2 points, based on personal experience: One, particularly in PhD programs, faculty must ultimately associate itself somehow with the student. And not as an inst

2 points, based on personal experience: One, particularly in PhD programs, faculty must ultimately associate itself somehow with the student. And not as an institution but as an individual professor, as supervisor or co-supervisor of the thesis. The places are fewer. The demand on the time of the professor is higher, per head. So there is virtually no room for "legacy" or "bring home the doughnuts" type of admissions. Think about it: the professor must ultimately sign his name on that kid's thesis or, even more publicly, publish together. Perhaps some professor profits individually and then makes a strong case for such a student because he figures his reputation is worth X amount or because he thinks the student is OK anyway, but institutionally I'd say no. Two, in master's programs, the likelihood of such admissions is higher. And within master's programs, I'd say MBA has a pretty high incidence. But don't forget, in particular in this setting, MBAs are not there for "education" in terms of books or skills or intellect. It's there for making connections. And so, legacy "kids" are kind of legitimate because they bring, almost by definition, already a very valuable asset with them.