assigning PhD advisors

[pgher]

I'm in engineering. MS students will be accepted, unfunded, by the department. Then it's up to the student to find an advisor. Or alternatively, a faculty member can proactively make an offer.

PhD students MUST have an advisor or they will not be accepted. The advisor is expected to provide the funding, unless the student has a fellowship. Again, faculty can proactively recruit, and indeed many (most?) of our PhD students come from active recruiting.

[Dismal]

We are developing a new system where a committee determines a medium size group of applicants who are acceptable and then shops them around to possible advisors who have funding. Then there are a couple of top students whose potential advisors don't have funding and so pool money is used for them. Then advisors who are willing get those students for free. Our advisors are assigned from the beginning but students can switch later.
In my previous department the committee would just make offers to the top students and then try to get those with funding interested in them later. That didn't work well IMO - when I have funding I would rather work with a top 20 student with interests in my area than a top 10 student who had little interest in me other than $$.