A troubled government can ill-afford to lose a by-election and even a seemingly secure party can be sent into a downward spiral by an unexpected defeat. Damage control can sometimes begin before the voting does, and the refrain that “governments lose by-elections” is often trotted out. But the truth of the matter is that a governing party is no more likely to lose a by-election than one on the opposition benches.
You can read the rest of the article on The Globe and Mail website here.
You can also read my column on why polls, even this far out from the next federal election, still have something to say, in The Hill Times. You need a subscription to read it online, but if you have a subscription you can find the article here.
In light of Ted Opitz's decision to appeal the Ontario Superior Court's voiding of his election in Etobicoke Centre, now is a good time to launch ThreeHundredEight's By-Election Barometer. The Barometer can be accessed by clicking on the image in the right-hand column. This will be constantly updated as new by-elections are called and new polls are released.
By-elections are notoriously hard to predict, and the Barometer is not a projection. But it is a way to track what the regional trends are pointing to for every by-election, as well as any polls that have been released for the ridings in question. The Barometer might also be an interesting measure against which to compare results to expectations. Kevin Lamoureux's win in Winnipeg North in 2010, for example, pointed to his individual appeal as a candidate. Julian Fantino's squeaker of a win in Vaughan and Robert Sopuck's landslide in Dauphin-Swan River-Marquette, on the other hand, were expected.