There is no evidence of the kind you ask, mainly because we have very few examples of recent changes of the kind you describe. Since most countries use fixed elections, to isolate fixed versus unfixed and look at the different voting turnouts would not work statistically.
There are two points I'd like to make on this. I have argued that with fixed election dates--and Mr. Dewar mentioned this--as part of a series of changes to address the democratic deficit, then I think it would be very clear because we would end up being similar to countries with higher turnouts, like the country I'm in right now. But that would not just be fixed election dates; it would include, for example, more proportional elections. If you combine those and a number of other specific matters, it's pretty clear that you would more likely improve turnout.
The second point--and this is the point I make in the article that you cite--is that what having fixed election dates does is make it easier to address particular groups that we have found are likely to be absent from voting. It makes it easier to address them and say, look, let's develop a strategy so that in the next election we can mobilize resources towards those people. If these are young people, for example, the resources to be mobilized would largely be through the schools, through planning civic education courses. If it's immigrant-based populations, it might be other means. If it's attracting women candidates, for example, I think fixed election dates would help in that regard, for the reasons I've explained and that have been mentioned in this committee.
In the very specific measures to improve participation, having fixed election dates is a good tool to facilitate that, but those are very specific, in that it's in the context of those kinds of very specific measures that fixed election dates would affect turnout. I'm not prepared to simply say that if this law is passed more Canadians will vote. There's no data to allow us to predict that.
Thank you.