Tuesday, September 13, 2011

Slim McGuinty lead means slim McGuinty majority

When Harris-Decima came out last week with an 11-point gap between the Liberals and the Tories, people scoffed. Indeed, that size of margin was, perhaps, a little on the large side, but two new polls from Ipsos-Reid and Nanos Research show the Liberals ahead, though barely.

The Ipsos-Reid poll for Global TV, Newstalk 101, and The Ottawa Citizen gives the Liberals a virtually non-existent one-point lead, 38% to 37% for the Tories. Compared to Ipsos-Reid's "flash poll" conducted in early August, however, there has been virtually no change. The Liberals are up two points and the NDP one, while the Tories are down two points. All within the margin of error.

The Liberals lead primarily because of the GTA, where they have the edge 40% to 33%. They trail the Tories in most other parts of the province, except the north. There, with a small sample size, the Liberals lead the NDP by 44% to 30%.

Interestingly, Ipsos is tracking how likely people are to vote. About 59% are absolutely certain, which sounds like a plausible turnout rate. Unfortunately, they didn't report the voting intentions of these people specifically, though looking at the crosstabs it seems the Tories hold the edge among decided, absolutely certain voters.

Nanos, in its poll for CTV, CP24, and The Globe and Mail, is showing a lot more change. Compared to their poll from last week, the Liberals are up 6.2 points to 38.1%, ahead of the Tories at 34.7% and the NDP at 24.3%. That is a drop of only 0.7 points for the PCs and a gain of 1.5 points for the NDP. Again, within the margin of error.

But where are those 6.2 points coming from for the Liberals? Last week, Nanos had an extraordinarily high "Other" result of 5.8%. It is now down to 0.2%. What happened there? Are we seeing Liberal gain or was there something wrong with the last Nanos poll?

Understandably, these two polls (the first taken since the campaign has started) have swung the projection quite dramatically.

The Liberals have picked up four points and now lead with 36.2% of the vote, ahead of the Progressive Conservatives who stand at 36.1%. That is a drop of three points for the Tories, a gain of four for the Liberals, and a very significant seven-point swing since yesterday's projection.

The New Democrats are up 1.4 points to 23.3%, while the Greens are down 2.4 points to 3.1%.

This results in the Liberals picking up 18 seats. They are now projected to win 55, only one more than is needed for a majority government. The Progressive Conservatives have dropped 17 seats since yesterday and are now projected to win 32. The New Democrats have picked up one seat and are now projected to win 20.

Most of the changes have come in and around Toronto. In the GTA, the Tories have dropped nine seats, eight of which have gone to the Liberals and one to the NDP. The PCs have also lost four seats in Toronto, all of which have gone to the Liberals. The Liberals also picked up two seats from the PCs in Ottawa and eastern Ontario, two seats in southwestern Ontario, and one seat in the central part of the province.

This means that the Liberals are now projected to win the majority or half of the seats in every part of Ontario except the central and northern parts, as well as the Hamilton/Niagara region. What a difference a day makes.

As you can see, this is the most dramatic shift in the campaign so far. But it is part of a trend. The Tories have been losing support progressively over the last month, while both the Liberals and New Democrats have been on the upswing.

Because of such a dramatic swing, a lot of the seats that were close last time are no longer very close. The Liberals now lead in 11 close races and trail in four, while the Tories lead in three and trail in 12. The NDP leads two close races and trails in one.

This puts the Liberal range at between 44 and 59 seats, so for the first time a majority is more than possible. The Progressive Conservative range stands at between 29 and 44 seats, so at the very least they can still expect to do better than they did in 2007. The New Democrat range is still 18 to 21 seats.
As you can see from the above chart, there isn't much over-lap in the PC/Liberal ranges anymore. They could still tie at 44 seats apiece, but with the numbers we're seeing it doesn't seem likely.
The chart above shows the daily polling averages of the campaigns so far. I am presenting the chart here, but in the future it will be updated in the right-hand column only.

No polls for the other campaigns have been released so far, so for now there are only pre-campaign monthly averages on the chart.

But for Ontario, we have some numbers. Dotted lines on the chart indicates days in which no polls were taken, or at least results for those days have not been released.

This chart is the same as the one I did during the federal campaign, it averages all polls in the field on a specific day. Unlike last time, every polling firm will be getting only one entry per day if they release multiple polls taken on some of the same days.

And, of course, here are the riding-by-riding projections. Hopefully we'll have some numbers out of Manitoba soon, as the projection was last updated at the end of July!


  1. Do any of these polls show any indication on why there's such a swing recently?

    I suspect it times with the writ being dropped and ads being launched; the most obvious interpretation to me is that Ontarians weren't thinking about provincial politics and were more or less reporting their federal allegiance until the campaign actually started, at which point they started thinking about provincial politics. But I'm curious if there's another narrative?


  2. As has been ever so common in the last two or three elections the Tories simply can't hold it together.

    So now we are seeing yet another Liberal resurgence. Within in sight of outright victory and well within the envelope of minority with NDP support.

    When will the Tories learn that Ont. may be Red Tory territory but it sure ain't CPC provincially !!

  3. So much for "Ford Nation".

  4. Brian,

    I think that is a plausible explanation. Another is that the pre-campaign voting intentions may have reflected more of an anti-McGuinty sentiment than a pro-Hudak sentiment.

    Now that Ontario is paying attention to the campaign, perhaps that alternative is not what was expected.

  5. Wow, what a swing. It looks like "foreign workers" may be the religious schools of the 2011 campaign. It was unbelievably bad politics for Hudak to play that card. Sure, it is red meat to the xenophobic old white men based of the party, but when the key to winning this election was the 905 and its large population of first and second generation immigrants, it stomped all over chances of winning there.

    I'd think the overall swing is due to the combination of the foreign workers business and the voting public starting to pay attention now that we are in the actual writ period.

  6. How can the projection change so much so fast????


  7. I really don't think your seat projection will be correct if the popular vote is like that. Mark my words - if the province wide popular vote in Ontario has the Liberals and PC within a point of one another - you will NOT see the Liberals with a 55 to 32 lead in seats. This is NOT Quebec where you have the Quebec Liberals piling up 97% pluralities in selected non-francophones seats. I just don't see the tory vote being that "inefficient". If the popular vote is close - the seat could will be close as well. take it to the bank.

  8. asdf,

    Consider that the new polls are the only polls that have been conducted during the campaign so far. The projection yesterday was based only on pre-campaigning polling, so these two latest sets of numbers from Nanos and Ipsos-Reid take up a lot of space in the projection.


    We shall see!

  9. Thanks for doing this Eric.
    I think that you should do a projection for each individual poll. Let the reader balance it or average it over regions.
    Otherwise you're repeating the mistake from the federal election.
    Remember polls lag, but especially so during the election. By massaging numbers, they lag even more and thus reflect the situation maybe 7 days ago.

  10. That was the problem in the federal election, but I've taken steps to address the problem. The Nanos and Ipsos polls alone, for example, take up more than 70% of the projection.

  11. One thing that seems to me to be becoming clearer all the time is that Federal politics and Provincial politics are not the same thing !!

    So projecting a victory in a Province based on the Federal results really is not very smart.

    But there are still those trolls out there who keep trying

  12. The federal results do give us ideas of where each party's relative areas of strength and weakness lie.

  13. Seems clear that the voters sleep and ignore provincial vs federal until an election is close enough to see signs up, then the average voter finally goes 'oh, there is a difference'?

    After seeing paper candidates win seats federally it is clear that things can change quickly and drastically and that the leader/general campaign for a party makes a major difference.

    The PC's are going negative right off the bat. The Liberals are doing the Harper thing 'keep the boat steady, don't rock it'. The NDP is playing the nice card ala the federal campaign (might say negative stuff, but quickly follows with either here is what we'll do or 'but we are nice'). The Greens, as federally, are completely ignored except by the odd member of the media who feels a party running in every riding deserves coverage.

    Wonder how the PC's would feel about FPTP if the end result is them winning the percentage vote (which under the plan we had a referendum on would've given them power) but losing the election to a Liberal majority? (ala Manitoba's current projections)

  14. Tim Hudak may portray himself as a nice family man (even has a picture of his wife and kid on his campaign bus!), but his politics is as dirty as it gets here in Ontario.

    Hudak hammered away claiming new Canadians are "foreign workers" at a time when voters are just tuning in to the Ontario election. This is definitely not a good first impression on Hudak.

    Hudak needs to stop talking about McGuinty and start talking about himself if he wants to get anywhere. The Liberals are campaigning on their record. The New Democrats are campaigning on being different. But the Tories, are campaigning on how bad McGuinty is. It's gotten to a point that McGuinty is starting to look good, and that isn't an easy feat!!!

    - Maple

  15. A big change from just a few days ago, but not altogether unexpected.

    As dull and unaimaginative as McGuinty is (and he is both), the province has been doing relatively steady economically, and a good alternative policy was needed to unseat the government. Instead, PC and Hudak have been playing the soundbites and then latched onto the "Immigrant Training" issue. The policy, which was poorly thought out and controversial, had the potential to become a wedge issue for the PC. By mishandling it with the "foreign workers" phrase, the PC have now turned it instead into a Liberal wedge issue. Meanwhile more important policies are not being seriously addressed.

    The race is still open to all the three parties (the NDP has not had any major drops in support) and it will be interesting to see how this will unfold. Anything, from an outright win by any of them to a "hung parliament" is possible.

    Eric, keep up the good work.

  16. I said as much in my post in the last article here. A tie vote=Liberal Majority or at least a significant seat lead over the PCs and a strong minority of ~50. If the NDP can stay in the mid 20s, 25-30 seats are within grasp.

  17. I heard that Hudak channeled Rob Ford recently by saying in a CTV interview "the war on the car is over". Not sure if that will play well at the moment.

    On the other hand, his campaign strategy seems to be more like Smitherman's, as in it largely consists of telling everyone how awful the frontrunner is, while promising to do mostly the same things but not as awfully.

  18. Could it be that the pollsters report the decided voters? So a lot of people didn't decide until now and you could see more change in the polls. PC voters are more firm than Liberal voters who might decide to vote for them at the last minute. I think that more people decide later other than switching their vote.

  19. Haha a new poll came out today. Abacus Data says that the PC is beating the lieberals 41 to 32! Bye bye Mcguinty!!! No more useless taxes for Ontario and a brighter future!!

  20. Eric, I wonder what you think of the Abacus "poll" put out tonight? Its methodology looks kinda goofy.

  21. Recent poll by Abacus, will it be included? it looks at Likely voters. Its much different that these polls. Confused

  22. Yes, that new poll by Abacus is interesting. A very different method of gauging voting behaviour.

    I think I will include it, but I will take a closer look at it and discuss it more at length tomorrow.

  23. I'm not sure that the Abacus poll has any ability to compare with polls asking the usual voter intention question. The approach is just to different. Also, I would think it would skew to over-represent devoted partisans, of whom the PCs are likely to have the most.

  24. The Abacus poll is interesting, but doesn't pose a surprise. The Tories usually have a stronger base, since there is only one option for right-wing voters. Centrist voters can choose between the Liberals and Tories, while left-leaning voters can choose between the NDP and Liberals.

    Also, the Liberal "surge" was only strongly evident this week. It is no surprise these voters aren't firmly picking the Liberals.

    - Maple

  25. Abacus Data CEO, Dr. David Coletto. University of Calgary PoliSci doctorate. Pollster for Sun News. Co-author of a report with Tom Flanagan.

    Maybe his work epitomizes impartial polling, but I'll take a large grain of salt with the Abacus results until I see those numbers confirmed by subsequent results from other pollsters.

  26. On the other hand, Abacus' CEO is apparently the "pollster for Sun Media" and also recently did an anti-CBC poll:

    Don't know if that's relevant, but makes me go "hmm".

  27. "On the other hand, Abacus' CEO is apparently the "pollster for Sun Media" and also recently did an anti-CBC poll:

    Which is a Canoe.ca publication and Canoe is owned by Quebecor. Does that tell you anything ????

    Which also means that Abacus is really an arm of Quebecor. And that REALLY should tell you something ???

  28. Having an arrangement to work with a media network does not make a polling firm part of that media network. Nanos isn't run by the Globe, Angus-Reid isn't run by the Toronto Star, Ipsos-Reid isn't run by Postmedia, etc.

  29. "Having an arrangement to work with a media network does not make a polling firm part of that media network. Nanos isn't run by the Globe, Angus-Reid isn't run by the Toronto Star, Ipsos-Reid isn't run by Postmedia, etc. "

    Which is incorrect given things that have happened south of us and here to a certain extent. Conservative pollsters, and they DO exist, are invariably tied strongly to Conservative media. Those you put up do polling all over the place on many different subjects. Does Abacus?? No I didn't think so !!

  30. Anonymous 12:28,

    Yes, Abacus does poll elsewhere. For instance, they did a poll on the monarchy for the Huffington Post back when the royals were visiting. And they undoubtedly have many private clients for whom they do research.


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.