So what’s new with the polls?
There was an interesting debate about the validity of public domain polls since we last looked at them in this space. I don’t conclude from that discussion that public domain polls contribute nothing to our politics. What I do think is that to make sense of all of this information, you need to look at all of it. It is the trends that usually matter, not any particular data point at a moment in time. This is doubly true for regional numbers and socio-demographics, which swing wildly in most public domain polls given small sample sizes.
In my view Alice Funke has written the definitive article on the value of formula-driven seat projections untempered by experience with riding-by-riding ground campaigning. Which is, basically, that they have almost no value at all.
Here are some of the most recent public domain polls in, let’s say, reverse alphabetical order.
Nanos Research (released Feb. 21):
Conservatives 39.7% – Liberals 26.6% – New Democrats 18.9%
EKOS Research (Feb. 23):
Conservatives 32.4% – Liberals 27.3% – New Democrats 14.8%
Angus Reid (Feb. 25):
Conservatives 39% – Liberals 26% – New Democrats 18%
Abacus Data (Feb. 25):
Conservatives 38% – Liberals 23% – New Democrats 19%
There are many ways to look at trends. For this exercise, let’s look at the most recent chart released by Nanos Research, which is sending pretty much the same messages you can find in long-term charts by all of the polling companies.
What to make of it?
As has been widely remarked, the Conservatives seem to go up whenever the Liberals again announce that it is now convenient for themselves to make Canadians vote in a federal election. But then Canadians seem to rebound away from the Conservatives as soon as the media reports that a Conservative majority is a serious prospect.
The Conservative Party of Stephen Harper is still in range of what Stockwell Day and Joe Clark got between them in the 2000 federal election (Mr. Day and his Canadian Alliance received 25.49% of the vote in that election; Mr. Clark and his Progressive Conservatives received 12.19% – a total of 37.68%). But the long-term trend in the Nanos numbers tell us they have crept up into that range and may now be a little ahead of it, if their current numbers hold.
It seems reasonable to speculate that they are achieving this by slowly leeching “blue Liberals” and former Progressive Conservatives from the Liberal column. Another way to think of these voters is what we might call Rob Ford suburbanites, socially-conservative new Canadian communities, and what in the United States are called “Reagan democrats” – working class voters, open to Conservative messaging on issues calculated to anger them and therefore to persuade them to vote against their own economic interests.
It also seems correct to say that the slowly creeping Conservative vote tilts male, and toward an older demographic. The bad news for the Conservatives is that this demographic needs to be fed periodic pieces of red meat, which repels other voters. The good news is that these voters have disposable income they are prepared to throw at issues that they are angry about (thus the Conservative fundraising advantage). And they can be relied on turn out to vote on election day (unlike, for example, youth voters, who traditionally tend not to vote).
The New Democrats have also been enjoying an incrementally good run, as reported in the Nanos chart. It’s a nice change after their near-death experience in the 1990s, when New Democrat support often hovered around 6%. The centerpiece of that collapse was in Ontario – due, I’d argue, to what we might politely call provincial political factors, plus a notable appetite to be done with the Mulroney government (Jean Chrétien was also assisted to three Ontario sweeps by the Canadian Alliance/Progressive Conservative split).
