Why Harper wants an election he says he can’t win outright

Unfortunately for a person of my political biases Stephen J. Harper is one smart cookie. The man runs a hell of a tight ship and is a brilliant electoral strategist to boot. While many politicians, to their detriment, tend to focus on short term gains, he always has a multi-year scheme up his sleeve. Case in point: in 2004 I read that Harper spent a month touring small-town Quebec, holding town-hall meetings to which only local and “ethnic” press were invited, assuring all that not only was he a family-values kind of guy, but one whose vision of a small federal government left lots of room for Quebec to handle its affairs internally from a position of near sovereignty. So I was among the minority who were not at all surprised to see significant gains for the Conservatives in Quebec in 2006.

As the Star noted this morning, an even longer-term plan of Harper’s has been to unseat the Liberals from their throne, and change the conventional wisdom which describes them as the “natural” governing party of Canada (which position they hold mostly by virtue of moving precisely to the centre of opinion polls on each and every issue, a strategy which has served them well over the years). The Liberal party is broke and spineless, having avoided voting against the Conservatives in every single confidence vote since this Parliament convened, because they know they would lose seats in the ensuing election. That hasn’t changed, although the Conservatives are still polling nationally in minority government territory. Harper’s disingenuous reassurance of this (“Don’t worry, even if we scare you, vote for us anyways — it’s not like we’ll win a majority”) is more subtle than it may seem. The Conservatives have just pulled ahead of the Bloc in Quebec, and if Harper can sweep Quebec like he does the West, he can easily win a majority of seats while gaining only a minority of the votes. This is far from certain, but it’s plausible enough to tempt him, and either way he ought to take more seats than in 2006. Even better, he’s describing this election as a referendum on Dion’s carbon tax proposal, which for my money is great plan (if a hard sell) and the first original idea the Liberals have had in a long time. If Harper can decisively defeat this scheme in a national election, look for the corporate money that keeps the Liberal party alive to jump the shark and move to the Conservatives, leaving the Liberals weak, leaderless, and likely to dissolve into yet another round of infighting with a new leadership race. As a group of people with nothing in common but a desire for power, the Liberal party stands to lose their armies of smart, ambitious backroom worker bees just in it for the perks, potentially leaving them foundering for years to come.

Less widely noted in the media is the fact that the U.S. is sliding into recession, meaning Canada will follow this fall, and in recessions the party in power is inevitably (if unfairly) unpopular. So if Harper were to wait until fall 2009 for an election (as he initially promised), he would have to contend with a year of bad reviews — today’s polls are as good as they’ll be for a while. The smart thing to do really is to hold the vote now, gaming the system by picking the election date most favourable to his party, for exactly the reasons that Harper had promised to abolish the flexible election date system. And he can, because he cleverly left a loophole in the law which he passed to ban these games which allows him to do this with a minority government. Like I said: one smart cookie.

Assuming the writ drops and an election is called next week as currently seems likely, here are two key websites to watch, for those who don’t already know them:

  • The Laurier Institute for the Study of Public Opinion and Policy, featuring poll analysis by Dr. Barry Kay, whose research interest is mathematical models for predicting Westminster-style election results from national poll data. This of course is a much harder problem in Canada than, say, in the U.S. presidential election — state-by-state polls translate directly into predictions of U.S. electoral college votes, but here you would need riding-by-riding data for equally dependable forecasts, and only Elections Canada has the money to do that. (Papers detailing his models are available from that site.)
  • The Election Prediction Project by election geek Milton Chan, using committee-reviewed crowdsourcing to call each riding individually, with equally impressive results.

Some people follow the Olympics, I follow politics. Happy election watching, everybody.