Thus far, we have not. We've talked about the need to do that. It's a much more complex request of StatsCan, but we're looking forward to doing more of that analysis. I think we started the study at a fairly basic level in terms of intersection, which is race and gender. With the scope, there is a necessity for us to do that.
The argument we're making is that the more disaggregated data is available, the greater is our ability to do the analysis you are asking us to do.