At the end of our discussion on Barnard’s proposal for an academic index for NESCAC, a current eph provides a thoughtful set of questions about the qualifications and performance of international students at Williams. Let me highlight some of his questions, make some comments and invite comments from others.

1) There is a great senior thesis to be written on this topic, advised by someone like Morty or Gordon Winston. If you’re an economics major interested in writing a thesis (especially someone considering graduate school), I urge you to consider it. As current eph notes:

If we’re trying to measure “performance” at Williams, I believe PBK is a rather poor way of doing it. We should really be looking at the entire grade distribution, and comparing like-students to like-students. In other words, we should be comparing internationals with non-URM, non-tipped, non-legacy USA students…which if I’m not wrong make up less than half of the class. Based on the data we have, we should consequently expect to see International students relatively evenly distributed among the top 40-50% of the class, falling a little more heavily at the extremes (because as previously discussed, d3 majors tend to get grades more towards the extremes). From an International class NOT affected by admissions bias, we should expect to see international students ranging from slightly below average, to PBK, with a slightly higher than average proportion falling under PBK (15-20%, rather than the 12% otherwise expected). I would bet that if you look at a 5-year average of grade distributions at Williams, you’d find results very similar to this.

This is an excellent point. Unfortunately, there is no public information that I am aware of on the topic, although every professor that I have heard from sings the praises of international students in the classroom.

A thorough study would need to be careful of what the fair comparison group is (all US students or, as current eph argues, all non-URM, non-tip US students) and adjust for tendencies to major in different fields, especially when some fields like Division 3 have higher variance in their grade distributions (therefore generating more PBK’s even if the average grades are the same). Note also that major choice depends on qualifications (how many AR 4s and below at Williams become physics majors?) so there is some endogeneity here. I do not think that the legacy issue matters much since legacies receive much smaller advantages than tips and URMs.

2) Although academic performance at Williams has not been studied closes, as far as I know, we have good evidence, I think, on the extent of bias in the admissions process. Recall what Jim Kolesar wrote:

But a college that gave itself over to educating mainly international students, which is eventually what would happen given the numbers, would have a significantly different mission, very different standing with U.S. prospective students, and greatly altered relationship with government, donors, etc.

I read this to mean: “If Williams did not have an international quota today of 6% of the class, an unbiased admission system would produce a Williams class of at least 30% international students.” This is consistent with second-hand reports I have heard about the large numbers and high quality of applicants from, especially, China and Eastern Europe.

Could I be reading this wrong? Sure! Perhaps Jim means that “Given the numbers of international applicants that would come in the future, Williams would become majority international 20 years from now.”

My main claim is that I have never heard anyone with knowledge of the admissions process deny that the current international quota of 6% at Williams is binding, that, in the absence of a rigid quota, Williams wouldn’t be much more international (10%, like Harvard/Yale? 15%? 30%?). But maybe I am wrong. Has anyone heard differently?

3) If it is true, that the international quota is binding, that a blind process would lead to 15% or whatever international, then we can be fairly certain that the admissions department took the best portion of that pool. That is, I would expect international students at Williams to have an average academic rank very close to 1. Since AR is an excellent predictor of academic success at Williams, it stands to reason that any group with AR 1 will do better at Williams than any other group with a lower average AR.

But, again, the comparison group is key, as current eph points out. If you get rid of tips, URMs and legacies, it is not clear that average AR for the remaining students is that far from 1 either. In fact, another way to look at it is that there is a quota for non-tip, non-URM and non-legacy (TUL) students at Williams. Indeed, there are only 250 or so spots open once those groups are taken care of. Is the cut-off for the top 250 US students that different than that for the top 30 international. Probably, but maybe not. Maybe the international quota isn’t really that binding because international applicants don’t compete against TUL applicants at all. International applicants only really complete against non-TUL applicants and, in that competition, the quota doesn’t matter much.

I appreciate current eph taking the time to outline these views. Informed comments are welcome.

4) I’ll leave aside for now the many comments on how one might measure “performance” at Williams. I think that everyone (current eph and Graham especially) has a useful perspective on this. I have argued elsewhere that the College, to the extent that it cares about such things, should do a better job of measuring them and then trying to tilt admissions toward them. I think that this is very hard to do (outside of spots and music), so I am suspicious of moves in that direction. In all of the above, I am just talking about academic performance which is only one aspect of one’s time at Williams.

Print  •  Email