The Bryan decision is still bugging me. I find the idea that a priori "logic and reason" should be preferred to actual empirical evidence difficult to take in the first place. But even a priori logic points in the opposite direction the Court took.
Suppose that the act of voting is a cost we engage in to get a government we want. Whether we live in Nanaimo or Bonavista, the chance that our individual vote is going to affect who becomes the local MP, let alone who constitutes the government is essentially zero. So on this assumption, no rational person would vote.
This argument leads to the expressive theory of voting. Instead of conceiving of voting as something we do in order to get a result, we vote to express our political identity (or perhaps as a civic duty). In economic terms, voting is consumption, not investment. It follows then that we would vote the same way regardless of who was going to win.
This argument may be wrong, since it is premised on people being instrumentally rational. Maybe they aren't. But I can't think of any way to reason about how people will react to information that is based neither on evidence of how they in fact do so or on an assumption of rationality. If Bastarache does, he should publish.