Friday, June 27, 2014

Incumbents favoured in three of four by-elections (updated)

Four federal by-elections are being held on Monday in Alberta and Ontario. If voters in these ridings can be bothered to cast a ballot on what is a long weekend for many of them, the results should be interesting.

The ridings in questions are Macleod and Fort McMurray-Athabasca in Alberta and Trinity-Spadina and Scarborough-Agincourt in Ontario (both in Toronto). These four ridings have been very safe ones for the incumbent parties in the past, and in three of them the incumbent is expected to prevail. That leaves Trinity-Spadina as the race to watch.

The Conservatives should have no trouble holding on to their two Alberta ridings, which they won by gargantuan margins in 2011. Of note may be the performance of the Liberals, who put in some surprising showings, and almost an upset, in two rural Manitoba ridings at the end of 2013. Fort McMurray-Athabasca, with its fast growing population and large labour force hailing from elsewhere in the country, could be interesting as well. The Liberals have been polling relatively well in Alberta - compared to their usual support levels, at least - and it will be interesting to see if they can attract a large number of voters in the province like they did in Calgary Centre in 2012.

Scarborough-Agincourt is a Liberal fortress, having survived the purge of 2011. Some might argue it was a Jim Karygiannis fortress, but polls suggest Arnold Chan will hold on to the riding for the Liberals. Trinity-Spadina, vacated by Olivia Chow as she runs for the Toronto mayor's office, would normally be an NDP lock. But with popular local councilor Adam Vaughan carrying the Liberal banner, the riding could swing over to him. Voters in Trinity-Spadina just recently elected a provincial Liberal, defeating long-time MPP Rosario Marchese, who had represented the riding since 1999.

Let's take a look at what kind of results we might expect in these four ridings. The first chart below shows the high and low projections based on the last three polls conducted in Ontario. These were calculated using the site's standard methodology for making riding estimations.
By these measures, the Conservatives look quite comfortable in Macleod and Fort McMurray-Athabasca, while the Liberals look solid in Scarborough-Agincourt.

June 30 update: An Angus Reid national poll was released over the weekend. Adding it to the high and low projections, the Conservative high would be boosted to 56% in Fort McMurray-Athabasca. The Liberal low would be reduced to 35% in Trinity-Spadina, while the NDP high would be increased to 44% there. In Scarborough, the NDP high would rise to 16%, while in Macleod the Green low would drop to 6%.

In Macleod, the Conservatives have ranged between 62% and 70% when employing the provincial swing from the last three polls. The NDP, Liberals, and potentially even the Greens could claim second spot.

In Fort McMurray-Athabasca, the Conservatives look likely to be reduced from the margin they managed in 2011, but would still take between 48% and 55% against just 22% to 31% for the Liberals. Still, in a northern Alberta riding that would be a good showing for the Liberals. The NDP have the inside track on third place, but the Greens could narrowly take it.

The Liberals look set to take a majority of the vote in Scarborough-Agincourt with between 52% and 65%, against 23% to 30% for the Conservatives. The NDP should finish third.

And in Trinity-Spadina, the Liberals and New Democrats are in a close race that is leaning Liberal. Vaughan could take 36% to 51% of the vote, and Joe Cressy between 28% and 41%. The Conservatives, at 13% to 15%, would finish third.

But these are estimates based on province-wide support levels. In by-elections, local factors are extremely important. Forum Research has, as usual, tried to gauge the support of the parties in three of the four ridings (interestingly, Forum has opted not to poll Fort McMurray-Athabasca, due to the difficulties in reaching its transient, cellphone-only population). Let's see how Forum measures the three races, again using the high and low numbers they have reported in its two (Macleod, Scarborough-Agincourt) or three (Trinity-Spadina) polls released in the last few months.
Here we see some small variations from the estimates based on provincial swing.

June 30 update: Forum released some new numbers on the eve of the election. In Macleod, the poll would reduce the low Conservative result to 54% and the NDP low to 4%. The Liberal high would increase to 16% and the Green high to 16%.

In Trinity-Spadina, the Conservative low drops to 11% and the Liberal low to 45%. The NDP high increases to 35% and the Green high to 9%.

In Scarborough-Agincourt, the NDP high increases to 10% and the Green high to 4%.

Forum released its first numbers for Fort McMurray-Athabasca: 41% Liberals, 33% Conservatives, 13% NDP, 8% Others, 5% Greens. Not sure about those, but we'll see tonight.

Macleod lines up about the same, with the Tories a little lower than the provincial swing would suggest.

In Scarborough-Agincourt, Forum is less bullish on the Liberals and more favourable to the Conservatives, but still gives it to Chan by a fair margin over Trevor Ellis. The most recent Forum poll (June 18) gives Chan the advantage among all age groups, suggesting turnout might not matter.

Trinity-Spadina is perhaps the most different, if only because it doesn't give Cressy much of a chance. Vaughan has managed between 52% and 54% in Forum's three polls, against 31% to 34% for Cressy. In the latest survey, Vaughan led in all age groups except the 18 to 34 year olds, and 44% of former NDP voters said they would cast their ballot for Vaughan. That does not look very good for the New Democrats.

But just how reliable are by-election polls? They can be hit or terrible-miss. On the one hand, Forum has generally had a good track record in Toronto-area ridings. But calling cellphones is impossible for by-election polling, since it is not possible to ensure that the phones being called belong to people who live in the riding. A lot of people are cellphone-only in an area like Trinity-Spadina. The question is whether these people are different from the landline using population. Perhaps they are, or perhaps Forum has gotten lucky in the past. Cellphone-only residents might be less likely to vote in a more rural riding, but not in the downtown core of Toronto.

Macleod is unlikely to be gauged incorrectly by Forum, despite the company's bad record in rural ridings. The Conservatives will win here, but don't be surprised if Forum's estimations are off by double-digits. Nevertheless, in all ridings the numbers generally line-up with where the swing suggests they should be (that was not the case in Brandon-Souris in 2013, the big miss by Forum) so a bad performance may not be in the works.

Another complicating factor is turnout. Much was made of the decent advance poll turnout rates in most of the by-elections, particularly Trinity-Spadina. But was that because of high interest, or because many voters knew they would be out of town on June 30, the day before Canada Day? If turnout is low, there is a greater chance that something unusual could happen if one party is more successful in getting their voters to the polling stations.

It should be an interesting set of results. The wider implications will be difficult to gauge. If incumbents win in three of four, and a popular local councilor defeats the incumbent in the remaining contest, will that say very much?

Wednesday, June 25, 2014

Chow maintains lead in Toronto race ahead of Ford, Tory

While the next election to be held in the country will be New Brunswick's in September, the mayoral election in Toronto scheduled for the end of October will be garnering more attention - as well as ballots. And the latest poll suggests that Olivia Chow continues to hold a lead over her two main rivals.

In the latest poll by Forum Research for the Toronto Star, currently the only polling game in town for the Toronto mayoralty, Chow held the lead with 34% support, down four points from Forum's last poll of June 6. Rob Ford jumped seven points to 27%, putting him narrowly ahead of John Tory, who was down four points to 24%.

David Soknacki was up one point to 6%, while Karen Stintz was down one point to just 3% support. Another 6% were undecided. Oddly, Forum is not releasing their numbers with the undecideds removed.

The shifts for Chow and Tory just straddle the margin of error, though Ford's are up quite a bit. However, it could very well just be a rebound from an abnormal result in Forum's last poll. If we look at the last four surveys from Forum, Ford has registered 27%, 24%, 20%, and now 27% again. Did Ford really hit rock bottom earlier this month, and has made up some of those losses? Or, more likely, was that 20% on the low side and Ford's support merely reverted to the mean?

Ford's approval certainly remains at an all-time low, at just 32%. It has been between 28% and 32% in the last three surveys, after previously having been in the mid-40s. The latest scandal, apparently, was the straw that broke the camel's crack, er, back.

If we look at how the numbers have been moving since candidates could officially register, things have been trending downwards for Ford. He was competitive with Tory and then Chow early on, but has since been on a relatively steady decline, the latest uptick notwithstanding. It was the first poll that ranked Ford second for quite some time.

Tory and Chow have been relatively stable, though Chow's numbers have been somewhat improved of late. Tory was leading before Chow entered the race, and her entry hurt him the most (along with Stintz, who went from respectable to fringe candidate). But Tory has rebounded a little, and the race is still more or less on between the three main candidates.

If Stintz and/or Soknacki drop out due to their low levels of support, it will not help Ford whatsoever. Forum suggests that of the nine points on the table between the two, Tory would get five of them and Chow would get three, with Ford getting just a single point (otherwise known as a rounding error). It suggests that if the race thins out, things will only get worse for Ford, not better.

Things would get a lot better for Tory if Ford drops out, as he would then be able to move to within two points of Chow.

Forum found that over 90% of Torontonians know who Chow and Tory are, compared to just 80% for Stintz and 63% for Soknacki. If we look at the approval ratings of the five candidates among all Torontonians, and not just those who know the candidates, which is what Forum reported, we get an idea of their respective ceilings.

The highest ceilings belong to Tory and Chow, at 56% and 55% approval, respectively. Next on the list is not Ford, but rather Stintz, who has 33% approval among all residents of the city. Ford is at just 32%, barely above Soknacki's 30%. It makes it very difficult to imagine Ford being able to get more than 1/3rd of the vote even under the best of circumstances. At this stage, re-election is simply not in the cards.

Or is it? Ford does trail Chow and Tory on approval and is behind Chow on voting intentions, but he does better on certain issues. He topped the list on being the best to handle the budget, at 31% to 27% for Tory and 26% for Chow, and was nearly tied with Chow on transit (27% to 26%, with Tory at 23%). These are the main issues, with 31% saying transit is the most important issue to them and keeping taxes low being the top issue for another 20% of Torontonians. But on having a vision for the city and, perhaps most importantly, on being able to leave the city in better shape after his or her term is over, Chow was well ahead of both Ford and Tory.

It appears that the race is still very much Chow's to lose, and she stands a better chance of avoiding defeat if Ford stays in the running. But it will be interesting to see if the recent uptick in Ford's support is an anomaly or not. For most candidates, being out of the spotlight for almost two months would normally be a bad thing. For Ford, it is his return that is unlikely to do him many favours.

Monday, June 23, 2014

Léger puts Liberals and NDP in tie in Quebec

With tomorrow being St-Jean-Baptiste Day, it is only fitting to take a look at the latest poll coming out of Quebec. It's from Léger, reported by Le Devoir on Saturday. The survey shows the Liberals and New Democrats in a tie in the province, but this masks Thomas Mulcair's sparkling personal numbers.

Léger was last in the field at the federal level in October 2013, so we don't really have any trends to monitor.

But this poll showed the Liberals and New Democrats in a tie with 34% apiece. That is not a bad number for the Liberals, who have averaged 33% in polls conducted in 2014 in the province, but it is a great number for the NDP, which has averaged 29%. If we exclude the polls from Angus Reid, which have been bullish on the NDP in Quebec this year, this is the best the NDP has done in any poll in Quebec since an Ipsos Reid survey conducted a year ago.

The Bloc Québécois was third with 17%, along with last week's Forum poll the worst the party has done in quite some time. Perhaps not coincidentally, both this Léger and last week's Forum poll were conducted after Mario Beaulieu was named the new party leader. The Bloc has averaged 21% in polls done in 2014, so this is on the low side for them by any measure.

The Conservatives were at 12%, well within the norm of recent polls in the province.

Among francophones, the New Democrats were well ahead with 39%. That is their best showing since a CROP poll from February 2013, before Justin Trudeau took over the Liberals. His party was at 25%, on the lower end of surveys since he became leader. The Bloc was at 21%, while the Conservatives continued to struggle with 12% support.

Among non-francophones, the Liberals dominated with 65% - actually a bit higher than they have been in recent months. Trailing at a distance was the NDP at 17% and the Conservatives at 12%.

With these levels of support, and especially considering the linguistic breakdowns, the NDP would retain most of their seats and win about 51, with 23 going to the Liberals and four to the Conservatives. The Bloc Québécois would be shut out of the legislature. The Liberal and NDP numbers suggest a huge degree of inefficiency on the part of the Liberal vote, but that is primarily because the party would be racking up super-majorities on the island of Montreal, and losing a lot of closer races elsewhere in the province.

Nevertheless, I would consider this level of inefficiency to be on the edges of the model's margin of error, and that the Liberals would do somewhat better. In four ridings, for instance, the gap between the NDP and Liberals stood at less than five points, so there is some potential for the Liberals to win closer to 30 seats, and the NDP closer to 45, at these levels of support. But we shouldn't be surprised if the Liberals have difficulty pulling a lot of seats out of Quebec, even if they win the most votes in the province.

This could be a problem for the Liberals in terms of their national seat count. If they only win some 25 out of 78 seats in Quebec, that puts them at a big disadvantage when the Conservatives can win 50 out of 62 seats in Alberta, Saskatchewan, and Manitoba.

Quick note on the provincial scene, rather insignificant considering the next election is four years away and the Parti Québécois is leaderless: the Liberals still lead with 40%, followed by the CAQ at 27% and the rudderless PQ at just 20%. The PQ is even running in third among francophones (24%), behind the CAQ at 32% and the Liberals at 29%.

Léger's poll included their quarterly 'barometer' of personal approval ratings for various federal and provincial politicians in Quebec. Among the federal figures, Mulcair finished head and shoulders above the others with 57% of Quebecers holding a good opinion of him, while just 12% held a bad opinion of the NDP leader. Another 17% were unsure, and 14% said they did not know who he was.

Next on the list was Trudeau, with a 42% to 36% split, and just 6% not knowing who he was. These numbers, for both Trudeau and Mulcair, are generally unchanged since Léger's December 2013 survey.

Stephen Harper's numbers were very low, with just 20% holding a good opinion of him but 68% holding a bad opinion. Elizabeth May had positive numbers, with a 24% to 10% split, but 24% did not know how they felt about the Green Party leader and 42% did not know who she was.

The Bloc's new leader has very little profile, as 54% said they did not know who he was and 22% did not know whether they had a good or bad opinion of him. Of the remainder, only 8% said they had a good opinion of Beaulieu and 16% said they had a bad one. The man Beaulieu beat to become leader was André Bellavance, who had an even lower profile (59% did not know him, 26% did not know what opinion they had of him). But 9% said they had a good opinion of him, and 6% had a bad one.

If we remove the undecideds and don't-knows for the three main leaders, Mulcair has a glowing 83% approval rating, with Trudeau at 54% and Harper at a woeful 23%. For the time being, that is even lower than Beaulieu's 33%, though many people have yet to form an opinion of him (early indications are it won't be positive when they do).

Compared to the provincial figures, Quebecers have very little clue about their federal politicians. Of the 15 listed, a majority of Quebecers recognized only nine of them. Of the 48 provincial figures listed, a majority of Quebecers recognized 32 of them.

The top five provincial or federal figures, rated by approval rating, were François Legault (CAQ leader), Mulcair, Françoise David (QS co-spokesperson), Philippe Couillard (Liberal leader and premier), and Trudeau. The bottom five, rated by disapproval rating, were Harper, Maxime Bernier (Conservative cabinet minister), Gaétan Barrette (Liberal health minister), Amir Khadir (QS MNA), and Sam Hamad (Liberal cabinet minister).

Just finishing out of the top five approval ratings was Pierre-Karl Péladeau. Just finishing outside the highest disapproval ratings was Trudeau.

It is interesting to look at the aggregate ratings of the federal figures listed by Léger. The Liberals come out on top in terms of approval ratings and recognition, but this is due to the high profile nature of the three names listed by Léger: Trudeau, Marc Garneau, and Stéphane Dion (they average 36% approval, 29% disapproval, 35% don't-knows/undecideds). Their total net score, however, is only +20. This is compared to the total net score of +63 for the four New Democrats listed (Mulcair, Alexandre Boulerice, Françoise Boivin, and Ruth Ellen Brousseau). Together, they average 25% approval, 9% disapproval, but also 67% don't-knows/undecideds. This suggests that, apart from Mulcair, the NDP still has a lot of work to do to introduce Quebecers to their NDP MPs.

The four Conservatives (Harper, Bernier, Christian Paradis, and Steven Blaney) totaled a net -83 rating, averaging just 16% approval, 36% disapproval, and 49% don't-knows/undecideds. It should be worrying for the Tories that their highest profile Quebec MP (perhaps despite their best efforts), Bernier, has such high negative ratings. It would have been interesting if Léger had included Denis Lebel, the Conservatives' actual top Quebec MP, in their listings.

It will make no headlines that the Conservatives are doing poorly in Quebec, and will be lucky to hold on to the seats they have there in 2015. But it is interesting to see the Liberals and New Democrats running such a close race (no turnout advantage to be had - the two parties were also tied among Quebecers 45 or older).

On the one hand, the NDP has an incumbency edge, but also a lack of profile for their incumbents. On the other hand, the Liberals have some more high profile Quebec MPs, but very few incumbents (and those they have were never going to be in danger of being toppled). Both Mulcair and Trudeau have good approval ratings, particularly Mulcair, but Trudeau has relatively high negatives as well. There is good scope for both significant Liberal gains in the province and wide retention rates by the NDP. Quebec will definitely be a battleground in 2015, but the importance of the seats won in Quebec will be determined in large part by how the election goes in the rest of the country.

Thursday, June 19, 2014

May 2014 federal polling averages

With the Ontario election campaign having kept us busy, it is time to finally take a look at the federal polling averages from May. There were only three polls conducted during the month, surveying just under 7,000 people. The numbers showed stability, the kind of stability that has been a feature of federal polling for much of the last 14 months.

The Liberals led for the 14th consecutive month in May with 33.9% support, a drop of 0.8 points since April. The Conservatives were down 0.5 points to 29.2%, while the New Democrats were up 0.2 points to 22.9%.

The Greens were up 1.1 points to 6.9% and the Bloc Québécois was down to 5.4%. Support for other parties stood at 1.7%.

How stable have things been? The Liberals have averaged between 34% and 36% support for the last five months. The Conservatives have averaged between 28% and 30% in 12 of the last 14 months, while the NDP has been pegged at between 23% and 25% for each of the last 14. That is a pretty settled electorate.

The Liberals narrowly led in British Columbia with 30.4%, a drop of 2.7 points. The Conservatives had 29.5% (+0.8) and the New Democrats 24.1% (-1.6), while the Greens picked up 3.2 points to hit 14.1% support. Apart from an anomalous December 2013, when only one poll was in the field, that 14% is the best the Greens have managed in B.C. since November 2010.

The Conservatives dropped in Alberta to 48%, a decrease of 8.7 points. The Liberals were up 1.7 points to 23%, while the NDP put up their best numbers since November 2013 with a gain of 2.6 points to 17.1%. The Greens were up 3.6 points to 8.5%.

Alberta has been, as usual, the least competitive region of the country in recent years. But the trend line for the Conservatives is certainly not a positive one. The Liberals are generally polling better than they did before the 2011 collapse, with support in the mid-to-low 20s. The same can be said for the New Democrats. Through they aren't at the 20% they routinely put up before and after Thomas Mulcair's leadership victory, they are usually registering in the mid-to-high teens, better than the 10% to 12% they could usually bank upon in 2009-2010.

Note that in the tracking charts, I have removed the markers denoting when interim leaders took over. I did this mostly to avoid cluttering up the chart, particularly in Quebec where the Bloc has had a musical chairs of leaders recently: Daniel Paillé in December 2012, André Bellavance in December 2013, Jean-François Fortin this past February, and now Mario Beaulieu.

In the Prairies, the Conservatives decreased 1.5 points to 41.2%, followed by the Liberals at 28.7% (-2.1). That is the lowest Liberal score in the Prairies since September 2013. The NDP was up 4.6 points to 22.8%, while the Greens were down 0.4 points to 6%.

The Liberals continued to lead in Ontario, up 1.3 points to 37.8%. They have been between 36% and 38% in the province since October 2013. The Conservatives were up two points to 34.8%, while the NDP was down three points to 19.8%. That is their lowest result since May 2013. The Greens were down 0.2 points to 6.1%.

For the first time since Justin Trudeau became Liberal leader, the Liberals did not hold an outright lead in Quebec last month. The party was down 2.8 points to 30.2%, their lowest since March 2013, while the New Democrats were up 1.3 points to 29.9%, their best since that date. The Bloc Québécois was up 1.2 points to 21.4%, and the Conservatives were up 0.2 points to 13.6%. Are the Tories picking up support in Quebec? They were at 12% in November-December, 13% in January-February, and have been around 14% since March. The Greens were down 0.2 points to 3.3% in the province.

The Liberals continued to dominate in Atlantic Canada, down 0.4 points to 53.7% (the Liberals have averaged 54% in each of the last three months). The Conservatives were down 4.6 points to 18.9%, their lowest since June 2013, while the NDP was up 0.5 points, also to 18.9%. The Greens were up 4.1 points to 7.2%.

With these levels of support, the Liberals and Conservatives would likely tie with about 128 seats apiece. That represents no change for the Tories compared to April, and a gain of one seat for the Liberals. The NDP was down five seats to 72, while the Bloc was up four seats to eight and the Greens were unchanged at two.

The Liberals made seat gains in Atlantic Canada (+2), Ontario (+1), and Alberta (+1), but dropped in Quebec (-3). The Conservatives were up in British Columbia (+3) and Ontario (+1), but were down in Atlantic Canada (-2), the Prairies (-1), and Alberta (-1). The NDP was up in the Prairies (+1), and down in Quebec (-1), Ontario (-2), and British Columbia (-3).

Stability is the word as we start ticking down on the 2015 election, now just 16 months away (or less than 12, if you believe that the Conservatives will pull the plug after a spring budget). One thing to keep an eye on in the next few months is the support of the Bloc. They are already quite low, but Beaulieu's start has been just short of a disaster. If that translates into a few extra points being made available to the other parties, that could be important in the multitude of very close ridings in the province.

Tuesday, June 17, 2014

Guest post: Frank Graves of EKOS Research discusses his likely voter model

The final poll of EKOS Research, whether it be the tally of all eligible voters or those of likely voters, told much of the story: the Ontario Liberals were on the cusp of a majority government. But the poll suggested the NDP would take a significant step backwards from its performance in the 2011 election, which turned out not to be the case. Why?

In this guest post, Frank Graves of EKOS Research investigates.


Frank Graves

Looking back at the Ontario election, we were very pleased to have predicted not only the victor, but also the majority. Only one other firm made this correct forecast. While we are very happy with the nearness of our final poll to the actual vote tallies for the two frontrunners, we were considerably less pleased with the performance of our Likely Voter (LV) model and our underestimate of NDP support. We were a bit over four points low on the NDP, but in a majority government this underestimate is of little practical consequence. The far more important challenges were identifying the winner, the leader of the opposition, and whether it was a majority or not; all of which we accomplished clearly and accurately. Yet in a closer race, the underestimate of NDP support could have been far more consequential and the challenge of forecasting who will vote and who won’t remains an important and unsolved puzzle.

As it turns out, the failure of the LV (LV) model, and the underestimate of the NDP are closely linked. The LV model significantly lowered NDP vote from a modest to large error and it moved the Liberal and Conservative votes from spot on to too high. Clearly, this wasn’t the intention and this error is somewhat baffling as the ingredients of the model are based on established historical patterns which show significant connections between these terms and likelihood of voting. To recap, we know that generally speaking (ceteris paribus), individuals who don’t know where there polling station is, who are younger and display lower socioeconomic status, who haven’t voted in recent elections and who don’t tell us that they are absolutely certain to vote, are less likely to vote. It is therefore extremely puzzling to see that applying these factors to estimate who will show up to vote not only wasn’t helpful, but magnified the error.

Further analysis of our final polls shows why this failure occurred and it is directly linked to the underestimate of NDP support. These could be quite separate issues, but an analysis shows they are interdependent. First, let’s look at the connection between our LV model and the final result.

The basic premise of our LV model involves assigning each respondent a “likely voter” score (maximum of eight points), based on a number of factors, such as age, past vote behaviour, knowledge of voting station, etc. Since we know that voter turnout is likely to be in the range of 50%, we isolate the (roughly) 50% of respondents who received the highest scores (giving us a “cut-off” of six points). Whether we apply the model weakly or strongly makes little difference. We find that different cut-off points make some impact but basically no model decreases error; they all magnify it. This is perplexing as we and others have shown clear linkages to these terms and voter turnout.

We then looked at the individual impacts of each of the terms of the model and none of the individual terms improved things. In part, the poor performance of the LV model is rooted in the paradoxical reverse linkages between turnout and the NDP vote. NDP vote appeared to be modestly rising in the later stages of the campaign, but we didn't make much of it because it was focused in parts of the voter spectrum that are not linked to high turnout: lower education, younger ages. 

Worse, the NDP vote was much more concentrated in non-voters from 2011, they were less likely to know where their polling station was located, they were considerably less certain to vote, and they were less likely to have voted in the advance polls. A few other notable findings: the NDP were performing more strongly with households with children at home and they were more likely to be union members. The link to the deep cuts to public sector workers may have caught the attention of these groups as the campaign matured.

One other interesting feature was evident in the data: NDP supporters were much more likely to only use a cellphone. We do sample cellphones, but we were under the population values and didn't weight this group up. This was because we were burned in the 2011 Federal Election for capturing cell only respondents who were much less likely to vote and less supportive of the Conservative government. In this instance, cellphone-only households were less likely to be Liberal or Progressive Conservative supporters, but they were not less likely to vote. If we had sampled more cellphone-only households, we would have probably been closer to the actual NDP vote.

In the end, we're left conclusion that even the most cautious and empirically informed attempts at creating LV models can be dangerous because things change; sometimes quite significantly. This harkens back to the classic problem of induction and the fact that the future will often not resemble the past. We are quite certain that presidential hopeful Mitt Romney was very surprised that he didn't win, as was Frank Newport of Gallup for precisely this reason. We therefore propose a more modest and less influential LV strategy that avoids the more ambitious approach we tried here. We also think that greater effort needs to be put into ensuring full coverage of all groups (e.g., cellphone-only households and younger voters), even if these groups hadn't been critical in past exercises.

We now have a reasonable handle on the reasons behind the underrepresentation of NDP supporters and the failure of the LV model. Our solution is a lighter less ambitious LV adjustment focusing on likelihood of voting. We also think that greater efforts to represent younger voters and cellphone-only households would be a prudent strategy.

We are, however, left with one big unanswered question. What is it in the latter stages of the campaign that seemed to engage what should have been a relatively disengaged NDP base to actually show up? That question is a very interesting one to which we have no real answer at this stage. A post-election survey would help to clarify this critical question.

What we do know is that this improbable appearance on Election Day was in defiance of historical patterns of likely voting and reinforces our conclusion to put more effort into creating the best random samples and less into seeking an elusive, unified field theory of voter turnout which we increasingly believe is a chimera.

Friday, June 13, 2014

Ontario election post-mortem: likely voter models fall flat, eligible tallies good

In the end, the race was not nearly as close as expected, with Kathleen Wynne's Liberals winning a majority government of 58 seats, with the Tories taking 28 and the New Democrats winning 21. The results have cost Tim Hudak his job, as his party took just 31.2% of the vote, with the Liberals improving on 2011 with 38.7% support and the NDP marginally upping their tally to 23.7%. The Greens had a good night, relatively speaking, with 4.8% of the vote.

Note: This post has been updated to reflect the reversed result in Thornhill, originally awarded to the Liberals but reverted to the PCs by Elections Ontario. The projection originally gave Thornhill to the PCs.

How about the polls and the projection model? The polls were not the miss that some commentators have suggested. As I write in The Globe and Mail today, the traditional numbers reported by pollsters actually did quite well, under-estimating the Liberals enough to put a majority in doubt but generally tracking the race accurately. However, the likely voter models employed during the campaign, and favoured by the projection model here, did not do the job at all. Every pollster that used a likely voter model did worse with it than they did with their estimates of eligible voter support.

As a result of this miss by the likely voter models, the projection model was off a fair bit. The projected support levels for the Liberals, New Democrats, and Greens fell within the likely ranges, which is a success. The PC number, however, was below even the 95% confidence interval.

In terms of seats, the Liberals ended up between the high and maximum expected ranges and the PCs between the minimum and low ranges. That is no coincidence, as the vast majority of the projection model's misses were seats projected to go PC, but that actually went Liberal. The NDP result was only one seat off the projection. So, a more mixed record there.

But the riding model itself did extraordinarily well. It only missed the call in 10 ridings, for an accuracy rating of 91%. That is the best performance of the model since the 2011 provincial election in Manitoba, when 56 of 57 ridings were accurately called. Taking into account the likely ranges (which should be the focus of anyone looking at these riding estimates), in only six ridings was the potential winner not identified. That ups the accuracy rating to 94%, or 101 out of 107.

Of those 10 misses, seven of them were seats projected to go PC, but instead went Liberal. One of them was expected to go PC, but instead went NDP, while two of them were expected to go NDP, but went Liberal. Half of the misses were in the 905 area code, where the PCs did unexpectedly poorly.

Nine of the 10 misses were called with 67% confidence or less, with five of them being called with less than 60% confidence. Only Durham, erroneously called for the Tories at 80% confidence, was a serious outlier.

The model would have done slightly better had I ignored the likely voter models. The vote projection would have been 37% for the Liberals, 33% for the PCs, and 24% for the NDP, with 50 seats going to the Liberals (or between 46 and 56 at the most likely range), 35 to the Tories (31-39) and 22 to the NDP (19-24). So, the Liberals and PCs would have still fallen outside the likely range, though less dramatically.

The model would have called 98 of 107 ridings correctly, for an accuracy of 92%, while the accuracy rating when incorporating the likely ranges would have been 102 out of 107, or 95%.

But what about if the polls had been dead-on? The seat projection model needs to be able to turn actual popular vote results into accurate seat projections, otherwise it would be hopeless in turning poll numbers into seats. On this score, the model did quite well.

The actual results for all three parties would have fallen within the likely ranges, with the Liberals on the higher end and the PCs on the lower end. This is one indication of how the Tories really had a poor night.

The overall call, instead of 'likely Liberal, possible PC victory', would have been 'Liberal victory, possible majority'.

The projection model would have called 98 of 107 ridings correctly with actual results, but when including the likely ranges at the riding level the model would have identified the potential winner accurately in 104 of 107 ridings, for an accuracy rating of 97%. The three ridings that bucked the trends? Cambridge, Durham, and Sudbury. The misses would have been called with an average confidence of only 55.4%.

All in all, the seat projection model performed as it should. The vote projection model did more poorly, but it can only be so good as the polls put into it. A look at the regional projections, though, gives us a look at where the polls missed the call - or, perhaps, where the parties over- and under-achieved.

The Liberals out-performed the polls across the board, but nowhere did they do so dramatically. Their actual result fell within the projected likely ranges in every region except eastern Ontario, where their 38.9% result was slightly up on the expected high of 37.2%. Nevertheless, they were up one to three points in each region.

The Tories under-performed in every region of the province. They were below expectations by three points in the north/central and eastern regions and four to five points in Toronto, the 905, and the southwest. Their result fell below the expected ranges in every region. They had a horrible night where turnout is concerned.

The New Democrats aligned quite closely with the aggregate projection, with differences of no more than 0.5 points in every region except the southwest, where they took slightly more of the vote than the higher likely range expected.

The Greens, shockingly, outperformed the polls in every region. That is a very rare occurrence.

But what if the projection had ignored those likely voter models? In most cases, the projection would have been closer. But still, the Liberals would have outperformed the polls in every region and the PCs would have under-achieved across the board. The NDP would have more traditionally under-achieved in most regions as well, as would have the Greens.

So, let's get to grading the pollsters. This should be done with a great deal of caution. Recall that most of these polls have a margin of error (theoretical or otherwise) of about two to four points for each party. And these sweepstakes, decided by a few decimal points, or not nearly as meaningful as many make them out to be. But let's put the results on the record.

In the chart below, I've ranked the pollsters by cumulative error for the four main parties, including both likely and eligible voter tallies. I've also highlighted in yellow every estimate that was within two percentage points of the result. In this regard, kudos needs to go to EKOS Research, the only firm to call two of the three main parties within two percentage points. In terms of the eventual outcome, their poll was probably the most informative, though they had the NDP too low.

Angus Reid Global's poll of all eligible voters of June 8-10 turned out to have the least total error, at just six points. Their numbers suggested a Liberal minority, however, as did Abacus's poll of eligible voters of June 9-11. would have ranked third among pollsters (or fourth among number sets), though if the likely voter models had been ignored the total error would have been just 4.4 points, putting the projection at the top of the list.

Oracle placed narrowly ahead of EKOS's eligible voter numbers, but the portrait of the race they painted (PC lead of one point) was not reflective of the outcome. After that, the errors become more serious, though both Forum Research and Ipsos Reid (eligible only) did have the Liberals in front.

You might be wondering why Nanos Research's poll is not included in the list. With a total error of just 1.5 points, the poll would have been - by far - the most accurate (Nanos had it as 37.7% OLP, 31.2% PC, 23.7% NDP, 5.3% GPO). But the Nanos poll was out of the field on May 26, 17 days before the election. While it is possible that voting intentions remained static during those 17 days, that is not something we can assume. And if we're allowing a 17-day-old poll to be used as a measuring stick, then the Abacus Data poll of May 28-31 (37% OLP, 30% PC, 24% NDP) was almost as good.

So where do we go from here? Clearly, the likely voter models are still in an experimental phase. When employed in Nova Scotia and Quebec, the first time we have seen them used in recent provincial elections, they only marginally improved the estimations, if they did not worsen them. We may come to the conclusion, then, that for the time being Canadian polling is not yet capable of estimating likely turnout with more consistent accuracy than their estimates of support among the entire population.

This is counter-intuitive, however. Likely voter models should improve things, particularly when turnout was only slightly above 50% yesterday. Going forward, should perhaps rely solely on those eligible numbers, until the likely voter models consistently prove their worth, and run a lesser, simultaneous model that takes into account likely voter estimates. This may provide the best of both worlds and give readers food for thought and all the information available, though it will not clarify things more fully. The challenges of polling elections in the modern age continues.

Thursday, June 12, 2014

Final Ontario projection: Likely Liberal, possible PC victory

Kathleen Wynne's Liberals stand the best chance of forming government in today's Ontario provincial election, with a likely return to a minority legislature. Tim Hudak's Progressive Conservatives will likely form the Official Opposition again, though they stand an outside shot of taking power themselves. Andrea Horwath's New Democrats should retain their role as third party in Queen's Park.

The likely outcome

The Liberals are projected to win between 42 and 55 seats, putting them just in range of a majority government (54 are needed), with between 35.4% and 40.6% of the popular vote. While this would be their fourth consecutive election victory, it could also be the least decisive of the four. The precise projection gives the Liberals 49 seats and 36.9% of the vote. That is little different from the 48 seats the party had at dissolution and the 37.7% the party took in 2011.

Final projection
The Progressive Conservatives are projected win between 33 and 44 seats with between 34.3% and 39% of the popular vote. This does make a minority victory by the PCs possible. The precise projection gives the party 36 seats and 35.8% of the popular, again little different from dissolution (37 seats) and their electoral performance in 2011.

The New Democrats are projected to win between 18 and 22 seats with between 20.4% and 23.8% of the popular vote. There is a chance, then, that the NDP could put up their best numbers since 1990. The precise projection gives them one more seat than they had at dissolution with 22, and slightly less of the vote than they earned in 2011 at 22.2%.

The Greens flirted with a seat for a period in the campaign, but in the end are projected to remain shutout of the legislature with between 2.9% and 5.1% of the vote. The precise projection gives them 4% support, which would represent their second best performance.

The range of plausible outcomes

With the polls at the end of the campaign - and this is something quite rare - showing greatly divergent results, it is especially worthwhile to look at the maximum and minimum projected ranges, encompassing 95% of likely outcomes. 

Projection ranges
When we stretch the projection to that extreme, we still get a likely Liberal, possible PC victory. The Liberals should be able to win between 29 and 68 seats, making a majority government or a return to Official Opposition status after more than a decade more than possible. Their vote range stretches from 34.3% to 47.2%.

The Progressive Conservatives could plausibly win between 25 and 60 seats, making a slim majority government possible for them. They should take between 32.2% and 45.4% of the vote.

The New Democrats do not seem to be in a position to move out of third place, with a maximum range of between nine and 25 seats, with between 18.4% and 25.8% of the vote. But some caution here: that maximum and minimum range does not even encompass all of the final polls, some of which had the NDP lower than 18.4% and higher than 25.8%. They are the wildcard in this election.

The Greens could take as little as 1.8% and as much as 6.3% of the vote,  but that does not put them in range of their first seat.

Regional breakdown

The Liberals are expected to finish first in the two most highly urbanized regions of the province, with the Progressive Conservatives placing first in the rest of Ontario.

Regional projections
In Toronto, the Liberals are projected to have between 44.1% and 50.5% support, enough to give them between 15 and 20 of the region's 22 seats. The PCs should finish second in the vote count with between 26.4% and 29.9%, but third in the seat count with between zero and three seats. The NDP, which has rallied in Toronto in the final stages of the campaign, should take between 20.6% and 24% of the vote and between two and four seats. The Greens are projected to take between 2.1% and 3.7% of the vote, their weakest region.

Regional projection tracker
In the 905 area code, which stretches from the Toronto-area suburbs through to Hamilton and Niagara Falls, the Liberals are projected to narrowly edge out the PCs with between 35.9% and 41.1% support. Their seat haul is projected to be between 16 and 18 seats. The PCs should take between 35.2% and 40% of the vote and win between eight and 11 seats. The NDP is projected to win five or six seats with between 19.5% and 22.7% support. The Greens should take between 2.6% and 4.6% of the vote. Support levels have been the most static in the 905 throughout the campaign.

The Progressive Conservatives have the most support in Eastern Ontario with between 40.1% and 45.5% of the vote, which should net them between seven and nine seats. The Liberals closed the gap in the last week of the campaign, and are projected to take between 35.7% and 40.9% of the vote, getting them between five and seven seats. The NDP is projected to remain seatless in the region with between 15.4% and 17.9% of the vote, while the Greens finish fourth with between 2.4% and 4.3% support.

Riding projections
In Southwestern Ontario, a relatively close three-way race that the PCs have led for most of the campaign, the Tories are expected to win between 12 and 14 seats with between 37.2% and 42.3% of the vote. The Liberals should take between 26.7% and 30.6% of the vote and two to four of the seats, with the NDP winning six seats with between 25.1% and 29.2% support. The Greens should take between 3.5% and 6.1% of the vote.

The closest three-way race of the entire campaign has been in Northern and Central Ontario. The PCs are projected to come out narrowly on top with between 31.8% and 36.1% of the vote and with six or seven seats, while the Liberals take between 31.3% and 35.9% of the vote and between four and six seats. The NDP is projected to win five or six seats with between 26.1% and 30.3% of the vote, while the Greens could put up their best numbers here with between 3.7% and 6.5% support.

The polls

Unlike the last two provincial campaigns that have taken place, in Quebec and Nova Scotia, the polls have been anything but consistent in Ontario. The last set of polls, at least among all eligible voters, were notably more consistent than earlier in the campaign, but by any other standards they were nevertheless quite divergent. It means that this projection is one of the least confident ones I have had to make, and there is a great deal of potential for a surprise tonight.

If we look at a mix of the final tallies that the pollsters will be judged against (their one-and-only numbers, or their estimations of likely turnout) we get a fair bit of agreement and disagreement. 

There is absolute consensus on the support the PCs will be able to capture: about 36%. All of the final polls put support for the party at between 35% and 37%. So on that score, at least, there is little mystery.

But that is not so for the Liberals and NDP. The governing party has been estimated to get as much as 42% of the vote or as little as 30%. That is the difference, of course, between majority government and Official Opposition.

Polling trends and intervals
For the New Democrats, the range is even wider: as high as 30% or as low as 17%. Here there is little consensus either, with polls landing everywhere between those highs and lows. And again, that makes a huge difference - perhaps not for the NDP, who would struggle to displace either the Liberals or PCs for second place even at 30% - but for the other parties. The PCs should hope for a high NDP number, as that will hurt the Liberals. A low NDP number, on the other hand, virtually assures a Liberal plurality, even if they find themselves short on the vote count.

The trends for the Liberals and PCs have been relatively clear and stable. The Liberals slowly inched up throughout the campaign, until taking a bit of a hit after the debate. The PCs dropped after initially leading, and up-ticked after the debate. The NDP, however, has been neither gaining nor dropping. The range of final outcomes, at least according to the polls, points to a disappointing campaign, a stellar breakthrough, and everything in between.

Why have the polls been so divergent? It has been a consistent trend in Ontario polling since Dalton McGuinty's resignation. Low engagement is undoubtedly a factor, as we all sit in fear of turnout dropping again from the record low of 48% in 2011. Part of it is, however, the increased variation added by the different turnout models. These are all designed differently and make different assumptions, whereas the polls of all eligible voters are based on the same assumptions and are done, for the most part, in similar ways. If we were only looking at eligible numbers in this campaign, we'd find the polls only slightly at odds.

How the leaders fared

Kathleen Wynne was the rookie leader in this campaign, as both Tim Hudak and Andrea Horwath took their second kicks at the can. Based on what the polls have showed, Hudak and Horwath had worse campaigns than Wynne did.

This is because Wynne's numbers have hardly budged since the pre-campaign period. Her approval and favourability ratings remained generally stable, and she continued to lead on who would make the best premier. Of 20 polls that asked this question throughout the campaign, Wynne led in 17 of them, or 85%. Of the 38 voting intentions surveys completed in this campaign, by contrast, the Liberals led in only 61% of them. She clearly has the leadership edge, averaging 31% on who would make the best premier against 25% for Hudak and 21% for Horwath.

It has not been a bad campaign for Hudak, however. The June 9 survey from Forum put his approval rating at 30%, the highest Forum has recorded it to be since October 2012. His approval among PC supporters has also improved, from under 60% before the campaign to 64-66% at the mid-point and finally 77% at the end of it. After having difficulty with their own leader, Tories seem to have come to terms with him.

But Ontarians are still not enamoured. The Angus Reid poll had 60% calling the PC leader 'unappealing', with just 21% saying he was 'appealing'. Those are rather bad numbers.

The NDP was banking on Horwath's personal popularity to carry them through this campaign, but the gambit may not have worked. Horwath's numbers have taken a hit as the campaign dragged on. Her approval rating had been well over 40% before the campaign began, but dropped to 34% in the last poll from Forum. Her disapproval rating is now nearing 50%, after being in the mid-30s before the campaign. A similar drop in favourability has also been recorded by Abacus, though she scored highest (33%) on being an appealing leader in Angus Reid's poll.

Wynne is given an 85% chance of winning her riding of Don Valley West, with an estimated support of between 52% and 60%, against 31% to 36% for the PC candidate.

Hudak's chances of re-election in Niagara West-Glanbrook stand at a comfortable 89% chance, with between 48% and 55% against 24% to 28% for the Liberal candidate.

Horwath is the most secure at a 98% chance of victory in Hamilton Centre, with between 58% and 67% support against 17% to 19% for the Liberal candidate.

Green Party leader Mike Schreiner is expected to come up short in Guelph, where the Liberals are favoured. He is projected to take between 9% and 16% of the vote in the riding, placing fourth. I would not be surprised, however, if he does better than that.

Missing the forest for the trees?

In past campaigns, it has been easier to gauge the state of the race as the polls were in general agreement. It was clear in Quebec that the Liberals moved ahead at the mid-point, and that the CAQ was making inroads in the final days. In Nova Scotia, it was obvious that the Liberals were going to win and that the NDP was in a fight for second place. Even in British Columbia, despite the overall miss, there was no mistaking the gains the B.C. Liberals were making.

Ontario has been different. The polls have actually been quite stable - but only from the perspective of each individual pollster. For the most part, the portrait of the race that they painted in the first week was the same as it was in the last week. But that portrait differed from pollster to pollster, with some consistently showing a Liberal advantage, others a PC edge, and yet others a tie. This has made it very difficult to tell the story of the campaign.

The aggregate has been clearer, with early PC strength and a generally stable Liberal lead afterwards that was reduced following the debate. If the election result mirrors the current projection, that will likely be the safest 'story' that can be told of what happened in the 2014 Ontario provincial campaign.

But because of the very different results from the polls, there has been a great focus on this cacophony, blaming the pollsters for being unable to consistently track the voting intentions of a disengaged and disinterested electorate. Because of the very different final tallies - caused by the drive to get it right after past misses - one pollster or multiple ones will be wrong. Some will be right, though. 

In all likelihood, the ones that were wrong will get the most attention, and Ontario will be chalked up as another miss. We can add the "Premier Hudak/Wynne" taunt to those of Premiers Adrian Dix and Danielle Smith. Never-mind that the polls accurately named Philippe Couillard, Stephen McNeil, Pauline Marois, Brad Wall, Kathy Dunderdale, Dalton McGuinty, Greg Selinger, Robert Ghiz, and David Alward as incoming premiers since 2010. Those two misses are all that count, apparently.

Some in the media will relish the opportunity to jump on whatever misses occur tonight, but pollsters have a role to play as well. Those that turn out to be right tonight must be humble, those that miss must not hide. Those who do well should not laugh at those who don't, those who don't should not call those who have a better night lucky. All of that infighting in the public sphere does nothing to help the image of the industry, and tarnishes both those who gloat and those who make excuses.

But the story of the 2014 Ontario provincial election should not be, of course, about the polls. It should be about the voters. If turnout is very low, or the number of those who decline or spoil their ballots very high, parties should ask themselves what they are doing wrong. It is not just the responsibility of organizations like Elections Ontario to get out the vote, but the role of parties to give voters an option they can be enthusiastic about. 

If the polls have said anything about this campaign, it is that it is likely to be a very close one and that every vote matters. The pollsters have tried to do their job to capture and report your opinions. Now it is up to you.

Final Ontario update

For those of you who stayed up late and those who are early risers, I have updated the site with the final projection for the June 12 provincial election in Ontario. You can get the full details of the final projection by clicking on the chart at the top of the page, or here.

A full, detailed analysis of the final projection will be posted on Thursday morning.

The polls added to the projection since the penultimate update are the following (from oldest to newest), with field dates and margin among eligible and likely voters:

Léger (June 8-9, PC/OLP tie)
Forum Research for the Toronto Star (June 9, OLP +7)
Angus Reid Global (June 8-10, OLP +4 eligible, PC +2 likely)
EKOS Research for iPolitics (June 9-10, OLP +6.4 eligible, OLP +7.9 likely)
Ipsos Reid for CTV/CP24 (June 6-11, OLP +3 eligible, PC +6 likely)
Abacus Data for the Sun News Network (June 9-11, OLP +3 eligible, PC/OLP tie likely)
EKOS Research for iPolitics (June 10-11, OLP +6 eligible, OLP +6.3 likely)
Forum Research (June 11, OLP +6)

Wednesday, June 11, 2014

2014 Ontario Election Projection

The following are's final forecasts for the provincial election in Ontario, held on June 12, 2014. These numbers were last updated on June 12, 2014, and reflect the best estimates as of June 11, 2014, the last day of polls included in the model. You can click on all of the charts below to magnify them.

Click here to read the detailed analyses from the main site concerning the Ontario election and new polls. 

The vote and seat projections in the central columns reflect the best estimates based on the available polling data. The low and high projections are based on the over-estimation or under-estimation of support the polls are likely to make, while the minimum and maximum projections are designed to include 95% of potential outcomes.

The chart below shows how each party is classified in the model for the determination of the high and low ranges, and the probability that the result will fall within any of the projection ranges.

Based on these probabilities, there is a 62% chance that the outcome of the election for the Liberals will fall between the best estimate (or average) projection and the high projection. There is an 71% chance that it will fall between the average and maximum projection, and so on. There is a 5% chance that the outcome will fall outside of the minimum and maximum ranges.

A detailed explanation of the vote and seat projection models and how the probabilities are calculated can be found here.

The projections are subject to the margin of error of the polls included in the model, as well as the inherent inability for the projection model to make perfect estimations of real-world dynamics. The projection ranges are a reflection of the degree of error polls have made in recent elections. The probabilities are based on how polls have differed from election results in the past.

The follow chart breaks the projection down by region. A description of what each region includes can be found at the bottom of this page.


The following chart lists the polls currently included in the projection model that make up at least 99% of the weighted average, as well as the weight each poll carries. It also lists the media outlet that either commissioned or first reported the poll.

By including polls in the projection, no representation as to the accuracy or equivalency of the methods used is implied, nor should inclusion be seen as an acceptance, endorsement, or legitimization of their results. However, the weighting scheme takes reliability partly into account.

The following is a list of the current projections for all 107 of Ontario's ridings. These are the best estimates of likely outcomes if an election were held on the last day of polling. The high and low results are  the estimates of likely floors and ceilings, based on the high and low vote projection ranges. The probabilities listed beside each riding is the likelihood that, if an election were held on the last day of polling, the winning party identified by the model would actually win. It does not assign any probability to a particular trailing party winning the riding - if a projection gives the leading party a 75% chance of winning, there is a 25% chance that any of the other parties could win (though, in practice, most ridings are only contests between two parties).

These riding projections are not polls and are not necessarily an accurate reflection of current voting intentions in each riding.


The chart below shows the evolution of the seat and vote projections and ranges since the start of the campaign.



Eastern Ontario - Carleton-Mississippi Mills, Glengarry-Prescott-Russell, Kingston and the Islands, Lanark-Frontenac-Lennox and Addington, Leeds-Grenville, Nepean-Carleton, Ottawa Centre, Ottawa-Orleans, Ottawa South, Ottawa-Vanier, Ottawa-West Nepean, Prince Edward-Hastings, Renfrew-Nipissing-Pembroke, Stormont-Dundas-South Glengarry

905 Area Code - Ajax-Pickering, Ancaster-Dundas-Flamborough-Westdale, Bramalea-Gore-Malton, Brampton-Springdale, Brampton West, Burlington, Dufferin-Caledon, Durham, Halton, Hamilton Centre, Hamilton East-Stoney Creek, Hamilton Mountain, Markham-Unionville, Mississauga-Brampton South, Mississauga East-Cooksville, Mississauga-Erindale, Mississauga South, Mississauga-Streetsville, Newmarket-Aurora, Niagara Falls, Niagara West-Glanbrook, Northumberland-Quinte West, Oak Ridges-Markham, Oakville, Oshawa, Pickering-Scarborough East, Richmond Hill, St. Catharines, Thornhill, Vaughan, Welland, Whitby-Oshawa

Northern Ontario - Algoma-Manitoulin, Barrie, Haliburton-Kawartha Lakes-Brock, Kenora-Rainy River, Nickel Belt, Nipissing, Parry Sound-Muskoka, Peterborough,  Sault Ste. Marie, Simcoe-Grey, Simcoe-North, Sudbury, Thunder Bay-Atikokan, Thunder Bay-Superior North, Timiskaming-Cochrane, Timmins-James Bay, York-Simcoe

Southwestern Ontario - Brant, Bruce-Grey-Owen Sound, Cambridge, Chatham-Kent-Essex, Elgin-Middlesex-London, Essex, Guelph, Haldimand-Norfolk, Huron-Bruce, Kitchener Centre, Kitchener-Conestoga, Kitchener-Waterloo, Lambton-Kent-Middlesex, London-Fanshawe, London North Centre, London West, Oxford, Perth-Wellington, Sarnia-Lambton, Wellington-Halton Hills, Windsor-Tecumseh, Windsor West

Toronto - Beaches-East York, Davenport, Don Valley East, Don Valley West, Eglinton-Lawrence, Etobicoke Centre, Etobicoke-Lakeshore, Etobicoke North, Parkdale-High Park, St. Paul's, Scarborough-Agincourt, Scarborough Centre, Scarborough-Guildwood, Scarborough-Rouge River, Scarborough Southwest, Toronto Centre, Toronto-Danforth, Trinity-Spadina, Willowdale, York Centre, York South-West, York West