Great discussion and very well articulated sudden. I agree with the vast majority of your points.
sudden wrote:
Let's be honest though, the schools really do not care about what is good for the candidates -- they are doing what is good for the schools. By minimizing the number of candidates interviewed, a school would maximize its own risk of failing to develop the best class possible. The schools are risk averse, this would never happen.
I agree completely. Schools are interested in minimizing their own risk and the candidates are an afterthought.
sudden wrote:
[Schools] select the best possible group from those available on a demographic adjusted basis as well as maintain a bench of qualified people "in the cue" in case they get stolen away by other schools. This is supported by the fact that Harvard and Stanford interview a smaller number of people because they basically know that almost none of the people given offers will turn them down.
I left out the parts of your quote that I agree with. However, this statement is not born out by the numbers. Below you will find a table I compiled that includes the top 20 schools' interview percentages, acceptance rates, post-interview acceptance rates (acceptance rate/interview percentage), yields, and US News rankings (all rates were pulled from the most recent business week stats). Note that I did not include schools that did not provide their interview percentage (Harvard, Stanford, Chicago, Tuck, Darden) or schools that interview all candidates (Kellogg and Kenan-Flagler). The one exception is Wharton who was included with an estimated interview percentage of 50% (as it was the basis of much of our discussion).
You can see from this table that the post-interview acceptance rate does not trend with yield or ranking as has been implied. You see top schools with high yields like Sloan and Stern on one end and top schools with high yields like Columbia and Wharton on the other end.
Further, the fact that lower ranked schools know that more candidates will turn them down would affect yield, not the percentage of interviewed candidates that are dinged. Unless you are arguing that schools use interviews in a large part to determine which top candidates will not attend (and ding those). If so, interesting premise and one that would be difficult to prove one way or the other.
sudden wrote:
Admittedly, the demographic argument is a bit of a black box, but it makes a lot more sense than your argument, which is basically that the schools are being unfair to people by getting their hopes up, as if the adcom is sitting around saying, "Hm, let's figure out how to crush this guy's hopes... oh, let's invite him for an interview even though we would never admit him." There is absolutely no motivation whatsoever to do this, and there is actually a disincentive in the form of not wanting to waste their own, their students' and their alums' time, not to mention that the last thing the adcom wants to do is create more admin work for itself. The fact that they interview so many people is a biatch, but there you go, that's life.
I have no counter to the argument that it makes no sense for them to interview candidates with no chance of acceptance. They are likely looking to reduce risk and, hence, invite candidates with an almost zero (but not zero) chance of acceptance.
sudden wrote:
Citing Haas, which is a small, quirky program does not prove or disprove anything. One school a trend does not make. Besides, Haas may be like Tuck in that they approach yield by making very sure that they only interview people that are deeply interested in Haas so that they do not become a back up school for people who fail to get into H / S / W. Some of these people will inevitably get into H / S / W anyway, hence the lower yield. I don't know that much about Haas, but this would not surprise me. If you still believe that one school breaks my argument, you need to provide a convincing reason.
Well looking at the table I compiled, schools such as Stern and Sloan (far from small and quirky) have a similar strategy to Haas (and in fact have an even higher post-interview acceptance rate).
I do agree with you that schools like Wharton and Columbia interview such a high percentage due to risk aversion. However, it still seems unlikely that they don't know with 99% certainty that a good number of those have no shot. Especially considering how other top programs are able to be successful while employing such a different (and more applicant-friendly) strategy. For this reason, I'd still like to hear an adcom explanation.