Let me just go back.
As I mentioned earlier, we cannot do evaluations of programs; we can only look to see if government is applying the management principles that they are expected to apply. If there is an expectation--and it could be simply a policy or even a broad commitment by government--that they are doing gender-based analysis in programs and policies, we can audit to see if that is actually happening. We can also ask them if they are doing the evaluations necessary to know if they are meeting the objectives that are set out.
You mentioned aboriginal. Aboriginal is different because the federal government has a clear responsibility. An example is education: they have a clear responsibility to provide education, so we ask them if they are providing it, how they are providing it, and how they know what's happening. We don't actually go in ourselves to assess the educational outcomes; we ask the departments if they know, and how they are producing it.
We could do the same sort of thing on gender-based issues, but we can't ourselves do the evaluation of programs, and I suspect that is where much of the frustration in the third-party groups is coming from; the outcomes are not what they would hope they would be, but to be a little blunt, that's the case with many programs.