Election autopsy

Opinion

Hey there, time traveller!
This article was published 25/10/2014 (3783 days ago), so information in it may no longer be current.

As the polls closed at 8 p.m. Wednesday, little did Winnipeggers know they were about to experience a political earthquake trembling beneath their feet.

The expectation among most Winnipeggers following this year’s municipal election campaign was that one of two possible outcomes would unfold by the time the votes were counted: Either Judy Wasylycia-Leis, the career politician and presumed front-runner for more than a year, would be sitting in the mayor’s chair, or privacy lawyer and rookie candidate Brian Bowman would pull off a shock upset and snatch victory by a narrow margin.

 

What transpired, though, turned out to be much different.

At 8:15 p.m., the first advance voting results came in showing Wasylycia-Leis with a 780-vote lead over Bowman, with Gord Steeves sitting slightly ahead of Robert-Falcon Ouellette. At about 8:35, Wasylycia-Leis’s narrow lead flipped into a 3,500-vote cushion for Bowman, with Ouellette firmly entrenched in third. And from there, the rout was on, as Bowman’s vote total soared as he left the rest of the field in his dust.

Not even the most starry-eyed and optimistic supporters of mayor-elect Brian Bowman would have expected their candidate to trounce Wasylycia-Leis so handily.

So what happened?

In my view, there were several factors that contributed to Bowman’s huge electoral victory:

 

Voter turnout variations

Municipal elections typically have lower turnouts than provincial or federal contests. There are also huge variations in turnout across Winnipeg.

Turnout data for the 2010 and 2014 municipal election campaigns showed turnout has been highest (55 to 60 per cent) in relatively affluent suburban areas on the edge of Winnipeg, slightly lower (in the 40 to 50 per cent range) in older suburban areas such as East Kildonan and St. James, and below 40 per cent in the downtown core, Point Douglas and the North End.

There is also a strong relationship between voter turnout and household income, with those living in some of the city’s most affluent neighbourhoods two or even three times as likely to cast a ballot as those living in the poorest areas of Winnipeg.

Data from two polls conducted by Probe Research during the campaign showed Wasylycia-Leis drew her highest levels of support from inner-city neighbourhoods as well as among those with lower annual incomes, while Bowman had much higher levels of support in southwest Winnipeg and among wealthier voters.

Although the sample sizes are small, there is a clear and obvious pattern: Bowman had significantly more support in wealthier, suburban areas where turnout was high, which was not fully reflected in polls that proportionally sample the city. What this means is the current way of drawing a regional sample from the city assumes the same proportion of people living in the William Whyte or Centennial neighbourhoods is the same as the proportion of those who will turn out in Tuxedo or Sage Creek — and clearly, this is not the case. This is an issue that we, as pollsters, will study in more depth and revisit in future election campaigns.

 

The ‘machine’ broke down — again

One of the accepted truisms of Manitoba politics is that the New Democratic Party and candidates affiliated with the party benefit from the assistance of the so-called NDP machine — a highly organized and effective group of NDP partisans and their allies in the labour movement who always get their voters to the polls on Election Day.

From 2010 to 2014, Wasylycia-Leis’s raw vote total plummeted by about 30,000 votes, with the “machine” unable to bolster voter turnout for her in areas where she should have had significant support.

Upon closer review, the recent record shows the vaunted NDP machine is not as effective as it is often made out to be. Going back to 2010, this machine has failed twice to deliver victory to Wasylycia-Leis in a municipal campaign; lost Wasylycia-Leis’s former federal riding of Winnipeg North in a subsequent byelection, and then lost longtime NDP bastion Elmwood-Transcona to the Conservatives.

The only win for Team Orange in the past five years was in the 2011 provincial election, which is increasingly starting to look like it was the result of facing a weak opponent (the Hugh McFadyen-led PCs) with an incoherent, confusing message as well as, quite frankly, extremely good luck (the NDP improbably won every potential swing seat in Winnipeg, including two — St. Norbert and Kirkfield Park — by only a handful of votes).

The bottom line is the NDP machine needs to spend some time in the shop, as it is breaking down in part due to factors such as limitations on how much unions can engage in political activities, as well as a lack of volunteer horsepower and an over-reliance on a relatively small group of campaign professionals. Obviously, this doesn’t bode well for the provincial NDP 12 to 18 months before the next provincial election.

Through hard work and good fortune, Bowman managed to position himself perfectly by the time the campaign entered its final, critical phase

 

Second choices and soft support

What was remarkable about polls conducted prior to the campaign was that Wasylycia-Leis’s support did not change much from the time she lost the 2010 election to the end of September: she received 43 per cent of the vote in 2010, and the last Probe Research poll conducted nearly a month prior to the election showed her with 42 per cent support among decided voters. In the end, she finished with only one in four Winnipeggers (25 per cent) casting a ballot for her.

So what happened in the past three weeks? It is apparent that whatever support she had that was reflected in earlier polling was never as locked-in as it appeared to be.

It is also worth noting a Probe Research survey conducted in August showed that among the entire field, Bowman stood out as being the most popular second-choice candidate. If we only concentrate on non-Bowman supporters who intended to vote for one of the other candidates, 28 per cent of those surveyed said Bowman was their second choice. In that same survey, Bowman and Ouellette were the only two candidates to have a significant percentage of Winnipeggers indicate they had a more positive view of them compared to when the campaign started.

Although Bowman’s overall support in this particular poll wasn’t that high (only 16 per cent, in third behind Steeves), there were indications that unlike Wasylycia-Leis or Steeves, he had the potential to grow his support to a much greater extent than his principal opponents. And, if many voters were genuinely undecided — or tepidly considering whether to vote for Wasylycia-Leis or another candidate — Bowman was a likely second choice if the case was made for them to consider other options.

 

#bowmentum

Looking back, Bowman benefitted from steady growth in public support, followed by a massive late shift in momentum when it truly mattered.

He entered the campaign as a relatively unknown first-time candidate: In a December 2013 Probe survey, only 11 per cent of decided voters indicated they would vote for him (compared to 45 per cent for Wasylycia-Leis). Another Probe poll conducted at the end of August placed him in third place with 16 per cent, while the last survey conducted in late September put him a clear, yet distant second at 23 per cent.

Bowman ran a tightly scripted campaign that was not only highly focused in terms of its messaging, but also made great use of data, particularly from social media, to identify potential supporters. Through a great deal of hard work early in the campaign to introduce himself to Winnipeggers and develop his political brand, he managed to position himself perfectly by the time the campaign entered its final, critical phase. His second-place showing in the last Probe poll and the collapse of the Steeves (and to a lesser extent, Paula Havixbeck) campaigns allowed him to successfully frame himself as the only candidate with a credible chance of defeating Wasylycia-Leis. He also benefited from the fact Ouellette’s support remained consistent and his supporters did not shift to Wasylycia-Leis in order to keep Bowman from becoming mayor.

It is also important to note that Wasylycia-Leis’s decision to more or less ignore the potential threat Bowman posed turned out to be a major strategic blunder, as she did not effectively parry his criticisms of her as an “old-school politician,” nor did she effectively convince supporters of other candidates — particularly Ouellette’s — to also vote strategically in order to keep a Tory-supporting, business-connected yet completely inexperienced and raw politician from becoming the next mayor. The effect was that the progressive side of the political spectrum split, which was ironic considering how much early hand-wringing there was about Bowman, Steeves and Havixbeck splitting the conservative vote.

But by the time the race became framed as a two-way fight, it was too late for her to effectively stop the so-called Bowmentum that shook the city’s political foundations and carried him to the mayor’s chair.

 

Curtis Brown is the vice-president of Probe Research Inc., a Winnipeg public opinion firm, and a former political journalist/commentator with the Brandon Sun and Winnipeg Free Press.