I have done some of the studies on British Columbia. There have been lots of other authors who have done similar ones, and they do try to grapple with the issues you're talking about. They try to say that there are all kinds of things that might affect greenhouse gas emissions or people's decisions in British Columbia, so how are we going to try to attribute any change in greenhouse gas emissions to a policy when we know there are all kinds of other things that could be driving that as well?
Basically, what those studies are doing is either taking data on individual households—such as how much gas a household consumes, or how far they drive—or taking data, for example, on gasoline sold each month in a province, and they're comparing it to what's going on in all the other provinces. The recession that happened in 2009, for example, is something that happened in other provinces as well. They try to control for other things that could be driving these trends, in addition to the carbon price, so changes in demographics or population or prices or other things. They're trying to do this comparison of other provinces, and also control for other factors to come out with these findings.
I should say that it's impossible to say with 100% certainty what the effect of any carbon price or any policy anywhere has been because we don't have two worlds. We don't have one world where this carbon price was applied and an exactly similar world where it wasn't applied. What we're trying to do with statistics is to create those different worlds and tease out the effect of the carbon price from those different data.