Canada held snap elections yesterday in which Steven Harper’s Conservative Party managed to secure a parliamentary majority.  The New Democratic Party (NDP) essentially supplanted the Liberals as the left opposition in the country, while the center-left Liberals, who had dominated Canadian life for decades, saw their seat total plummet.  Likewise, the separatist Bloq Quebecois were reduced to a small four seats.

Canada isn’t normally thought to be an interesting country (Something which they should take as a complement; interesting countries tend to have more problems!) but buried in the lackluster coverage are a few gems worth election wonks pondering over.  The first is the fact that the final results were far different than initial polling would suggest.  When snap elections were called in March, the Liberals were still the leading opposition party with the NDP solidly behind.  This is pretty strong evidence that the campaign mattered. Not shocking to many, but certainly to those who are aware of the significant  literature that suggests campaigns are really only important at the margins. Recall the recent British elections where the Liberal Democrats under Nick Clegg’s leadership surged in the polls, only to wind up right where they started when the results were tallied.   The LibDem’s performance, of course, may be partially attributable to tactical voting, which brings me to my next point regarding Duverger’s Law.

In Les Partis Politique, Marice Duverger explained how plurality votes in single-member districts would bring the effective number of competitive candidates to two. ‘Duverger’s Law’ as it was dubbed by William Riker  was taken by some to mean that (1) SMD systems would always produce only two viable parties, and (2) that there would only be two effective parties at the national level.  These two misconceptions have led many to incorrectly state that Canada and India are proof Duverger’s law has been broken.   The problem, I think, is that what Duverger was explaining wasn’t really a law so much as a force, and is in this respect completely true.   Gary Cox in Making Votes Count does a good job of rescuing Duverger while expanding on his theory with his ever helpful equation, N+1, to predict an electoral system’s impact on the number of candidates.  (N being the number of available seats in the district, the number of candidates would be one more).   A quick glance at the results seem to indicate that Duverger’s Force was certainly in effect.  I’m guessing once the NDP took the mantle of the leading non-conservative party, voters evaluated it as their best option in a single member district.  It’s hard to say if Canada’s party system will stay like this after the next election, but I think there is a decent amount here for us to digest for now.