Earnscliffe Election Insights 2019 - Issue 2

Earnscliffe Election Insights 2019 – Issue 2


Welcome to our second issue of Election Insights, taking a look at the first few days of the campaign and casting ahead to how the polls, narratives and ballot questions are being shaped.  Across Earnscliffe, our multi-partisan advisors help look beyond the headlines and explain what’s behind the policy, party and specific strategic issues and their impacts.

Politics, predictions and the quirky polls of 2019. 

By Doug Anderson

For those who follow politics, the deluge of polling results released during a campaign can be exciting, overwhelming and at times, perplexing.  Depending on one’s hopes or affiliation, one can be encouraged by Company A’s poll one day and discouraged by Company B’s poll results the next.  Worse, people will be frustrated and confused when they see two polls conducted at the same time showing different or even conflicting results.

I can hear the cries of “Hang all pollsters!” already.  And pollsters get it.  Some outrage and skepticism are understandable.  It’s never been easier to conduct a poll, but it’s also never been harder to conduct a good poll.  However, before we throw all the polling evidence out with the pollsters, let’s make sure we know what we are looking at and how to fairly understand each poll and decide “what is good.”

First, remember that in Canada, the 30s are a “finnicky” polling zone.  Getting two percentage points more or two percentage points less when a party is in the 30s can translate into a heck of a lot of change in electoral outcome.  When a poll says there is 36 per cent support for a party, the reality at that instant might reasonably be 38 per cent and flirting with a majority…or 34 per cent and possibly not even forming government at all.

We’re likely to see a LOT of results in the 30s (certainly for the Conservatives and Liberals), so take a deep breath, remain calm and be ready to scrutinize each poll carefully.

Second, scientific research is all about replicability.  Two pollsters who do exactly the same thing at exactly the same time, should deliver results that nearly replicate each other.  Having support for a party two per cent higher than in another survey is entirely expected and need not disqualify either study as inaccurate.

However, and this is important to help you win arguments about election polls, no two polling companies seem to be doing the same thing these days, making it more difficult to compare one poll with another.

Reviewing the methodologies and approaches used by each of a dozen companies that have released polls recently, one would expect to see general commonality, but in fact, no two are identical…and the differences from one pollster to another can be pretty important.

The differences go well beyond the question of how genuinely representative of the population their study is: whether they are doing an online poll, using traditional live interviewers (Computer Assisted Telephone Interviewing) or robocalls (Interactive Voice Response).  The differences include:

  • How is the sample constructed that defines who is being polled?  Online polls recruit respondents from different sample providers who have created their panels differently and may produce unique skews;
  • How will different questions prompt different responses?  While a few pollsters are using fairly similar question phrasing, many are using unique phrasing, including some downright strange approaches;
  • Are the responses of all voters included?  Some just ask and report on “decided” voters, while others ask the undecided respondents a secondary question on “leaning” and report on “decided and leaning” voters;
  • Does the pollster prompt all the party names?  Almost all prompt respondents with all five/six (depending upon region) “major” parties—Conservative,  Liberal, NDP, Green, Bloc Quebecois, People’s Party of Canada (PPC), but some exclude PPC and one does not prompt for any party at all;
  • How are the response categories framed?  The response categories respondents are given vary, with a few providing a leader’s name as well as the party and the party names themselves are expressed in a few different ways. By the way, including leaders’ names is a potentially biasing influence.  No one gets to vote directly for a party leader so the appropriateness of including their names is highly debatable.  Finally, some pollsters provide a variety of other response categories (e.g., will not vote), others identify the parties and provide little response option beyond that.

Many of these variations should cause little or no impact on a respondent’s answer, but some will in at least a very small way and without more evidence, it’s impossible to say which approach is causing a skew in the results in any particular way.

Since the leading parties will likely be in the finicky 30s and several parties will be battling for smaller percentages, a very small difference from one poll to the next-maybe even the difference caused by including the name of the party’s leader (which again, does not appear on a ballot)—can look like a really important shift in results.

At the very least, know that although we will be comparing apples to apples among voter intention polls, one will be a Macintosh, another a Granny Smith, and another will be a Red Delicious, so don’t assume all day-to-day differences between different pollsters are reflecting actual voter shifts.

Unless or until a lead opens up for one party or another, if you see a slight change in standings one day to the next, temper your emotions by considering that it may not actually be due to a change in support.

Simple, eh?

The economy, cost of living and affordability

In last week’s Election Insights, we noted that the Conservatives intended to target the Liberals on Mr. Trudeau’s strong suit, the middle class, by focusing on the cost of living and appealing to those Canadians anxious about their economic future. Internal polls for the major parties have identified the affordability of daily life as the primary public concern. In the first week, the policy initiatives of virtually all parties signaled that this will be a major theme of the campaign, along with measures designed to help consumers by regulating fees and costs levied by a variety of service providers.

If Bill Clinton’s campaign guru James Carville was right in 1992 when he proclaimed, “It’s the economy, stupid,” the Liberals are starting with a very strong hand. Since 2015, the Canadian economy has created more than a million new jobs, an impressive record. The Liberals can also speak glowingly about the Canada Child Benefit and their middle-class tax cut.  To protect future prosperity and support growth, the Liberals have also completed the Comprehensive and Economic Trade Agreement (CETA) with Europe, negotiated Canada into the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTTP) and secured the Canada-United States-Mexico Agreement (CUSMA) against very significant odds.

Still, many Canadians are afraid for their future and the future of their children: widespread apprehensions about economic security persist among a wide swath of Canadians. These are prompted by a variety of issues, including job losses due to plant closures, slowdowns in the resource sector, possible fall-out from Trump-inspired trade wars, the burgeoning gig economy and the future job threats resulting from robotics and the growth of artificial intelligence. For the most part, political parties and governments cannot prevent these kinds of economic disruptions and dislocations, but they can spend and regulate to ease their pain on voters and consumers.

Into the bidding war

In a move to boost the purchasing power of home buyers in high-cost markets, the Liberals last week announced a tweak to their First-Time Home Buyer Incentive Program, which provides a top up for buyers with a down payment of less than 20 per cent of the purchase price.  The proposal would raise the cap on the price of eligible homes from $500,000 to $800,000 and allow buyers with a maximum combined income on $150,000—up from $120,000—to be eligible for the incentive.  They also promised a national tax on foreign speculation to drive housing prices down as they have in British Columbia. The Liberals followed up with a promise to create hundreds of thousands of before- and after-school child care spaces for children under ten years of age and to lower child care fees by ten per cent across the country.

In the past week, the Conservatives rolled out three tax credit promises aimed directly at hard-pressed voters, the creation of a maternity and parental leave benefit and the reinstatement of the public transit tax credit and the children’s arts and learning tax credit, both which were Harper initiatives abolished by the Liberals.  Andrew Scheer promised a major reduction in the tax rate on the lowest tax bracket, taxable income under $47,630 from 15 per cent to 13.75 per cent.  The party said the cut would save the average single taxpayer about $444 per year and two-income couple about $850 per year over a four-year phase-in.   By choosing this form of tax cut, the Conservatives ensured that the program would be universal with everyone gaining the benefit regardless of income.

The NDP has waded into the cost of living/affordability territory too.  The NDP vision document promises investments in health care, affordable housing, and childcare to “tackle the things that are keeping people up at night worrying – and make sure everyone can afford a good life.”  First up last week was a plan to cap cell phone and internet prices to save families an estimated $250 per year and the introduction of a Telecom Consumers’ Bill of Rights.  The NDP also promise to “require service providers to offer basic plans that meet the needs of Canadians, end caps for internet plans and require companies to offer affordable, unlimited data plans for cellphones.”

The Green Party released its voluminous platform yesterday with several measures that deal with affordability and the anxiety over the future of work. They promised to cap credit card interest rates, limit ATM fees and force greater competition and lower prices for telecom services. They also promised significant investment in transition programs to deal with workers being forced out of work by the change to a green economy.

These promises are just the beginning, a down payment on what the leaders will be announcing in the coming days and weeks.  One thing is certain: some of these commitments are hugely expensive.  The Parliamentary Budget Office estimates the Conservative tax cut promise will cost about $14 billion in lost revenue between 2020-21 and 2023-24.  The Liberal child care promises are un-costed as yet but look to be substantial.  All of this raises the stakes on the parties’ tabling of their rolled-up costed promises later in the campaign.  It’s easy to make promises; explaining how they will be paid for is another thing!


Perils of Bubble Campaigning

By Rick Anderson

There are three circumstances in which political parties adopt “bubble campaigns”, meaning playing it safe and shielding their leaders from pesky embarrassing questions from journalists, debate hosts, one another and voters at large:

  • They are frontrunners (or think themselves such), looking to nurse the campaign across the finish line with the least-possible risk (can be a dangerous strategy if deployed for more than a few days);
  • They wish to avoid particular questions or topics; and
  • The campaign lacks confidence in the leader’s knowledge and/or skills.

There are certainly those around who paint the Liberals as frontrunners, and perhaps that is the dominant sense in their war room. But I think they are smarter than to believe that. This is a close election, our closest in years.

Public opinion data do not really support the notion of Liberals as frontrunners. The Liberals and Conservative are locked in a virtual 34-34 tie today, as indeed they have been for the eighteen months since the unusually long Trudeau honeymoon came to an end in March 2018. The Liberals did occasionally see leads of five or six points during autumn 2018, but by January of this year those had shrunk to just two or three points. And that was even before the SNC-Lavalin issue exploded into headlines and resignations.

Since then, throughout the spring, the Conservatives led the Liberals by as much as six points. During the summer, the Liberals scrambled back to today’s margin-of-error dead heat[1].

Some seat projections may give the Liberals more assurance of victory, thanks to contemporary “wisdom” that vote distribution favours the Liberals. This of course depends on using national data to look at local races, a dangerous proposition. It also depends greatly on turnout assumptions, of which almost no one speaks although who turns out is often determinant to outcomes.

Young people helped Trudeau win in 2015 and may do so again in 2019. That’s a comforting memory for Liberal souls.

On the other hand those with longer memories recall 1972, Pierre Trudeau’s first re-election campaign. Relying upon “The Land is Strong” campaign theme, Trudeau’s Liberals beat Stanfield’s Progressive Conservatives by three points (38% to 35%).

But Pierre Trudeau barely eked out a minority win, with 109 vs 107 seats.

Or, the nightmare 1979 election, which Pierre Trudeau lost to Joe Clark, 136 seats to 114. The popular vote? Liberals actually won that, 40% to 36%.

Even if all types of voters turned out evenly (a rarity) the idea that any party can win a majority with just 34% begs analysis.

The three slimmest-vote majorities in the last sixty years were three of Canada’s most recent majorities:

  • 1997    Jean Chrétien             38.5 %
  • 2015    Justin Trudeau            39.5 %
  • 2011    Stephen Harper          39.6 %

CBC’s poll-tracker currently has the Liberals in second-place (really, tied) at 33.6%.[2]

The only time any Canadian has ever won a majority with a number anywhere like that was John A. Macdonald in 1867.

John A won with just 34.8% – but he did that by:

  • Running simultaneously as the leader of BOTH the Conservative Party and the Liberal-Conservative Party (winning 71 seats under the Conservative banner and 29 seats under the L-C banner);
  • Running against a Liberal Party with no official leader, and whose informal leader, elder statesman George Brown, was not then nor ever elected to Parliament (nor any legislature pre- or-post-Confederation); and
  • Running against 143 candidates of independent or unknown affiliation, precisely zero of whom were elected.

That bizarre 1867 setup is the only “precedent” for predicting that today’s poll numbers translate into a majority for anyone.

Published polling data should not be giving savvy Liberal campaign strategists that frontrunner feeling, nor the sense that a peek-a-boo campaign with limited leader availability is in order.  More likely, they appreciate their campaign needs exactly what Andrew Scheer’s campaign needs: to gain an important four or five additional points of public support. No one is going to relax in this campaign until they pull into the 37-38-39-40% range typically needed to form a majority government.

So why the peek-a-boo campaign?

Two other plausible theories come to mind.

First, the Liberal war room must have watched last week’s campaign launch unfold with dismay, dominated as it was by questions about the hopefully-past SNC-Lavalin topic, RCMP enquiries and blocked access to sought materials.

Second, approval ratings.

It is hard to recall a campaign in which the leaders of all three of the heretofore traditional parties had net approval ratings sub-zero. But that’s what we have today.

According to the CBC’s “Leader Meter”[3], the average of the last ten polls has these four leaders’ net rankings (approval minus disapproval) ranked this way:

  •  Elizabeth May            + 17
  • Jagmeet Singh            – 10
  • Andrew Scheer           – 12
  • Justin Trudeau            – 22

As things stand, the fact is, no one yet enjoys frontrunner status in this election. One suspects the voters may like it that way, for now at least.

Likewise, all the participants know that, if they’re as smart as we think they are. And, one suspects, most likely their hunches are like Yogi Berra’s – that this election won’t be over till it’s over.


LISTEN: Earnscliffe’s Kathleen Monk, NDP strategist, joins Vassey Kapelos from CBC’s Power and Politics to discuss party strategy when vetted candidates come under scrutiny.




WATCH: As candidates from various parties begin to drop out of the race, Earnscliffe’s Charles Bird, Liberal strategist, joins CBC News to speak on how the prevalence of social media affects the vetting process. 


Campaign taglines and the ballot question

By Geoff Norquay

The fight to establish the ballot question is the critical task for every party in a national election.  The ballot question is each party’s key differentiator—the strategic phrase that summarizes its positioning vis-à-vis its competitors, and frames the choice to be made by voters.

Once the ballot question is settled, it must be effectively framed in a tagline or slogan, a phrase that sums up the essence of the campaign for each party.  Hopefully, that slogan is catchy and concise, even better if it incorporates a rhyme or is alliterative.  It must also be accessible—easy to understand and remember. If it makes people pause and think for a moment because it can be read on several levels, it will likely be even more effective.

Political parties around the world have created some memorable campaign slogans over the years, some successful and some not so much:

  • A new deal: Franklin Delano Roosevelt, U.S. Democrats, 1932 – The first modern tagline, it completely changed political campaigning.
  • I like Ike: U.S. Republicans, 1952 – A play on the nickname of their candidate, Dwight D. Eisenhower.
  • The land is strong: Canadian Liberals, 1972 – Actually, the land wasn’t strong; it was in a recession and everyone knew it. The slogan was ridiculed and Pierre Trudeau was reduced to a minority.
  • Keep the bastards honest: Australian Democrats, 1977 – The direct approach!
  • Labour isn’t working: U.K. Conservatives, 1979 – Combined with a picture of a long unemployment line, it was hugely effective. Conservatives claim it won the election for Margaret Thatcher.
  • The answer is Liberal: Australian Liberal Party, 1990 – A bit of a disaster: Labour Leader Bob Hawke demolished it with the comment, “If the answer is Liberal, it must have been a bloody stupid question!”
  • HOPE: Barack Obama, 2008 – Twinned with a brilliant image of Barak Obama, it was simply iconic.

In the current campaign, no party has hit that home run, but they have displayed some strategic thought:

  • The Liberals are going with “Choose forward,” which in the early days of the campaign has allowed them to invite comparisons between themselves and Stephen Harper and Doug Ford, even though they are not on the ballot.
  • Andrew Scheer and the Conservatives have chosen “It’s time for you to get ahead,” which links directly to their campaign emphasis on the cost of living, affordability and voters’ concerns about future job security.
  • The NDP has chosen “In it for you,” which allows them to characterize both the Liberals and Conservatives as being more concerned about vested interests and big business than common folk.
  • The Greens are going with “Not left. Not right. Forward together,” an appeal for voters to leave the tried and true and choose a new banner.
  • The People’s Party has chosen “Strong and free,” a traditional approach for party appealing to traditional values.

May the best tagline win!

Media and the search for narrative in Election 2019

By Elly Alboim

An election is often a contest between competing story lines or narratives. Politicians try to create a ballot question, weave a story around it and associate themselves with it. Media does the same but with different criteria and objectives.

Media culture favours story telling as its primary vehicle for organizing and delivering information. It tries to create story arcs to situate facts, events and personalities within compelling dramatic narratives to make them more understandable and attractive to news consumers. The end result, as many media critics have said, is to have the interesting crowd out the important.

Media approaches that task in predictable ways. It favours “news” (and its implied immediacy and urgency) despite elections being more about repetitive and incremental exposition of policy and promise by the candidates and their parties.

The more significant strategy though, is the imposition of media narrative to frame the exercise and imbue it with compelling features to generate interest. The media narrative has two dominant components — the state of the competition (the horserace) and as been true since the days of Greek drama, the story arcs that characterize the personalities who are the leading actors in the overall drama.

Polling has established that currently the outcome is too close to call and may not even produce a clear result. It has legitimized that most prized story line: the horserace or sports event with all of the opportunity that creates for focusing on strategy, performance and momentum.  And in a two for one special, there isn’t just the race for first, there is also the race for third between the NDP and the Greens.

As Week One ends, the attempts by leaders and parties to frame their ballot questions and policy announcements have taken a decided back seat to horserace handicapping.

What hasn’t yet solidified is the other leg of the media narrative stool – the story lines defining the leaders, although some are emerging.

Justin Trudeau

In 2015, Mr. Trudeau was the “new politics” insurgent promising an end to the dour Harper years. Obviously, that is well behind him now. His efforts to have the electorate look ahead have been blunted by media (and Opposition) insistence on looking back. He can yet become the hero who overcame significant difficulties to once again be the peoples’ choice. Or he might represent the classic fall from grace of a leader struggling for redemption and a second chance. Less dramatic and perhaps most likely, would be the simple story of contrast – becoming the default choice among a set of leaders who fail to inspire.

Andrew Scheer

Media has struggled as well in trying to find a consistent way to characterise Mr. Scheer. At this point, the most common framing is that he is quite low key and has yet to define himself or be defined by others.  Liberals have worked hard to demonize him as a dogmatic social conservative and an acolyte of cutting back government a la Stephen Harper and Doug Ford. Media has reported on those themes but thus far at least, seems unpersuaded.

Mr. Scheer himself is promoting the narrative that he is the leader who will make life more affordable for Canadians.  By defining his overall framing that narrowly, Mr. Scheer leaves open broader questions about how he would manage the economy in uncertain times and how he would represent Canada on the world stage.

Jagmeet Singh        

Jagmeet Singh’s debate performance has made a difference in media narrative largely because of the inappropriately low expectations they held of him. Although the audience for the debate would have been quite small, the emerging media story line for Mr. Singh seems to be the clearest of all at the moment. Pending some significant set-back or diminishing poll results, he has become the 2019 election turnaround personality – exceeding expectations, performing well and rejuvenating the NDP’s chances.

Elizabeth May

Ms. May’s first week has suggested that she faces a lot of scrutiny and hard assessment as media try to take her measure. The fact that the Greens are more competitive and may hold the balance of power has turned their attention to her policy set, team and set of personal convictions with decidedly mixed results. In one short week, media framing of her has shifted to focus on her vulnerabilities as opposed to her core proposition.  There is no doubt of her authenticity, fluency and comfort debating, but the first week has shown she is in danger of being marginalized by media with a narrative of overreach and being not quite ready for prime time.

Media framing matters

Media framing is important to shaping the world for people who are not deeply partisan or ideological. Media cannot tell people what to think – they are more than capable of deciding for themselves. But it does provide clues about what is important for them to think about. And media narratives tend to become conventional wisdoms.

Click here for a link to the full article on Capital Current. 


LISTEN: Earnscliffe’s Mary Anne Carter, senior consultant, sat down with Amanda Connolly of Global News’ Parliamentary Bureau to discuss vetting election candidates in the age of social media on their new podcast, “Wait, There’s More.”




LISTEN: Earnscliffe’s Yaroslav Baran, crisis communications advisor, weighs in on what happens when crisis hits during a campaign launch, and how to move past controversy and back onto your message track on CBC’s The House.


The Polls

In Formula One racing championships, there are usually a couple of teams with a realistic chance of victory and everyone else jostling to be the best of the rest.  In this federal election, we have two teams with a realistic chance of winning, but unlike F1, how the rest of the parties do will actually impact who ultimately wins this election.

The latest polls tell us that the Liberals and Conservatives are neck-and-neck and neither appears to have more than 35% of the vote.  The lowest proportion of vote that has ever won an election in Canada was the 35.9% recorded by the Progressive Conservative Party in 1979.  In that election, the Liberals actually received many more votes (40.1%) but only won the second most seats.

As it stands, the performances of the NDP, Green, BQ and/or PPC – particularly, regionally – are key metrics to watch.

Election 2019 Key Dates

  • September 26, 2019, 2:00 p.m. local time – Candidate Nomination deadline
  • October 1, 2019 – Munk Foreign Policy Debate
  • October 2, 2019 – TVA French-Language Leaders Debate
  • October 7, 2019 – Advance Polls Open
  • October 7, 2019 – English-Language Leaders Debate
  • October 10, 2019 – French-Language Leaders Debate
  • October 21, 2019 – Election Day


READ : Earnscliffe’s Robin Sears, NDP strategist, argues that candidates must be honest about their past during the election campaign vetting process in his latest column for the Toronto Star.


[1] https://newsinteractives.cbc.ca/elections/poll-tracker/canada/

[2] https://newsinteractives.cbc.ca/elections/poll-tracker/canada/

[3] https://www.cbc.ca/news2/interactives/leadermeter/