Harper and the Mulroney Coalition

In two items in the May 23rd editions of the Globe and National Post, the narrative about a coming Harper majority based on breakthroughs in Quebec is once again articulated.

In the Globe Harper advisor Tom Flanagan suggests the coalition was outlined in a Stephen Harper speech 10 years ago:

“Mr. Harper marshalled historical evidence to show that all winning Conservative coalitions in 20th-century Canadian history had been built around three main elements: populist reformers, strongest in the West but also present in rural Ontario; traditional Tories, strong in Ontario and Atlantic Canada; and francophone nationalists in Quebec. “

The problem with this generalization is that there are only three cases and this fact alone makes generalizing problematic. One case, Borden’s victory in 1911, came prior to the World War I conscription crisis, which so defined Quebec politics in the 20th century, and should therefore be excluded. Flanagan’s argument does not accurately describe Diefenbaker’s 1958 sweep. Dief won Quebec votes based on the Quebec electorate wanting a share of power in an election where the outcome was generally known in advance. However, he was not at all successful in Quebec in 1957, 1963 and 1965. The unilingual Diefenbaker simply did not build a coalition with Quebec nationalists (although he received some support from the Union Nationale). Only Mulroney built the coalition Harper/Flanagan describe, sustaining it through two elections, and it always had the potential to disintegrate in the way it did in 1993 because the values of his western supporters and those in Quebec, while similar on free trade, were polar opposites on language and national unity issues. It does not have the potential to be a formula to make the Conservatives a “natural governing party” as suggested by Mr. Flanagan. Quebec might buy into Conservative decentralization but it rejects the rest of the package.

The big difference between today and 1984 is that there was no nationalist party in Quebec that stood between Mulroney and those nationalist votes. Now we have the Bloc Quebecois. Pro-independence nationalist parties have commanded thirty-nine percent or more of the Quebec vote, federally and provincially, since 1993 and 1976 respectively with the single exception of the Quebec provincial election in 2003 when the PQ received 33%, losing some soft nationalist votes to the ADQ.

The other obstacle facing Conservatives in Quebec are that province’s values. Whether it is Kyoto, Afghanistan or gay marriage, there are a host of issues where the Harper Conservatives are at odds with the views of a large majority of Quebeckers.

An Ipsos-Reid poll out May 23 is headlined as saying the Conservatives would win a majority if an election were held today. And indeed T.C. Norris’ forecast model agrees it would give the Harper Conservatives a majority of about 175 seats. But it is not based on gains in Quebec.

But in today’s National Post story on the poll we find the Quebec-as-source-of-majority narrative continuing:

“Basically what's happening is that Stephen Harper is recreating the Brian Mulroney majority," Ipsos Reid president Darrell Bricker said in an interview.
"And the way he is doing that is by breaking through in the province of Quebec. It's very much that kind of coalition -- Quebec and the West."

I think Mr. Bricker should study carefully the results of his own poll because it gives the Conservatives their greatest gains in B.C., Ontario and Atlantic Canada. That is where the majority is created. Indeed the Conservatives are in first place everywhere except Quebec. The Conservatives are up in Quebec, but they still trail the BQ and would pick up only six seats according to my model.

I think Harper could win a majority if an election were held today, but if he is going to do so, then this poll tells us it is likely to be outside Quebec. Harper needs significant gains especially in Ontario and Atlantic Canada plus smaller gains in B.C., combined with a more modest advance in Quebec, to win his majority.

This poll has a sample size of only 1,000 and therefore the regional numbers are suspect due to very high margins of error. To take just one example, the Ipsos poll says the NDP are at 11% in Ontario while two weeks ago, the SES poll had the NDP at 24% in the province. The SES poll would produce a strengthened Conservative minority of 139 seats, but nowhere near a majority.

The Ipsos poll also was taken prior to the Kyoto flare-up, and the Afghanistan vote and Gwyn Morgan fiascos, so it probably reflects the positive media around the budget, but not the more recent bad news days for the Conservatives.

The overall total for the Conservatives of 43% has a margin of error of ± 3.1% meaning it could be less than 40% and still be within the margin of error. The range of 40-43% is really in the area somewhere between majority and minority, so one cannot say for sure that this is the result the poll actually forecasts. I would say that a poll giving the Conservatives 45% is highly likely but not certain to predict a majority (given the margin of error).

Will our pundits notice the weakness and ambiguity of this week’s majority assertions? Since they seem incapable of noticing that Quebec is the one place where the Conservatives don’t have a lead, I don’t hold out much hope.