I think it's done jurisdictionally.
In Quebec, for instance, there is a venue in which the provincial government sits down at the same table with the faculties of medicine and with hospitals, and they look at expected vacancies. They actually look at the number of surgeons expected to retire at the hospitals in the jurisdiction, and they allocate their residency quota accordingly.
Ontario has quite an elaborate model. They've tried to look at the data in a more robust way, weighing it against expected changes in the population, and again that table exists at which the provincial government sits down with the faculties of medicine and they make decisions about what the quota will look like.
So, again, I would repeat the point about the system-wide approach to this. There are more residency positions in geriatric medicine than are filled by residents. Graduating medical students are looking at that specialty and they're making decisions based on whether they would be doing a lot of on-call, whether they are going to be paid in a way that they want to be, whether they will have opportunities to teach and do research. I think there's a bit of a package that has to be looked at.