On Tomorrow's Byelections

Four byelections will be contest tomorrow: one in suburban Vancouver, another near Riviere-de-Loup, a third in Montreal, and a fourth in the Nova Scotia riding of Conservative-turned-independent Bill Casey.

This article by Joan Bryden tips the NDP to win one, the Bloc to win two, and the Cons to regain the NS riding. But it also goes on to outline how party strategists are claiming that these elections are a test of the health of official opposition. Is such a claim credible? Do by-elections act as a test of the popularity of the government and other parties?

Fred Bastien and I have an article forthcoming in the Canadian journal which tests this proposition, among others. You can read it here. We find that while the success of government parties in byelections was linked to the government's national popularity (as measured in national vote intentions) until 1993, there is little evidence that it is anymore. We don't have a strong explanation for this, except that the period since 1993 was characterized by a very fragmented opposition, strong regional variation in support for parties, and the lack of a credible national alternative. Accordingly, we are inclined to think that these elections were more likely to turn on local issues. Now, much has changed in our federal politics in the last five years, so this may no longer be the case. But for the meantime, view with caution any self-serving explanations that these byelections provide some measure of the vitality of parties nationally.