OPINION | How the Harper model worked, and why it's now broken

It is the irony of Conservative politics that people in Calgary need to worry about how the Conservative party is selling itself around Toronto.

To win the next election, Conservatives need to add voters in the 905 without alienating voters in Calgary

Prime Minister Stephen Harper waves to the crowd after winning a majority government on election night in 2011. According to Ken Boessenkool, the strategy used to win in 2011 won't work today. (Jonathan Hayward/The Canadian Press)

This column is an opinion from Ken Boessenkool, one of the original authors of the Alberta 'firewall letter' and a former senior campaign advisor to prime minister Stephen Harper.

The Conservative Party of Canada is now in the early stages of a leadership selection process. 

This is the result of a number of factors, the most important of which is the desire by Conservatives — particularly Western Conservatives — to see themselves once again reflected in the national government as they did during the halcyon Harper era.

If that's to happen, Western Conservatives are going to have to come to grips with a difficult fact, namely that the Harper model for winning elections is broken.

Harper built a coalition of voters and regions that got him a minority (2006), another minority (2008) and then a majority (2011). That majority required a coalition of targeted Conservative voters across targeted ridings across targeted regions.

At the broadest level, Harper's model combined a strong British Columbian and prairie base with a smattering of seats in Quebec and Atlantic Canada and a good base of seats in rural eastern, southwestern and northern Ontario.

The critical difference between winning and losing for Conservatives has been one particular region: the 905. That region is a collection of suburban Ontario ridings surrounding Toronto. Harper swept the 905 in 2011, but was wiped out there in 2015.

It is the irony of Conservative politics that people in Calgary need to worry about how the Conservative party is selling itself around Toronto. It's something that Harper understood. It's also something Conservative voters should understand as they consider their choices in the leadership race.

Because any prospective leader who doesn't have a plan to do that, merely has a plan to lose again.

How Harper won: The 905

The Harper strategy for winning the 905 was first outlined by the brilliant political strategist Patrick Muttart in the aftermath of Harper's first attempt to form government in 2004, an election Harper lost. 

In the wake of that election, Harper went dark and hinted publicly and privately that he would resign unless his campaign team fixed the many mistakes made during his maiden effort. 

Muttart, who had played a relatively minor role in Harper's 2004 campaign, was integral to building the new strategy.

Conservative leader Stephen Harper throws a balloon in the air following his speech to supporters on election night in 2004. After losing the election, Harper hinted that he would resign unless his campaign team fixed the many mistakes made during his maiden effort. (Adrian Wyld/The Canadian Press)

His central insight involved breaking the voting public into a series of voter groups that dominated swing ridings.

Along with pollster Dimitri Pantazopoulos, Muttart created groups that were needed — and equally importantly, those that were not needed — to get to victory.

How Harper won: voter groups

Muttart created three broad groups: core voters, swing voters and non voters. Each communication piece or ad was targeted at one or more of those groups. 

For some voter groups, the strategy was to increase turnout. For others it was to convince them to vote for us rather than the other party. For the rest, the goal was to tarnish our opponents enough that they might not vote at all.

Each category of voters contained two to four archetypes. Each archetype had its own name and picture that were spread through the campaign office, and each communication piece or ad was targeted at an archetype. 

We knew precisely what proportion of each archetype we needed to either attract from our political opponents or entice into coming out to vote to get us to a target vote number.

Or, to put it another way, we needed to get enough core and swing voters to achieve a minimum target national vote number, one that Muttart calculated would be necessary to win a majority government. 

That number was below 40 per cent.

This analysis was particularly detailed in — indeed it was largely driven by — 50 swing ridings that would get us to a majority. 

The biggest group of swing ridings, by far, were in the 905. These urban and suburban ridings around Toronto held enough of our swing voters that, if we attracted them or increased their turnout, it would give Harper his majority government.

Who were those swing voters?

'Dougie' and 'Heather'

Reaching our vote targets required that we increase the turnout of "Dougie," the beer-swilling, single construction worker.

Dougie had voted in three of the last seven elections. He voted when he was angry at government waste. When he voted, he was fairly reliably a Conservative voter, but he wasn't a reliable voter. 

And so the campaign ran ads on TV programs Dougie watched. Ads that focused on the sponsorship scandal and the "billion dollar boondoggle" at HRDC. 

Also, the more pictures of Stephen Harper holding a beer (he didn't enjoy drinking them), driving a skidoo or a farm quad, the better — particularly if they ran in the Sun chain of papers that Dougie checked for sport scores.

We also needed to attract "Heather." She was a married mother with three kids who lived in the 905 region.

Heather worked part time at Canadian Tire as a cashier and her husband was an accountant at a small local firm. Heather was concerned about balancing the family budget while doing the best for her children. Every year she had to decide whether her kids played soccer or took piano lessons.

She always voted, and was open to voting Conservative, but had not consistently done so. The thinking was she could be attracted away from the Liberals with policies that showed the Conservative Party understood the challenges she faced. 

So the party ran ads on the "$100 per month for each child under six" and "the children's sports and arts tax credit" policies on TV programs popular with soccer moms.

This voter targeting was brilliant so far as it went, but it relied on one external factor almost entirely beyond the party's control.

The strength of the NDP.

Vote splitting

Particularly within the 905, the NDP vote was critical.

In a three-party system, winning a majority with less than 40 per cent of the vote requires the third-place party to garner somewhere north of 20 per cent of the vote. As the NDP vote rose above 20 per cent in the 905, it drained votes directly from the Liberals and gave seats to Conservatives.

Heather and Dougie and the electoral math won the day for Harper in 2011. But it won't work again. 

The model is broken.

Dougie and Heather are doing their part. The NDP isn't.

Conservative leader Stephen Harper walks off the stage after addressing the crowd on election night in 2015. The collapse of the NDP delivered nearly the entire 905 – and the government – to the Liberals. (Jonathan Hayward/The Canadian Press)

In 2015 Harper, then nearly 10 years in office, dropped a mere three-and-a-half percentage points in the popular vote in the 905, compared to 2011. But the collapse of the NDP from 25 to 10 per cent delivered nearly the entire 905 — and the government — to the Liberals.

In 2019, the Conservatives shed another four points, which was larger than the very small NDP gain of less than two points, and as a result shed another couple of seats in the 905.

The key point is worth summarizing: Had the NDP held their 905 vote from 2011, Harper would have remained prime minister in 2015 and Scheer would be prime minister today.

Winning the next election

If Conservative voters across Canada want to win government, it will require a critical modification to the Harper model. 

The party will need to shift from a three-party 905 strategy, to a two-party 905 strategy. They cannot assume that the third (and fourth and fifth) parties will skim the necessary number of votes off their primary opponent.

That means the Conservatives will need to increase their vote total by at least 10 per cent in the 905 to form the government.

Can it be done?

To find out, prospective Conservative Party leaders should be asked two questions.

First, who are the new voters you will attract to get the party an additional 10 per cent of the vote in the 905?  These groups will be the new archetypes the party will need to target.

While this is a question where detailed polling will lend a strong hand, the answer will almost certainly be some combination of young professionals, new Canadians and new, aspirational mothers. 

Conservative Leader Andrew Scheer acknowledges supporters following a speech on election night in 2019. Had the NDP held their 905 vote from 2011, says Ken Boessenkool, Scheer would be prime minister today. (Adrian Wyld/The Canadian Press)

Second, how do you propose to attract these voters into the Conservative coalition? 

This will require hard work, keen analysis and old-fashioned intuition. It will mean finding out what their core concerns are, and then proposing conservative solutions.

The offerings do not need to be restricted to a specific grab bag of new policies. They could include a new approach and presentation of policy (moderation, incrementalism) or a shift in tone. 

But the party has to add and not subtract.

Where social conservatives fit in

Shedding social conservatives (likely at least 10 per cent of the Conservative vote) as a strategy to gain a new 10 per cent of the vote, will just result in treading water.

Conservatives will also need to add voters in the 905 around Toronto without alienating the voters in Calgary. 

Success does not require abandoning conservatism. In fact, it requires that we hold on to existing social conservatives, fiscal libertarians and traditional red Tories. Nor does it mean riding some temporary populist wave.

Instead, growing our vote in the 905 by 10 per cent will require the application of conservative principles to the concerns of that additional 10 per cent of voters.

That's not easy. But for the next leader of the party, it's necessary. 

This column is an opinion. For more information about our commentary section, please read this editor's blog and our FAQ.


Ken Boessenkool is a lecturer at the Max Bell School of Public Policy at McGill and a Research Fellow at the CD Howe Institute. He was one of the original authors of the so-called Alberta Firewall letter in 2001. He was a senior campaign adviser to prime minister Stephen Harper, and has worked or volunteered for Preston Manning, Stockwell Day, Jim Dinning, Ric McIver and Christy Clark.