Tuesday, July 22, 2014

Summer hiatus

We're in the dog days of summer now and, as we all know, polls are for dogs. So ThreeHundredEight.com will be taking a brief summer hiatus, returning after the civic holiday long weekend.

In the meantime, you might want to check out some of the recent articles posted to this site that you may have missed:

- A look at how the Ontario provincial election campaign would have been perceived differently if the likely voter models that performed so poorly had been ignored.
- A breakdown of federal party support by age and gender, and an investigation of how this support has shifted since the 2011 federal election.
- A history of the riding of Kingston and the Islands, from John A. Macdonald to Ted Hsu.
- All the latest polls from the upcoming Toronto mayoral race.

And if you really need a polling fix, time to finally check out Tapping into the Pulse, my ebook retrospective of 2013's year in politics and polling. What happened with the polls in British Columbia and the Brandon-Souris by-election? Why were the polls so much better in Nova Scotia? What was Justin Trudeau's first year as Liberal leader like, or the last year before the Quebec and Ontario provincial elections? You can find the links to order here.

You might want to explore the site a little bit as well. After six years and almost 1,500 posts, there is plenty to keep you occupied!

See you in August! Elections in New Brunswick and Toronto are just around the corner - it should be an interesting autumn.


Friday, July 18, 2014

The alternate and truer history of the Ontario provincial election

The conventional wisdom is that the polls did not do a very good job of predicting the result of the Ontario provincial election last month. But that is not entirely true. The likely voter models did a very poor job of estimating what the result would be. The polls of all eligible voters - the numbers that in the past have been the only ones reported - did a decent job.

But how did those likely voter polls influence the narrative of the campaign? What was their role in creating the surprise of election night?

To answer these questions, I re-ran my projection model for the entirety of the election without using likely voter numbers. At the time, I had favoured likely voter results when they were available. This time, I've used only the eligible voter numbers. The chart below shows how the two projections would have stacked up against one another. (Note that throughout this post, the dates refer to the final day of polling that would have been included in the model, rather than the date of publishing.)

 It makes for quite a different story. The difference would have been felt as early as May 9, when the first Ipsos Reid likely voter poll was published. Instead of the wide lead the Tories would have been given - one that they would hold until May 16, the PCs would have been given a narrow lead of just under four points. They would have lost that lead as soon as May 12, regaining it again on May 14, and then losing it - never to have it again - on May 15.

That is quite a different story than the one the likely voter polls told. Instead of relinquishing the lead for good on May 15, the Tories had been awarded the lead for the first three weeks of the campaign. They were given the lead again, though briefly, on May 21, May 27, May 29, and were tied with the Liberals on June 6. But according to the aggregate of the eligible voter polls, the PCs never held the lead in the last four weeks of the campaign.

It is interesting to note that the proposed plan to cut 100,000 jobs, which has been credited as the key moment that cost Tim Hudak the election, was announced on May 9, only a few days before the PCs irrevocably lost the lead.

So instead of a yo-yoing, too-close-to-call campaign, the Liberals would have been considered to be in a good position from mid-May straight through to election day. The Liberal lead would not have been large, but with the exception of the post-debate period, the PCs would never have been pegged to have more than 34.9% support after May 14, with the Liberals never below 35.3% after May 16.

The tightening that occurred after the debate (the news about the police interviewing Dalton McGuinty might have had more of an impact) would have still been recorded in the eligible voter polls, but instead of the close race persisting in the last days, the Liberals would have moved comfortably ahead again on June 9.

That the Liberals took 38.7% of the vote would have still been a bit of a surprise, as the aggregate would have never given the party more than 37.8% of the vote, except on May 20. But the PC fall to just 31.3% would not have been too much of a surprise. Over the last three days of polling, the PCs were steadily falling, from 35.4% on June 9 to 33.8% on June 10 and finally 33.1% on June 11. From there, the drop of 1.8 points is easily explained.

Instead, the PCs were pegged to be far higher. Dropping, yes, from 39.2% on June 8 to 36.9% on June 11, but that is a far cry from the 31.3% they achieved on election night. We were expecting a close race. If we had been ignoring the likely voter numbers, we would have expected a Liberal victory - the only question being whether it would have been a minority or a majority (the projection would have had a likely range of 46 to 56 seats for the OLP).

The New Democrats were, for the most part, under-scored throughout the campaign due to the likely voter models. The difference was negligible to the end of May, but afterwards the effect was more significant. The trend lines would have been mostly the same, with the New Democrats dropping steadily from May 31 to June 5, but a big jump would have been recorded from June 5 to June 7, instead of the slump that the likely voter models suggested.

From June 4 to June 9, the likely voter model aggregate pegged the NDP to be under 20% support, bottoming out at 17.6% on June 8. Without those models, the projection would have only once put the party under 20% (on June 5). For the vast majority of the campaign, the New Democrats would have been estimated to within a few points of their final result. The performance of the eligible voter polls, the aggregate at least, would have been right on the mark on election night: 23.9% instead of the actual 23.8%.

Going forward, I intend to focus primarily on the eligible voter numbers in order to try to tell the story of what the polls are saying is the current state of opinion, rather than what the pollsters are guessing will happen. That is what these likely voter models are in the end, a guess, as no one can estimate with certainty how humans will behave in the future. The huge variation in approaches that each pollster took was testament to that.

I will be recording the results of these models in future elections to provide a bit of a clue to readers as to how turnout might look, but will be keeping the focus on the main numbers reported by the polls until it is proven with consistency that these models will work in Canada. In the Ontario election, the likely voter models were more misleading than informative. This time, a look at the eligible voter polls, like the one above, seems to have told the story more truthfully.

Wednesday, July 16, 2014

Party support by age and gender

The polls currently suggest that the Liberals enjoy a small lead over the Conservatives nationwide, with the New Democrats running in third place. But when the Canadian population is broken down by age and gender, a different story can be told - including how the political mood has shifted since 2011.

During the Kickstarter campaign at the end of 2013, one of the backer rewards was the option to choose a topic of analysis I would look at here. The first request I received, from Keith Neuman of the Environics Institute, was to take a look at federal support broken down by demographics. This is that analysis.

Before getting to it, a few words about the project behind this analysis. Tapping into the Pulse: Political public opinion polling in Canada, 2013 was published as an ebook earlier this year, and is a retrospective look at politics and polling in 2013. The ebook is listed at $4.99, and can be downloaded directly from GumroadAmazon (for your Kindle) and Kobo (for your Kobo eReader).

To give this exercise some context, it would be useful to know how Canadians voted in 2011 by their demographic profile. There are ways to estimate this, such as exit polling or post-election polling, but it is impossible to know with certainty how exactly Canadian demographic groups voted in the last election.

I decided to look at the post-election polling conducted by Abacus Data. From May 2011, just after the election had been held, to August 2011, Abacus Data conducted three polls on federal voting intentions. The numbers matched up quite closely to the 2011 federal election result: an average of 39.6% for the Conservatives (exactly on the mark), 32.2% for the NDP (+1.6), 17.2% for the Liberals (-1.7), and 5.3% for the Greens (+1.4).

Using these three polls, and adjusting for the discrepancy between the election results and the polling numbers, we can get a good idea of how Canadians voted in the last election according to their age and gender.

Not all pollsters release full demographic information, and few use the same age brackets. So to compare voting intentions today to the last election, I've again focused on three polls from Abacus Data (in order to get a large enough sample with which to work). These polls were conducted between January and July 2014. That is a long period of time, but the numbers in Abacus's polling were consistent and, overall, there has been little movement in the polls in 2014.

The Liberals averaged 34% in those three polls, against 29% for the Conservatives, 23.3% for the NDP, and 6.3% for the Greens.

Now that we have these baselines, let's look at the demographic breakdowns.

Though the Conservatives won 40% of the vote overall, beating the NDP by nine points in 2011, the New Democrats might have taken more of the female vote than did the Tories, at 36% to 35%. The Liberals captured just 19%, with 4% of it going to the Greens.

The Liberals have made huge inroads among women, up 15 points. This came from both the NDP (12 points) and the Conservatives (7 points).

Interestingly, men and women's support is now rather uniform. The Liberals have 34% support among men, with 30% supporting the Conservatives and 23% the NDP. But in 2011, the Conservatives had an enormous advantage among male voters, with 45% support to just 26% for the NDP and 19% for the Liberals.

The Liberals have again picked up 15 points here, but most of that came from the Conservatives, who are down 15 points. The NDP has slipped three points.

Looking at the numbers this way gives us an indication of where support is coming from for the Liberals. They have improved their position equally among both men and women, but that female vote came from the NDP and the male vote came from the Conservatives. The effect has been to level the playing field, with each party garnering roughly half of their supporters from each gender. An interesting shift.

It should come as no surprise that the NDP won the youngest cohort of voters in 2011, with 38% support against 28% for the Conservatives and 20% for the Liberals. Today, however, the Liberals lead in this group with 34%, up 14 points. The NDP has dropped 11 points to 27% while the Conservatives have fallen seven points to 21%. The Greens are up four points to 10% among 18- to 29-year-old Canadians.

In 2011, 30- to 44-year-olds were somewhat representative of the population as a whole: 39% voted Conservative, 33% voted NDP, and 17% voted for the Liberals. They are still broadly representative, with 35% now supporting the Liberals, 27% the Conservatives, and 23% the NDP.

The Conservatives have suffered most among this group, dropping 12 points. The NDP has also fallen, by 10 points, while the Liberals have picked up 18 points - more than doubling their support among these voters.

This is a target group for all parties, considering the preponderance of family-friendly language used and family-focused policies proposed. It would appear that the Conservatives did a good job attracting these voters in 2011, but the Liberals are now managing it best.

The next group, who may be nearing retirement and have children heading to university, voted solidly Conservative in 2011: 41% to 30% for the NDP and 18% for the Liberals.

Now, this group is mostly split between the Liberals (33%) and the Conservatives (30%), with 22% supporting the NDP. The Liberals have picked up 15 points among these voters, with the Conservatives down 11 and the NDP down eight.

Finally, there is the oldest cohort of Canadians, an important demographic as they can be most counted upon to cast a ballot. They overwhelmingly supported the Conservatives in 2011, with 50% backing the party. The NDP took just 23% of the vote among this group, while the Liberals took 20%.

The NDP has held firm at 23%, but the Liberals have gained 15 points to move into a tie with the Conservatives, down 15 points, at 35%. If the Tories cannot win this voting block, they will not win the election.

When we break it down by age, we can see where the parties have suffered losses. The Conservatives have generally lost in uniform proportions, though perhaps took a bit more of a hit among older Canadians. The New Democrats have lost primarily younger voters, who have gone over to the Liberals, while older Canadians who backed the New Democrats in 2011 are mostly sticking with the party. The Liberals have generally gained in similar proportions across all demographic groups, but may have gotten their biggest uptick among Canadians who, coincidentally or not, are about the same age as Justin Trudeau.

All of this is relatively intuitive. The Conservatives are traditionally the party of older male voters, while the NDP would be expected to be the party of younger, primarily female Canadians. The Liberals enjoy uniform support levels among both genders and all age demographics, which seems to align with the broad appeal the party is seeking and is best placed to capture as the party in the centre. The Liberals remain in the best position one year before the next election. These numbers make it clear which demographic groups the NDP and Conservatives need to re-capture if they hope to win in 2015.

Monday, July 14, 2014

Riding History: Kingston and the Islands

During the Kickstarter campaign at the end of 2013, one of the backer rewards was the option to choose a riding the history of which I would profile, as I have previously done for Toronto-Danforth and Calgary Southwest. The first request I received, from Justin Irwin, was for the riding of Kingston and the Islands. This riding also holds a special place in my heart, as I lived there for four years when I studied at Queen's University.

Kingston is an old, charming city of about 160,000 people located where the Cataraqui River flows into the St. Lawrence and where the St. Lawrence flows out of Lake Ontario. Kingston was once the most important city in Upper Canada and was briefly the capital of the British colony. Some of the relics of Kingston's time in the sun still remain, like Fort Henry, built in the 1830s to guard the St. Lawrence and the entrance to the Rideau Canal from an American attack, and Kingston City Hall, built in the 1840s to house the legislature of a fledgling colony.

John A. Macdonald
Politically, Kingston has been the home riding of one premier, one Speaker of the House of Commons and a clutch of cabinet ministers. But it is perhaps most famous for being the riding of Canada's first prime minister, John A. Macdonald.

Before getting to the history of the riding, a few words about the project behind this profile. Tapping into the Pulse: Political public opinion polling in Canada, 2013 was published as an ebook earlier this year, and is a retrospective look at politics and polling for 2013. Now that we are in the second half of 2014, a history book on 2013 may be a little out of date. But there is much in the book that is timeless: the story of the provincial election campaigns in British Columbia and Nova Scotia, the federal Liberal leadership race and the first year of the Liberal Party under Justin Trudeau, and discussion of the polling debacle in B.C. and in a few federal and Ontario by-elections, as well as the polling success of the Nova Scotia campaign. A foreword by Paul Adams, journalism professor at Carleton University and former journalist with the CBC and The Globe and Mail, is worth the price of admission alone, as it goes over the evolving history of the relationship between pollsters and the media.

The ebook is now listed at $4.99, and can be downloaded directly from Gumroad, Amazon (for your Kindle) and Kobo (for your Kobo eReader).

The riding of Kingston and the Islands began its history in 1867 as the riding of Kingston, limited to the City of Kingston itself. Its first MP was the man who represented the riding in the colonial legislature: John A. Macdonald.

There is little reason for me to recount the storied career of Canada's first Prime Minister here. Macdonald served, in his first stint, from July 1867 to November 1873, when he also took on the role of the Minister of Justice and the Attorney General of Canada. Leader of the Liberal-Conservative Party from 1867 to 1891, in his first election in the new Dominion of Canada Macdonald took 83.8% of the vote, or 735 votes against 142 for a physician named John Stewart.

Macdonald was re-elected again in 1872, this time in a closer contest with just 54.9% of the vote against John Carruthers, who took the other 45.1%. By 1874, when Macdonald had lost power and was then sitting as the Leader of the Official Opposition, Macdonald won again against Carruthers by a whisker: 51.2% against 48.8% - a margin of 38 votes. The election was declared void and a second vote, held at the end of 1874, was won again by Macdonald, but this time by just 17 votes.

Alexander Gunn
Macdonald's luck ran out in 1878 - at least in Kingston. Though his party was returned to power, Macdonald lost Kingston to the Liberal Alexander Gunn, another Scot, by 53.7% to 45.9%. Perhaps sensing the danger, Macdonald also ran in the ridings of Marquette in Manitoba and Victoria in British Columbia, a rules quirk of the age. He was elected in both, but not in his home town.

In 1882, Gunn was re-elected against the new Conservative candidate, Michael Sullivan, by a reduced margin. Luckily for Sullivan, though, he was named to the Senate by Macdonald and would serve until 1912, when he was disqualified.

In 1887, Macdonald took another shot at Kingston, though still keeping his options open in the riding of Carleton. This time, Macdonald defeated Gunn - but again the margin was just 17 votes. Macdonald was also re-elected in Carleton, but he opted to represent Kingston again.

Macdonald and Gunn squared off for the last time in 1891, but this time Macdonald finally won a comfortable majority with 57.3%. It was his last election, and his best in Kingston since 1867. Macdonald passed away in June 1891 after suffering a stroke. He was replaced in the riding by James H. Metcalfe, who was acclaimed in 1892.

Kingston riding boundaries, 1895
In the post-Macdonald era, the riding of Kingston turned to the Liberals under Wilfrid Laurier, who would hold the riding for the next 15 years. Metcalfe opted not to run in 1896, and was replaced by Liberal Byron Moffatt Britton, who had briefly been the Mayor of Kingston from 1876 to 1877. Britton, whose father-in-law was Luther Hamilton Holton, a four-term Liberal MP for the riding of Chateauguay in Quebec and a member of the National Assembly at various times between the 1850s and 1870s, was re-elected in 1900 by a similarly narrow margin against the Conservatives.

Kingston, 1900
By 1902, Britton had been named a Judge of the Court of King's Bench for Ontario. Metcalfe made a failed attempt at a comeback, as William Harty, Commissioner of Public Works for Ontario from 1894 to 1899 and Liberal MPP for Kingston from 1895 to 1902, took a stab at replacing Britton. He was very successful, taking 60.8% of the vote, which still ranks as the highest share of the vote won by a Liberal in the riding. He would not win by the same margins in 1904 and 1908, but he was re-elected in both cases.

Harty decided not to run again in 1911, and William Folger Nickle, Conservative MPP for Kingston from 1908 to 1911, and later from 1922 to 1926 when he was named the Attorney General, ran as the Conservative candidate. He was elected, as Robert Borden's Conservatives were swept to power. In 1917, running under the banner of the Unionists at the height of the First World War, Nickle was re-elected with 77.7% of the vote.

Nickle stepped down in July 1919, and Henry Lumle Drayton was acclaimed in his place that October. Drayton had been named the Minister of Finance and Receiver General in August, and so needed a seat. He would hold these titles until December 1921.

Drayton did not run again in Kingston in the election held that year, moving over to York West instead where he would be elected three times before becoming the Chairman of the LCBO. Arthur Edward Ross, who first ran against Harty in 1908, took his place in the riding.

Ross had a distinguished military career, serving in the Boer War and being awarded the Croix de Guerre for his service in World War I, where he was mentioned in dispatches seven times. He ended the war as Director of Medical Services of the Canadian Expeditionary Force, retiring with the rank of Brigadier-General.

Arthur E. Ross
He had a distinguished political career as well. An alderman from 1904 to 1907 and mayor in 1908, Ross served as the Conservative MPP for Kingston from 1911 to 1921 and was the provincial Minister of Health from 1919 to 1921.

But he only won the 1921 election by a margin of 171 votes, taking it with 50.7%. He would subsequently be re-elected three more times, with far more comfortable margins.

He was defeated in 1935, however, in the election that saw R.B. Bennett's Conservatives drop from 48% to just 30% against Mackenzie King's Liberals. Norman McLeod Rogers defeated Ross by seven points, and would serve as Minister of Labour until September 1939, when he took over the Ministry of National Defence. Rogers was a military man, too, having served in the 6th Nova Scotia Mounted Rifles during the First World War. Notable in 1935 was the first appearance of the CCF in Kingston. The party took 2.8% of the vote.

Rogers was re-elected in 1940 with ease but died in office shortly thereafter at the age of 45. Angus Lewis MacDonald was acclaimed in his place that year. MacDonald, who was the Premier of Nova Scotia from 1933 to 1940, and again from 1945 to 1954, claimed the seat after being named Minister of National Defence for Naval Services in July 1940. He served in that capacity until the end of the Second World War.

MacDonald returned to Nova Scotia and the Conservatives took the seat back for the first time in 15 years. Thomas Ashmore Kidd won the riding with 50.9% of the vote in 1945. Another veteran of the First World War, Kidd was the MPP for Kingston from 1926 to 1940 and Speaker of the Ontario legislature from 1930 to 1934.

Kidd's tenure in office did not last long, however, as his vote share dropped to 42% in 1949 when William James Henderson was elected instead. A member of the Royal Canadian Signal Corps from 1942 to 1944, and then of the Legal Section in Italy in 1944, Henderson would be re-elected again in 1953 and 1957, that year fending off a comeback attempt by Kidd.

He was not able to fend off the Progressive Conservatives again in 1958, however, when Ben Allmark, an alderman from 1953 to 1958, was elected with 51.9% of the vote. But Allmark would not stay very long, being defeated by Edgar John Benson of the Liberals in 1962. This election was also the first one by the New Democrats, and the party took 4.6% of the vote.

Benson, a sergeant in the 1st Survey Regiment during World War II, was re-elected in 1963, 1965, and 1968 by comfortable margins. He served as the Minister of National Revenue from 1964 to 1968, President of the Treasury Board from 1966 to 1968, Minister of Finance from 1968 to 1972, and finally as Minister of National Defence in 1972.

But Benson resigned that year to become the President of the Canadian Transport Commission, and this was the opportunity for Flora MacDonald of the Progressive Conservatives, who won handily with 53.4% of the vote. She was the first woman to represent Kingston in the House of Commons.

MacDonald, the third with that name to represent the riding, would be re-elected four more times, serving until the 1988 election. MacDonald stood as a candidate in the Progressive Conservative leadership race in 1976, and was named Secretary of State for External Affairs, the first woman to get the post, from June 1979 to March 1980. She was named Minister of Employment and Immigration in Brian Mulroney's first cabinet, and then served as Minister of Communications from 1986 to 1988.

This period also featured a steady growth for the NDP, from just 10.9% in 1965 to between 18% and 20% in the 1979, 1980, and 1988 elections. But in 1988, MacDonald was defeated by the Liberal candidate Peter Milliken, who took 40.6% of the vote to 35.9% for MacDonald, the worst Conservative performance since Confederation at the time.

Milliken would go on to be the longest serving MP for Kingston and the Islands and its predecessors. His family had a political background, with his cousin, John Ross Matheson, being the Liberal MP for Leeds from 1961 to 1968 and Order of Canada recipient.

Milliken was re-elected six more times, taking a majority of ballots cast in 1993, 2000, and 2004. The Progressive Conservatives were reduced to around 20% support in the 1990s here, as the Reform and Canadian Alliance parties took between 12.5% and 15.4% of the vote between 1993 and 2000. With the merger that formed the current Conservative Party, that support increased to 23.1% in 2004, 26.1% in 2006, and 32.5% in 2008. That election was the closest for Milliken, as he took 39.2% of the vote - his lowest share.

But he had become a political fixture in Canada by this time. He was parliamentary secretary to the House Leader until 1996, and became Deputy Speaker from September 1997 to October 2000. In January 2001, he was named Speaker of the House of Commons, a role he would fill until June 2011. That made him the longest serving speaker in Canadian history.

Milliken retired in the run-up to the 2011 election, and Ted Hsu of the Liberals was elected with 39.3% of the vote. The Conservatives came close, with 34.9% of the vote (their best since 1988), while the New Democrats took 21.5%, their best ever. The Greens, after growing steadily from 1.7% in 1997 to 10.8% in 2008, dropped again to 4.2% of the vote. 

Kingston and the Islands, and its predecessors, has had a rather stable political history. Only the Liberals and the Conservatives have ever represented the riding, and their vote share has merely wobbled back and forth before the breakdown of the traditional party system in 1993. After Liberal dominance in the Milliken era, the riding seems to be returning to a closer Liberal-Conservative contest, as the NDP returns to the level of support the party was enjoying in the Ed Broadbent years.

Overall, Kingston and the Islands is a Liberal riding - the party has won it 24 times against 19 for the Conservatives, and has averaged 46.1% support since 1935, compared to 39.9% for the Conservatives and 11.5% for the CCF/NDP. It was closely contested in 2011, but that was a historically bad election for the Liberals. Will it return to being one of the safest ridings in the country for the party in 2015?

Thursday, July 10, 2014

June 2014 federal polling averages

The number of polls released each month drops dramatically during the summer, down to two or three polls per month in 2012 and 2013. June usually sees a rush of polling before the summer break, however. Last year, six polls were released in June. In 2012, that number was eight. In 2014, the total number of national polls released was just two. That makes the monthly averages for June a bit of an anomaly, as they are based on only two data points outside of Quebec (two Quebec-only polls were also released). So, we should exercise caution when looking at these results. You've been warned.

The total sample size for the month of June is relatively large, however, at just under 10,000 Canadians thanks to the mega-poll conducted by Angus Reid Global, which alone sampled over 6,000 people. But the only two pollsters in the field, the other being Forum Research, showed very different results, with Forum giving the Liberals an eight-point lead and Angus Reid judging the Conservatives to hold a one-point advantage.

This inconsistency has been, well, consistent. In May, Forum gave the Liberals a six-point lead when Angus Reid still gave it to the Tories by one point. In April, it was a nine-point edge for the Liberals from Forum and a two-point advantage for the Conservatives from Angus Reid. What this means, then, is that these two pollsters have been showing the same general things for the last few months, regardless of how that compares to the findings of other pollsters.

Let's get to June, then. In these two polls, weighted for sample size, the Liberals averaged 33%, a drop of 0.9 points from May. The Conservatives were up 1.8 points to 31%, their highest result since March 2013, before Justin Trudeau became Liberal leader. The NDP was up 1.4 points to 24.3%, while the Greens were down 1.2 points to 5.7%.

The Bloc Québécois was down 0.7 points to 4.7% and 1.3% of Canadians said they would vote for another party.

The Liberals have now led in the polls for the last 15 months, though the Conservatives do seem to be experiencing some wobbly improvement.

In British Columbia, the Conservatives reached their highest level of support since September 2013 with a 4.3-point gain to 33.8%. The NDP jumped 4.7 points to 28.8%, while the Liberals were down 3.6 points to 26.8%. That was their third consecutive month of decrease. The Greens were down 4.5 points to 9.6%.

The Conservatives led in Alberta with 55.4%, up 7.4 points since May, while the Liberals were down 0.3 points to 22.7%. The NDP hit a 4.2-point slide to 12.9%, while the Greens were down 1.8 points to 6.7%.

In Saskatchewan and Manitoba, the Conservatives hit their highest level of support in a year with a 2.3-point gain to 43.5%, while the Liberals were up 2.2 points to 30.9%. The NDP was down 1.8 points to 21%, and the Greens were down two points to 4%.

The Conservatives have experienced a bit of a rebound in the Prairies in the last few months. Support was oscillating back and forth before Trudeau came along, but after that point the Tories were on a general slide, cratering at the end of 2013 and actually briefly relinquishing the lead to the Liberals. Since then, though, the party has returned to the low-40s where it has stood for most of the post-election period.

Support for the Liberals has been steady in Ontario for nine months now, at between 37% and 38%. In June, the party averaged 37.7%, a drop of 0.1 point, while the Conservatives were up 0.5 points to 35.3%. The NDP was up 0.9 points to 20.7%, while the Greens dropped 0.4 points to 5.7%.

The New Democrats captured their first outright lead in Quebec in June, the first time they had done so since Trudeau took over the Liberals. The NDP was up 4.3 points to 34.3%, their best number since October 2012. The Liberals were up 1.3 points to 31.2%, but the last two months have been the worst for the party under Trudeau (though that still ranks above the performance of the Liberals from mid-2009 until the beginning of 2013).

The Bloc Québécois was down 3.2 points to 17.9%, their worst since August 2013. The result at that point, and their average in June, marks their lowest support levels since October 2011. Mario Beaulieu is experiencing the opposite of a honeymoon. The Conservatives were down 1.9 points to 12.3%, while the Greens were up 0.3 points to 3.3%.

Note, since both Léger and CROP reported in June, the average is more robust for Quebec than it is elsewhere. Also, I had missed recording a CROP poll from May, so the monthly average has been retroactively adjusted. It is reflected in the tracking chart above.

And in Atlantic Canada, the Liberals took a 3.7-point drop to 50%. The Conservatives were up 5.9 points to 24% and the NDP increased by 1.4 points to 20.3%, while the Greens were down 1.9 points to 5.3%.

Last month, a 4.7-point national lead for the Liberals translated into a tie in the seat count with the Conservatives. This month, the two-point lead translates into a 13-seat deficit.

The Conservatives would win around 129 seats at these levels of support, up one seat from the May projection. The Liberals fell 12 seats to 116, while the NDP was up 18 seats to 90. The Greens were unchanged at two, while the Bloc fell from eight seats to just one.

The Conservatives gained one seat apiece in British Columbia, Alberta, and the Prairies, and picked up two in Atlantic Canada. They dropped four in Quebec, however.

The Liberals were down one seat in Ontario and Alberta each, two in both Quebec and Atlantic Canada, and six in British Columbia.

The NDP was up 13 seats in Quebec, five in British Columbia, and one in Ontario, but was down one on the Prairies.

Overall, then, it would seem to be a rather poor month for the Liberals. But is that fair? A completely unweighted average of the Forum and Angus Reid polls from May and June would reveal a 1.5-point gain for the Liberals, rather than a loss. But the party has been losing steam in British Columbia for the last few months, and the NDP does seem to be making inroads in Quebec again (primarily at the expense of the Bloc Québécois). The Conservatives are slowly returning from the doldrums of the high-20s, but are still very far from a re-elected majority government. Will the summer see any of this change, or will the numbers solidify as people turn their attentions elsewhere?

Tuesday, July 8, 2014

Nanos poll puts Tory ahead

The polling from Forum Research has dominated the Toronto mayoralty race, with the last non-Forum poll (by Ipsos Reid) dating from November 2013. In those polls by Forum, Olivia Chow has consistently led by a comfortable margin, with John Tory mostly placing second since the latest Rob Ford fiasco. But a new poll by Nanos Research turns that on its head: Ford is still out of the running, but Tory is given the lead, with Chow running in second.

What's going on? It is difficult to say on first glance that the Nanos poll is an outlier, as we have only heard from Forum. That all the polls from Forum have been showing something similar does not, alone, discount Nanos's findings. We'd need to have multiple pollsters agreeing with Forum before we could say that Nanos is out of step. For all we know, Nanos could continue polling every month, just like Forum, and show the same consistency in its results. We need a third opinion before reaching any conclusions.

This Nanos poll was commissioned by the Ontario Convenience Stores Association, though published directly by Nanos. According to the Nanos report, "the vote and issue module was asked first in the survey followed by some proprietary questions related to convenience stores." This should ensure that the sample was unbiased by the sponsor.

The poll found that 39% of decided and leaning voters supported Tory, followed by Chow at 33% and Ford at just 22%. Karen Stintz had 4% support, with Sarah Thomson at 2% and David Soknacki at just 1%. About 11% of respondents were undecided and not leaning towards one candidate.

This is quite different from the portrait Forum has been painting. In recent polls, Chow has ranged between 35% and 40%, with Tory between 25% and 29%. The Ford numbers are a little lower than most of Forum's recent polls, but not out of the ordinary. Soknacki's 1% is also on the low side.

If we take into account the margin of error from Nanos's survey of decided and leaning voters (about +/- 4%), Tory could be as low as about 35% and Chow as high as 37%. That puts Chow in the ballpark of the Forum polls but Tory is still quite high. Forum was last in the field on July 2 and had Chow at 38% among decided voters, with Tory at 28%. Could there have been dramatic movement on July 3-5 or is there a methodological issue causing the difference?

We can probably assume that methodology is the main factor here. Forum conducts its polls via interactive voice response (robo-dialing) and over a very short time period (usually a few hours), which rules out the possibility of call-backs. This is when a pollster tries to get in touch with a randomly-dialed number when the respondent at first doesn't pick up. This lessens the error that could creep in from only sampling a subset of the population at home on a certain day of the week.

Nanos, on the other hand, conducted this poll with live-interviewers and dialed both cell phones and landlines (Forum does not make clear whether it does so) over a few days, which allows it to carry out call-backs (Nanos points out that it conducted a maximum of five call-backs). The differences in the sampling methods may thus be behind the variation between the polls.

It isn't the first case of divergent views from Nanos and Forum. In their gauging of the Ontario provincial scene, both firms could rarely come to any agreement. In January, Nanos gave the Liberals an eight-point edge over the Tories. In the same month, Forum gave the PCs the lead by three points. A disagreement on who led in Ontario occurred again in February, and again in April, as well as on numerous occasions in 2013. So to see Nanos and Forum at odds here is nothing new. Nanos did not poll late enough in the Ontario provincial campaign for us to measure its performance versus Forum, however.

What about in the 2010 mayoral election? Both Forum and Nanos were active, but well before Election Day. The final result of that contest was 47% for Ford, 36% for George Smitherman, and 12% for Joe Pantalone. Forum did a poll 11 days before the vote, and gave 44% to Ford, 38% to Smitherman, and 16% to Pantalone. Not a bad performance. Nanos, polling nine days before the vote, gave 44% to Ford, 41% for Smitherman, and 15% for Pantalone. A little worse, but we're talking margin of error differences.

(Note, there has been some chatter on Twitter about the reliability of this poll, since it was commissioned by a lobby group that has some past links to Tory. The people casting aspersions on the poll have every political incentive to do so, considering they are supporters of the Chow campaign. What they are implying is that the poll has been manipulated in Tory's favour to please the OCSA, but the logical extension of that smear is that Nanos Research manipulated its own poll. No one is saying that directly because they have no proof and to suggest such a thing would be libelous. It is also implausible. Nanos's success in business banks on being reliable and objective, and it is ridiculous to believe that the company would purposefully manipulate a poll and then release it publicly, data tables and all. The poll was not even published by the OCSA, which would raise some flags, but directly by Nanos, something that the polling firm would have absolutely no reason to do if the numbers were manipulated to please a client. It might be too much to expect from some political operatives, but they should exercise a little caution before they accuse individuals of being lying shills.)

The poll did some have other interesting tidbits. Like all its polling, Nanos asks respondents to list their first two choices. Combining them gives us an indication of each candidate's potential ceiling. Tory's is the highest (his approval rating has been the highest in Forum's polling as well) at 57%, with Chow not far behind at 46%. After that, no one has a real chance of pulling out a victory. Ford's ceiling tops out at just 27%, Stintz's at 17%, Soknacki's at 10%, and Thomson's at 4%.

Nanos also asked respondents to name their top issue - unprompted. This means survey takers had to come up with an issue themselves, rather than choose from a list. By far, transit was the top issue at 35%, followed by property taxes at 17%, jobs and the economy at 16%, and traffic at 14%. The embarrassment of Ford took only 4%, but that is not too surprising. Ford's antics might be a reason someone won't vote for him, but that doesn't change the election being about a more important issue that is relevant to a voter's daily life.

In terms of whether it is Tory or Chow who is leading, we're better off to wait and see what other polls are going to say about this race before determining who is off the mark (both could be, of course!). It will also be interesting to see if Nanos will continue polling, and if so, what its results will be. For now, we can just say that the race remains primarily between Tory and Chow.

Friday, July 4, 2014

Chow still leads, Ford drops to third

A new poll from Forum Research for the Toronto Star shows that Olivia Chow still holds the lead in the race for the Toronto mayoralty, and that the spike in support recorded in Forum's previous survey for Rob Ford seems not to have been the start of a new trend. He is now back in third behind John Tory.

Chow was up two points from Forum's last poll conducted on June 23, and now leads with 36% support among all respondents. Tory was up three points to 27%, while Ford dropped one point to 26%. David Soknacki, down two points to 4%, and Karen Stintz, unchanged at 3%, were well behind. Another 4% of respondents were undecided.

If we remove those undecideds, Chow improves to 38%, followed by Tory at 28% and Ford at 27%. 

None of these shifts in support appear to be statistically significant, and if we look at the chart above it is clear that the candidates are mostly just wobbling to and fro within the margin of error. Chow, for instance, has been between 35% and 40% since she officially launched her bid, and Tory has been between 25% and 30% since Ford's support cratered in April. 

Ford's numbers, just like the man, have been a bit more erratic, but are mostly hovering around the 25% mark.

So it would appear that Torontonians have mostly settled into their voting intentions, at least for the time being. That is good news for Olivia Chow, who holds a comfortable lead. But while it is possible to see a road to victory for Tory, it is not possible to see the same thing for Ford.

Ford's approval rating is down to 31% among all Torontonians, putting him just one point up on Soknacki (who is penalized mostly for not being well known). If Stintz and Soknacki drop out of the race, Ford's numbers only improve to 28% against Chow and Tory, still too low to be in a position to win. And when asked directly whether they will vote for Ford, 69% of Torontonians said no. It is not possible for Ford to win with the remaining 31% (some of whom were undecided, rather than willing to vote for Ford) unless Chow and Tory split the vote almost evenly, and Soknacki and Stintz up their support levels.

It is difficult to imagine how such a scenario could come about. If something like that happens in the last weeks of the campaign and the polls record it, we are unlikely to still see a large portion of the electorate voting for candidates who have no hope of beating Ford. Instead, either Chow or Tory would likely get a surge from being the consensus anti-Ford candidate.

Even Ford's 'rehab' won't save him. Asked again if Torontonians would vote for Ford if he stays clean and sober (which only 26% think he will over the next three months), still 67% will not vote for him. He has alienated two-thirds of the electorate, and that will sink him. Ford may have his fans, but his boosters do not love him as much as the rest of the city is sick of him.

But what about Tory? Ford is the biggest obstacle to him winning the job. When Ford is dropped from the race, Tory moves into a two-point lead over Chow (38% to 36% among all voters, 46% to 44% among decideds). He retains that two point edge if Soknacki and Stintz also drop out. This is the first time Tory has led Chow in these hypothetical scenarios since February.

Tory's approval ratings are also the highest at 63% among decided voters, or 60% among all voters (an increase of four points since June 23). Chow is not far behind at 56% approval among all voters, but it suggests that Tory has the higher ceiling of the two. If this race was not dominated by the Ford circus, Tory would have a very good shot at becoming Toronto's next mayor.

But Chow is still a force on her own, since the narrow lead Forum gives Tory in a non-Ford contest is well within the margin of error. Tory may have an ace up his sleeve, however, in that he is much more competitive among older voters. While the edge over all is worth nine points for Chow, that shrinks to one to four points among voters aged 35 to 54 and over 65, while Tory has a three-point edge among voters aged 55 to 64. The Ford factor is likely to encourage high turnout among younger voters, but if primarily older people vote the margin between Tory and Chow could shrink considerably.

That is as competitive as the race is likely to get. At best, the contest is two-headed with Chow holding a healthy advantage. Ford is merely taking some 25% of the vote off the table and may play the spoiler, but he no longer has the support to put him in a position where he can actually win. As long as they don't do something crazy like, say, smoke crack on video, slur racial, homophobic, and sexist comments while on a bender, or associate with known criminals, either Tory or Chow will win. How sad that pointing something like this out in the past would have been a joke.

Wednesday, July 2, 2014

Liberals big winner in by-elections

In the end, the by-elections on Monday went just about as expected, with the Liberals picking up Trinity-Spadina and making impressive gains in rural and northern Alberta. The turnout was low - in two cases setting new records - but the results were nevertheless in line with what the national and regional polls suggested should be the case.

This would seem to validate the results, despite the low turnout. It also seems to validate the national polls that have been giving the Liberals the lead in voting intentions. In every case since Justin Trudeau took over the party, the Liberals have made important gains in by-elections, just as the party has done in national polls. Yes, after 2011 the Liberals had nowhere to go but up, but in many cases the numbers the Liberals have been putting up in these ridings were similar to the vote shares the party managed in the days of Jean Chrétien. In other words, the Liberal surge is real and not just a rebound from historical lows.

Let's go through the by-election results, riding by riding. The following charts show the change in vote share and actual votes compared to the 2011 federal election.

Amazingly, the race in northern Alberta was the closest of the four.

The Conservative David Yurdiga prevailed in the end, with 47% support against 35% for Kyle Harrietha of the Liberals (the party's best performance since 1968). The New Democrats took 11% of the vote, while the Greens took 4%.

That was a big swing between the Tories and Liberals. The Conservatives dropped by 25 points while the Liberals picked up 24.9 points. Despite the record-setting low turnout, the Liberals increased their raw vote count by 1,301 ballots. The Conservatives dropped 16,043 votes and the New Democrats 2,604. The margin between the Conservatives and Liberals, standing at 18,798 votes in 2011, fell to just 1,454.

The changes were less significant in Macleod, where turnout was also quite low. John Barlow of the Conservatives took 69% of the vote against 17% for the Liberals' Dustin Fuller (the best performance for the party since 1957). The Greens placed third here with 6% support, and the NDP finished fourth with 4%.

The Conservatives dropped a more modest 8.7 points, but the Liberals picked up 13.3. Much of that came from the Conservatives, but also from the NDP, which was down 6.1 points.

In raw votes, the Conservatives fell from 40,007 to 12,394, a decrease of 27,613 votes. The Liberals, however, increased their total from 1,898 to 3,062, a gain of 1,164 votes. The Greens lost about half of their votes, and the NDP about 85% of theirs.

The Liberals did not suffer from the departure of Jim Karygiannis in Scarborough-Agincourt. Instead, their vote share under Arnold Chan increased by 13.9 points to 59%, while the Conservatives' Trevor Ellis dropped 4.9 points to 29%. The NDP fell even more dramatically, by 9.6 points to 9%. The Greens took 1% of the vote (and actually placed fifth, behind an independent candidate).

The Liberals could not increase their vote total here, however, dropping 5,669 ballots to 12,829. The Conservatives shed 7,586 votes, falling to 6,344, while the NDP dropped 5,532 votes to just 1,844.

The marquee match-up of the evening went very comfortably to the Liberals, as Adam Vaughan won 54% of the vote in Trinity-Spadina, an increase of 30.3 points over the party's 2011 performance and the best the party has done since the riding was created in 1988. Joe Cressy of the NDP took 34% of the vote, down 20.4 points from Olivia Chow's performance. The Conservatives fell 11 points to 6%, while the Greens were up slightly to 5%.

In terms of raw votes, the Liberals increased their total here by 3,271 to 18,547, while the NDP dropped 23,799 votes to 11,802 (less than what the Liberals took in 2011). The Conservatives lost 8,954 votes, falling to just 2,022.

Overall, it was a stellar performance for the Liberals across the board. They averaged 20.7% support across these four ridings in 2011, but on Monday night they averaged 41.3%.

The Conservatives placed second last night with an average of 37.7% support, after capturing an average share of 50.1% in 2011. The NDP was down from 24% to 14.6%, while the Greens held steady at 4%.

For the Liberals, this was their best performance since before 1988. The closest they have come since then was an average of 38.8% in 2000.

This was the Conservatives' worst performance since before 1988, if we combine the vote totals of the Progressive Conservatives and Canadian Alliance/Reform between 1988 and 2000. Their previous low in that span was 2004, when the Conservatives took an average of 41.2% support in these four ridings.

For the New Democrats, this was their worst performance in these four ridings since 2000, when they averaged 12.8% support.

Considering how last night's results stand against historical performances over the last 26 years, the numbers are remarkable. So perhaps they are a fluke?

I don't think that argument can be easily made. Before the by-elections, I posted what the high and low ranges were when applying this site's projection model to the ridings using the last 30 days of Alberta and Ontario polling. At the time, I hadn't included the latest Angus Reid survey. But the chart below includes that survey, and shows that the results fell well into line with what the province-wide polls suggested could be the case.

As you can see, the bulk of the results fell just outside or well within the expected ranges.

In Macleod, the Conservatives performed within expectations, while the Liberals over-achieved a fair bit and the NDP and Greens under-achieved.

In Fort McMurray-Athabasca, the Conservative result was just below where the polls suggested it could fall, and Harrietha's was just above the high-end. Perhaps in this we see the push the Liberals made in the final days and the multiple visits from Trudeau.

The Liberals just out-performed expectations in Trinity-Spadina, no doubt due to Vaughan's high-profile, while Cressy performed just about as expected. The Conservatives, not in the running, were well below expectations.

And in Scarborough-Agincourt, both the Liberals and Conservatives performed as expected, with the NDP and Greens falling just below their low range.

This seems to back-up the results, suggesting that they were not a fluke but generally representative of wider trends. It also suggests that the national polls are not off the mark, and that what they are recording is in the ballpark of how Canadians feel right now.

But other polls did not do so well. As usual, Forum conducted riding-level polls in each of these four ridings, despite the myriad of difficulties involved with doing so. One particular issue is the inability to geo-locate cellphones, leaving them entirely outside the sample. Here is how Forum's polls did:

Overall, they did rather poorly. Even accounting for the +/- 5% margin of error that Forum reported, in only two cases (the Liberals in Macleod and the NDP in Trinity-Spadina) were the two leading parties in each riding gauged accurately.

The winners were identified in Macleod, Trinity-Spadina, and Scarborough-Agincourt, but these were relatively easy to call. The Tories were underestimated by 15 points in Macleod, and the Greens over-estimated by 10 points. In Fort McMurray-Athabasca, the Liberals were wrongly identified as the winner, while the Tories were underestimated by 14 points.

In Trinity-Spadina, the Liberals were under-scored by nine points and the Conservative overestimated by five. In Scarborough-Agincourt, the Liberals were underestimated by 11 points and the Conservatives overestimated by eight.

Not a strong night for Forum. But there is some credit worth portioning out. Forum did record that the Liberals were doing well in Alberta, greatly above historical levels. And throughout the campaign Forum pegged Vaughan as a game-changing candidate. Its polls there made Trinity-Spadina a much easier call, even if they missed on some of the final numbers.

Polling by-elections is perilous, and Forum should probably stop doing it as it has not helped its reputation. But these by-election polls have been broadly informative, rather than precisely accurate. They told us much about Vaughan's influence in this round of by-elections, and the surprising Liberal gains to be made in Manitoba last fall. We just need to know how to read these by-election polls better, if they will continue to be published. Look at the broad strokes and the general narrative, forget the calls on who will win and by how much. If we do that, these polls may prove useful after all.