ABSTRACT
The Canadian Election Study is a case in point, because circumstances certainly allowed it, but could be said to have forced it, to become the prime site for the study of campaigns. The Canadian Election Study series is among the oldest in the world, dating back with only one gap to 1965. But the Canadian granting agency has resisted creating a presumption in favor of the studies, such that proposals are left to the decentralized initiative of the research community. The 1984 election's campaign dynamics, as captured in commercial polls, loomed particularly large in the decision to shift the Canadian Election Study to a campaign-sensitive mode. Canadian elections seemed ripe for studying media effects and campaign dynamics. Strategic considerations in voters' minds may also be relevant and may also shift. This is especially likely in multiparty elections under an electoral formula that punishes "coordination failure" (Cox 1997), which is precisely the Canadian situation.
