I haven’t written about polls much here on the blog during this campaign. I take polls as significant but transient: they give useful information about the state of play today, but of course the state of play can change. Trends need time to change, however, so the closer you get to election day, the less time trends have to change. You can load a .pdf of all that here.
• The Conservatives have polled above their 2008 election-day share of the national popular vote on every day of this campaign. The stability of the Conservative vote is simply extraordinary. It went from roughly 29% to 36% over the Christmas holidays in 2005-06, notched up a couple of points in 2008, and is now a couple of points higher. That growth is very slow but as a rule of thumb, it doesn’t reverse.
• Regional numbers have bounced around more than that, but not really all that much. The Conservatives have been trending down in Quebec, where they don’t have all that much to lose; arguably trending up in Atlantic Canada, where they haven’t got loads to gain; and recovering from a mid-campaign slump in Ontario. The Liberals trended up fairly steadily in Ontario for the first week, held steady for a while, and have been declining since April 11, the day of the English-language leaders’ debate.
• “Slump” is relative: the Conservatives have polled below their 2008 election-day share of the Ontario popular vote precisely once, on April 13. (Yeah, yeah, it’s a tracking poll, so that’s the day the slump of the previous couple of days was reported. Yeesh. Sticklers.) They’ve been above 2008 levels in the most seat-rich province for the rest of the campaign.
• The Conservatives are polling 9 points higher than their 2008 score in Ontario, the Liberals 5 points lower than their own 2008 level, and the NDP a point lower than theirs. Even with a few losses in Quebec, that will surely be enough to put Harper within a few seats of his majority. My hunch is it puts him over the top.
• Oh-ho! But if he misses his majority, don’t opposition parties have a shot at taking him out in a Parliamentary confidence play? Only if they can present a coherent alternative. Jack Layton will have a really good Tuesday night if he wins, say, 70 seats, but he’ll have a hard time presenting for prime minister. And the more he grows, the more he’ll want a say in the composition of any Liberal-led coalition-collaboration-thingie. Opposition deadlock, to Harper’s advantage, seems likely even if Harper falls just short of a majority.
• But here’s what’s most striking. What’s the biggest obstacle to a really transformative NDP breakthrough? It’s Ontario. NDP support there is 4 points lower than in the Atlantic, 7 points lower than in the Prairies, 10 points lower than in British Columbia, 13 points lower than in Quebec. This is looking less like a Quebec-led NDP wave than like a national wave from which Ontarians are opting out. Ontario is the only region of the country where NDP support is no higher than on the day of the English-language debate.
• Why? My strong suspicion is that Bob Rae’s 1990-95 government has left a lingering bad taste in the mouths of many Ontarians who might otherwise consider an alternative. When I test-flew this hunch on Twitter this morning, some people wondered why Ontarians would be generically mad at New Democrats after Rae, and not at Conservatives after Mike Harris. Well. Leave aside any subjective evaluation of the quality of the two provincial governments, because that would just cause pointless shouting matches (No, your guy sucked worse). Look at the numbers. Bob Rae’s NDP lost nearly 18 points of popular vote to lose with 20.6% of the vote in 1995. Ernie Eves’ post-Harris took an 11-point reduction to lose with 34.6% of the popular vote in 2003. That is a much less severe rebuke from voters. Since their respective losses, the provincial Conservatives have more or less stabilized, while the provincial NDP is, I’m sorry, a basket case. I haven’t heard anecdotes from the door in this election, but in other recent federal elections, federal New Democrats were still getting their ears bent by voters about Rae Days. The man may have moved on; the brand is still in difficulty.
• Still, my explanation for the NDP’s much weaker showing in Ontario can only be conjecture. The facts are stark. The Harper Conservatives are in good position to improve their showing in this election. I’d appreciate a fact-check on this, but I believe that in so doing he would become the first prime minister since Confederation to increase his seat standing three elections in a row. (UPDATE: good thing I asked. Colleague Coyne reminds me that Wilfrid Laurier went from 118 seats in 1896, to 132 in 1900, to 139 in 1904. The Liberals grew from 1963 to 1965 to 1968, but they had to change leaders to do it.) If he does not win a majority he will probably face a divided opposition poorly positioned to take power from him through parliamentary manoeuvring.
This can all change. But the most important feature – steady support for the Conservatives – has now held for four-fifths of this campaign.
UPDATE, Monday evening: Ekos has numbers so different from Nanos‘ that you can argue all night over who’s right!