We will know that in this one as well, quite easily, because we'll have the data from the census part of the exercise. We will know exactly who was supposed to respond in the NHS, and where they lived, and we'll have basic information about them.
Response rates vary across population groups and across the country. We had a tremendous problem, for example, in northern Alberta in 2006. That's par for the course. The issue is whether the data are usable, and that's—