[swimfar]

And I'm sure some of that is because the children of Harvard graduates tend to be more qualified (in terms of other application criteria) than the general population or applicant pool.

If an additional 1000 2.0 students applied to Harvard next year this number would look even worse. But would the change be meaningful?

That's not to say they don't take legacy into account. But, as presented, those numbers can be misleading.

[JoshTriplett]

Some of it also likely occurs because the children of Harvard graduates will have a better idea of what the process is looking for, not least of which because they have an N=1 daa point of a successful application.

Similarly, if you're trying to have a successful interview with a company, it helps to talk to people who have interviewed there before, preferably people who have interviewed there successfully.

(The process may well have other issues that make "legacy" applications more likely, but a higher acceptance rate of such applications doesn't inherently suggest a problem.)

[michaelt]

If children of graduates are more qualified _naturally_ then why does Harvard _need_ it to be _policy_ that they take legacy status into account? MIT and Caltech seem to do fine with legacy-blind admissions.

[arethuza]

I'm not from the US so don't really know what I'm talking about - but my impressions from this side of the pond are that MIT and Caltech seem to have rather different reputations from Harvard?

[albntomat0]

Probably proportionally based for each one.

Legacy applicants are X% more qualified naturally (well educated parents, etc), and get Y% from the admissions. Given the numbers for Harvard, MIT, etc, there are plenty of qualified applicants for each spot.

[usepgp]

I would imagine that both MIT and Caltech have a higher acceptance rate for legacy for this reason. However a side effect of them not keeping track is that we will never have the breakout of those rates.

[fjsolwmv]

That's backwards reasoning.

[montrose]

Than the general population possibly, but not than the applicant pool, which is the comparison that matters.

It seems more likely that in the applicant pool the non-legacy kids would be smarter. Imagine how confident of your ability you'd have to be to apply to Harvard as a random kid from a public high school in Iowa. Whereas if you'd grown up expecting to go there because your parents did, the threshold for applying would be pretty low.