I don't have any numbers for that one. That's probably going to come out of administrative sources as much as anything, if one kept good data on that, but I'm always worried about the administrative sources, because they have very special functions as to what they collect; when I turn to them as a demographer, I often think that it's not that the numbers are wrong, but just that they're very limited and very specialized in their use.
On March 26th, 2007. See this statement in context.