I'll try to answer quite briefly.
The comments I was referring to were made by Glenn Drover, who was representing the Canadian Association of Social Workers, which has done a lot of work on women in the economy recently. He was referring to standard measures that both Statistics Canada and HRSDC produce on poverty. There are LICOs, LIMs, pre- and post-tax--the standard kinds of things--and the market basket measure. All these things use household measures, which means that the power imbalances within households are not reflected. There may be women living in very straitened, almost desperate circumstances in households that actually do have some money and wouldn't fall under the thresholds for any of those indicators.
Now, there are only so many things you can do with any one indicator. Again, it seems to be an area of convergence among people working in indicators that we are not going to find the poverty line. We need several measures, a suite of measures, not a gazillion, but a few key ones, and more than one, that will give us a better understanding.
If you, for example, took key poverty measures, if we picked three--most countries seem to have done something like that--and complemented those with things like the economic gender-equality indicators that show what's going on in the labour force, how the tax system is impacting gender equality, and what time use looks like, then you'd get a much better sense of why and how women always end up featuring more prominently in the poverty statistics.
In that aggregate collection of things there's no one measure that's going to give an answer, but those couple of key things--unpaid work, Suzanne mentioned violence as the other key.... It's not just a matter of disaggregating. It's making a deliberate attempt to build statistics about something that we traditionally didn't do for a long time. Those are the two key areas--the non-market and whatever--and with those, I think we could do a good job.