not up to snuff.

      Online polling is perfectly fine and has become an accepted method of polling in the industry.

      Delete
    9. Sorry to answer your question, Real Clear Politics, actually the closest analogy to what you do, doesn't count online panels such as Zogby. What Nate Silver does with his statistical modeling is nothing like what you do. Real Clear Politics averages polls which is what you do and they don't count online panels.

      Delete
    10. Eric, great post. Thanks for letting me know why you're doing it. I read on your other post you're not a stats guy, which I have to say I did not know. My bad. With someone who's not a stats guys, you do a great job analysing the stats. Keep up the good work.

      Delete
    11. As a many decades subscriber to the Globe and Mail, and going back even further than that with my family, I have to say I've started to get worried about the paper for the first time in my life. It started with their strange pro-Prentice reporting during the Alberta campaign, and has continued with this one. Like the Calgary Harald, somewhere they decided that their readers were older, so they should slant their political coverage to represent older more traditional conservative voters. That's rather shocking to those who have always enjoyed the Globe's objective reporting over the years. They've done better with their Duffy Trail reporting so hopefully that will seep into their regular campaign reporting. But I have to say, considering all the slants they're using to try and help out Harper in this campaign, I'm worried that this might be the beginning of the end of The Globe. Canada already has a National paper with a less objective right wing slant in the National Post. Why would the Globe want to got there to lower their pool of potential readers? And I've somehow found the Post's coverage even more objective than the Globe's this election, which is shocking.

      Delete
    12. After this week with the Duffy trial, it feels like what happened after the Prentice Notley debate, what happened when the Chretien sponsorship scandal broke, and what happened after Kim Campbell's ill faded TV. commercial. It feels like the downfall of a political party.

      It may take a while for it to be reflected in the polls, but I think it will eventually throughout the Canadian political spectrum. Many did not see the Alberta debate, but eventually the poll results of it became the poll results on election night. The sponsorship scandal did not rock the Liberals over night, but eventually it took them down to third party status. And most did not see the now internationally famous Tory commercial making fun of Chretien's face as it was taken off the air after only 7 viewings, but eventually it helped destroy the PC party as we knew it.

      It's starting to feel like that after the last two days of the Duffy trial, even if it might take some time for the polls to reflect it. This might end very very badly for Harper, as history seems to be repeating itself.

      Delete
    13. shoshana,

      I'm not too concerned with what RCP does, and no RCP is not the closest thing to what I do. RCP averages polls without any adjustments whatsoever. I think they just take the last three or five, and treats them all equally.

      I weight the polls by sample size, date, and track record. What Silver has done in the past is also add a factor related to pollster bias (easy to do in a two-party system with many elections, polls, and jurisdictions, virtually impossible in our system). He also takes into account demographics, which again is virtually impossible to do in our system.

      Delete
    14. 11matt11

      But I'm beginning to wonder if that isn't the result of rather than the force behind the countries slow movement to the right.

      And how much of that is the influence of the RICH and powerful who control far to much of the economy ??

      Delete
    15. Online opt-in panels, IVR polls, live caller polls and randomly recruited online panels all bring a bit of different information to the table.

      No one is able to create a true random sample anymore, there is always going to be some bias based upon the recruiting method. Some people screen calls and other people will never join an online panel. Mixing different methodologies should yield a better overall sample. Unless there is a very compelling reason to believe a pollsters methodology is systematically biased I see no reason to exclude them (In the past a case could have been made to exclude Rasmussen's results in the US ).

      Delete
    16. 11Matt11,

      I disagree that "change is upon us". The campaign is running about as smoothly for the Tories as they could hope. Yes, Duffy has been an unwelcome distraction but, it ended a week early! The NDP leads and much more focus will be placed on a campaign with more than a few holes. The NDP's difficulties in Ontario says it all, I would expect those numbers to get worse for the NDP. The Tories finished their worst week of the campaign with 30%-if that is their base and the only way is up-can we really be all that far away from a second straight strong stable Conservative majority Government? David Cameron won a majority on 36% of the vote with the NDP and Grits duking it out for opposition supremacy Harper may well pull off a similar feat.

      Delete
    17. Harper has proven himself to be extremely dishonest. Harper is heading for third place. Not so sure that is a smoothly running campaign. The CPC campaign by all accounts has been riddled with errors .

      Delete
  17. How do you handle riding polls when they are conducted by forum? I am also curious what you do if a poll does not release specific regional numbers but they release provincial ones? You mention something about this on the explanation page but I'd like some more detail if you could:


    "You might wonder why I don't adjust the model to project sub-regionally in Quebec. The issue with that is that no other pollster is releasing sub-regional numbers for the province. And in 2011, CROP exited the field over a week before the end of the campaign. It would seem unwise to turn the Quebec portion of the model into a sub-regional one, when in all likelihood I won't have information to plug into it."

    ReplyDelete
    Replies
    1. I treat riding polls the same way no matter which pollster conducts them.

      I'm not sure what you mean about provincial numbers.

      Delete
    2. So in Quebec you don't project sub-regionally. I am wondering what you do in other provinces if a pollster doesn't release sub-regional numbers but only provides provincial ones?

      I guess I am wondering if you can use a riding poll why there is no added benefit from the sub-regional results in Quebec provided by CROP?

      Delete
    3. I don't use sub-regional polls. I use polls with breakdowns for BC, AB, PR, ON, QC, AC, or any one of those.

      I use riding polls when they are saying something different than the model.

      Delete
    4. So what do you do if a pollster only releases results for all of Ontario (with no regional breakdowns)? Does that poll get ignored or do you still have a way to use it?

      Delete
    5. Polls done only in, or having numbers for, the following regions are used:

      British Columbia
      Alberta
      Prairies (MB+SK)
      Ontario
      Quebec
      Atlantic Canada

      So a poll done just in Ontario can be used. I don't use sub-regional breakdowns. A poll of the GTA, for example, is not used by me.

      Delete
  18. And now after this weeks devastating testimony you think soft Conservative voter will drift away Eric ??

    Sure hope you are right. Time to lose this lying, deceiving Govt !

    ReplyDelete
    Replies
    1. this may not be an ideal government Peter, but then few are.

      I'd much rather have Harper and a majority government than either of the pinkish alternatives!

      Delete
    2. Well Earl that's your personal opinion and that does not seem to be the National opinion as born out by the polls we're seeing.

      Delete
    3. Quite true Peter but, come election day I think Earl's view will prevail. Conservatives want Canadians to keep their TFSAs; Liberals and Dippers want o eliminate TFSAs and replace them with a payroll tax a la Kathleen Wynne's OPP (harebrained) scheme.

      Delete
    4. Of note Pete is how Canadians have shown they are willing to pay payroll taxes for defined programs over the years - even accepting high EI rates that are just going to pay for Conservative promises right now, much like people didn't scream bloody murder back when the Liberals used it to balance the books in the late 90's/early 00's. Payroll taxes hit those who make under $50k a year the hardest but they are also the least likely to complain about it as they know they are the ones who will need EI and CPP the most.

      Those who complain loudest about taxes tend to be those who make $100k a year and up, who also have the time to figure out things like how government spends EI funds on stuff other than EI. CPP on the other hand has always been dedicated to CPP thus most can live with it so the Cons fighting CPP like they are is a losing battle. I say this a bit sadly as I hate payroll taxes as I see them as a regressive tax on jobs.

      Delete
  19. I think that the Liberals are going to do better than expected. I think that the Liberals are going to win most of the seats in the GTA, because the NDP is unpopular except for a handful of ridings near downtown Toronto and Hamilton. I see Liberal signs in Toronto everywhere. The NDP will have a hard time winning without support in Canada's largest city.

    ReplyDelete
  20. The only reason the Liberals are doing as well as they are right now is because of the 905, and while there may be some room for growth there isn't much outside of that. The NDP do much better outside of the immediate 905 (including Hamilton and Niagara which are outer 905) than the Liberals do, and even then there are a few ridings the NDP could potentially pick up or retain in Peel and Durham. If the 905 was onboard with the rest of the country supporting the NDP, we'd be talking about a majority government right now, but the 905 on its own is not going to be enough for Trudeau to form any government at all.

    ReplyDelete
  21. The only reason why the Liberals are doing as well as they are in the seat count projections right now is because of the 905, and there's really not much beyond that for them to unexpectedly do better. Granted, while the Liberals may win the most seats in the 905, the 416 is a tossup between them and the NDP in Old Toronto and Scarborough, and everywhere in Ontario outside of the 905 - but including Hamilton and Niagara which on the outer fringes of the 905 belt - is where the NDP are the alternative to the Conservatives. While in the 905 will probably at least allow the Liberals to double their seat count from 2011, they're still looking at possibly losing seats in Quebec and perhaps even the few they have in Southern Ontario outside of the GTA to the NDP. It's not hard to do great when you've already hit rock bottom, but that still won't be enough to form government.

    Furthermore, if the NDP do win - which based on current numbers looks likely, but could of course change - they most certainly will have the support of Canada's largest city, as there are now and will remain several NDP seats in Toronto proper even if the suburbs vote overwhelmingly Liberal. Certainly, they will have more support in their minority government from Toronto than Harper has in all of Quebec with his majority government right now.

    ReplyDelete
  22. I posted this twice and didn't realize. Sorry about that guys.

    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.