What really happened in the 2004 election, and can it happen again?
By Ira Basen, CBC.ca Reality Check Team | Dec. 20, 2005 | More Reality Check
We all know what happened in the 2004 election campaign, right?
We know that the Conservatives were cruising along, riding the wake of the sponsorship scandal towards a minority and possibly even a majority government, until the party got gob-smacked by some verbal gaffes by Stephen Harper (Paul Martin soft on child pornography) and erstwhile allies like Ralph Klein. Then the Liberals unleashed a nasty set of "attack ads" that successfully "demonized" Harper and "scared" voters away from the Conservatives.
In the last weekend of the campaign, there was a massive strategic shift of frightened NDP voters into the Liberal camp, resulting in a Liberal minority government that surprised pollsters, politicians and the media. Meanwhile, Quebecers were so angry about the unfolding sponsorship scandal that they deserted the Liberals in massive numbers and embraced the Bloc Quebecois.
We know all this because we have been told some variation of this story by the press and the pundits since the day after the June 28 election, and it remains the master narrative through which this current campaign is being interpreted.
But what if the story is wrong? What if we've been sold a bill of goods about June 2004? Would that not increase the possibility that much of what we think about how this campaign will unfold might be wrong too?
The possibility that the story on 2004 could be wrong is strongly suggested by the Canadian Election Study, a massive undertaking conducted by a well-respected group of non-partisan academics, headed by Prof. Neil Nevitte of the University of Toronto. Their findings were based on thousands of interviews conducted daily throughout the campaign. Let's look at what they found…
Conventional Wisdom: The Liberals' attempt to "demonize" Harper, to make him appear "scary," succeeded in the final days of the campaign, thanks in part to the negative attack ads.
Survey Says: In English Canada, almost half of voters surveyed agreed with the statement "Stephen Harper is just too extreme." But that perception was there from the beginning of the campaign, and significantly, Harper's ratings did not go down in the last days of the campaign. Moreover, the Conservative leader's rating was never much worse than Martin's. Attitudes about the Conservative party also remained relatively stable throughout the campaign. The attack ads did not appear to significantly alter voters' perceptions of Harper or his party.
Conventional Wisdom: The sponsorship scandal cost the Liberals their majority.
Survey Says: This is probably true, but the scandal's impact may be overstated. About 40% of English Canadian voters surveyed said they were "very angry" about the sponsorship scandal, and roughly another 40% said they were "somewhat angry." If the vast majority of those people took out that anger by voting for the opposition parties, the Liberals would undoubtedly have been defeated. But they didn't. A third of the people who voted Liberal in 2000 said they were "very angry" about the scandal, but half of them wound up voting for the Liberals again. And in the final days of the campaign, large numbers of people who were "somewhat angry" decided to put that anger aside and vote Liberal.
Conventional Wisdom: "Scared" NDP voters, concerned about a possible Conservative victory, deserted their party in droves in the final days of the campaign in an outbreak of mass strategic voting.
Survey Says: Fifteen per cent of all voters surveyed at the beginning of the campaign favoured the NDP, and of those, more than three-quarters wound up voting for the party. Liberals and Conservatives each gained 10% of the defectors. There was no massive strategic shift away from the NDP in the last few days of the campaign. Many of those who did leave did so because they preferred Harper or Martin to Layton.
Conventional Wisdom: Anger over the sponsorship scandal caused a substantial vote swing from the Liberals to the Bloc in Quebec.
Survey Says: Not exactly. The study shows that the scandal hurt the Liberals about the same in Quebec as elsewhere. Quebecers went into the campaign with a lower opinion of Martin and the role he played in the scandal than English Canadians, but with a higher opinion of the honesty of his predecessor, Jean Chretien. The major difference between Quebec and English Canada was that in Gilles Duceppe, Quebecers found a non-Liberal leader that they could embrace. This never happened in English Canada. Duceppe's popularity skyrocketed after the French-language debate in 2004 and it stayed high, while the English debate did not provide a long-term boost to the fortunes of either Layton or Harper.
The findings of the 2004 Canada Election Study offers some intriguing hints about the last election and about how things might unfold this time.
- The Liberals' main problem in Quebec may not be anger over the sponsorship scandal, but the popularity of Duceppe. A more popular Liberal leader and a less popular Bloc leader could have resulted in better results for the Liberals, despite the revelations of corruption.
- To the extent that there was a flight from the NDP in the closing days of the campaign, it had less to do with strategic voting than with the inability of Jack Layton to connect with voters beyond the core NDP constituency.
- Negative ads and other "scare" tactics do appear to work, but they don't work
nearly well as the conventional wisdom of the 2004 campaign suggests. The
Liberal attempt to "demonize" Harper was only successful because it
reflected deep-seated uneasiness about the Conservative leader that was present
from the beginning of the campaign. Is that uneasiness still there? The
answer to that question could decide the outcome of this election.
- Always be skeptical of conventional wisdom.