Wednesday, September 20, 2006

And Dartmouth Does Not Follow

From today's The Dartmouth, we learn that Dartmouth is unlikely to follow Harvard and Princeton by giving up its binding early decision program. Quoting Dean of Admissions Karl Furstenberg:

"Every time we've [reviewed our admissions policies] in the recent past, we've come to the conclusion that early decision works well for Dartmouth and its students," Dean of Admissions Karl Furstenberg said. "There's no immediate need to change."

Although Furstenberg believes that Princeton's decision may put more pressure on other schools to make a change, he maintains that early admissions programs continue to offer certain benefits.
The article does not list those benefits. More from Furstenberg:

"We have worked hard to diversify the early decision pool with some success in recent years," Furstenberg said. "At the same time, the overall racial and socio-economic diversity of the entering class has increased in recent years."

According to Furstenberg, Dartmouth admits approximately 35 percent of its class early, as opposed to about 50 percent admitted early by Harvard and Princeton. In judiciously managing its early applicant pool, Furstenberg said Dartmouth ensures that incoming classes will represent many backgrounds.

In addition, Furstenberg believes that Harvard's claim regarding the disadvantages of early admissions with regards to financial aid packages is lacking when it comes to schools in the Ivy League, pointing out that Dartmouth offers need-blind admissions and extremely attractive financial aid awards to early applicants.
Here's some speculation as to what a binding early decision program can do. This is my own speculation, and it does not necessarily match how the Dartmouth admissions office views the issue.

Suppose that a university has two groups of applicants, A & B. Group A is stronger academically on average, but Group B has some other characteristics that the university wants to ensure are represented in its incoming class.

If there is a binding early decision program, then the university needs to accept only one applicant from Group B for every space in the incoming class it wants to be filled by a member of Group B, if it admits them at the early deadline. That leaves more admits for the academically more qualified Group A applicants, whether at the regular or the early deadline.

If there is no binding early decision program, then the university needs to admit more than one applicant from Group B for every space in the incoming class it wants to be filled by a member of Group B, since not all of them will come. But given a limit on the total enrollment of each incoming class, these extra admits from Group B necessarily crowd out admits from Group A or pose the risk that an incoming class will be too large for the facilities.

Under this arrangement, a binding early decision program can actually make it possible to raise the average academic credentials of the incoming class. This is not how the early admissions programs at Harvard and Princeton have been discussed this past week, which accounts for much of my skepticism of what's being reported about those programs. It is more reflective of what is reported in the article above.


Anonymous said...

Note that Harvard wouldn't derive the benefit you suggest because its program was early action, not early decision. Without the binding commitment you mention, Harvard still would have to scale up its Group B admits (though possibly by a smaller factor if they're more likely to attend than regular admits).

Kenji said...

An interesting post. I discuss a proposal in response to this analysis at my blog, for anyone interested.

Andrew Samwick said...

As noted in this article, Harvard gets a yield of about 80 percent on its admits, even without the binding early decision. So we don't expect much change at all there. Princeton, with a binding early decision on half its admits, got "only" a 60 percent yield. It will be interesting to see how it plays out there.

Kenji--post a link to your blog.

Anonymous said...

During my time at college and graduate school, I never found the number of "Group B" students to be a constraint on my education. I thought the goal of your admissions department was to yield the best class. The fact is that some of your best candidates are Group B students. A disproportionate number of these students do not apply using your early decision program and this is largely not in the control of your admissions department. Your suggestion seems to work fine if you don't really care which Group B students you admit and you are satisfied admitting lower quality Group A students. I would also suggest that certain Group B students would like to keep their options open and not commit to Dartmouth before hearing from other schools. This option is taken away if you admit all Group B students in the early decision round.