Generally speaking, the biggest hurdle is data availability. When we're doing measures, we try to reach all different sources. I'm a tax policy person, so we actually have better access to data in that we have the tax returns, and that's divided up by gender. We know filers are male or female. We know what their income is. We have all that available to us.
On other measures, it can be difficult just to actually get the underlying data as to how many women would benefit from a certain measure. We go to different sources in trying to analyze that. We use StatsCan data quite a bit when the measures deviate from the tax system, etc. But getting good data is one of the hardest aspects.