I don't necessarily think that you're going to be hurt. The reality with survey research, and even with the census, is that it changes over time. It's an organic process. You learn as you go along. As far as I can see, the idea of going to a voluntary census, or actually a voluntary sample, carries with it certain risks. I think they've been well identified by the people who are here on the panel. The question is whether or not they're unmanageable risks.
Based on my professional experience doing this research all over the world, I can tell you there are people who manage these risks all the time quite successfully. It may take a couple of cycles, as Mr. Drummond indicated, but it's not impossible to move to a voluntary census and generate very high-quality data that would be of as high a standard as anywhere in the world. As I said before, an 80% response rate is about as high as I've ever seen on anything.