Asian discrimination at Harvard?

Relative to White applicants, Asian applicants to Harvard have higher quantitative academic scores. Yet remarkably a statistically smaller portion of them are offered admission. Is this discrimination?  Better yet, is this a problem?  I am an Asian Harvard alumnus (graduate department of statistics) and another Asian Harvard classmate (physicist Peter Lu) reached out to me on this recent class-action court case.  While there isn’t enough public information to engage on every imperative aspect of this, here are some notes from analyzing the dueling court briefings and model statistics (many thousands of pages).  In a nutshell, the picture is quite multi-dimensional; let’s elaborate on some of the things that stand out.

The plaintiff’s (student applicants) case rests mostly on the idea that the types of candidates within a race are equal, and hence contrasting quantitative averages is an easy exercise.  But there is more to this story, and some of it goes beyond admissions data.  The very best Asians apply to Harvard.  They do exceed other ethnic/race category on quantitative measures.  Is this not to be admired?    

The briefing material asserts Asians appear less diverse as a subset of the applicant pool (e.g., geographically, academic interests, legacy or athletic attributes)  How does one balance getting the best of the best Asian applicants, though a more concentrated set of Asian applicants?  This is just one of many issues to sort through.

Now we have the circumstance (mathematically analyzable) of giving preferential boosts based on personality scores or whether someone was a lineage applicant (i.e., legacy).  Some ask if this is a fair variable to begin with, and whether it should be open to public discourse.  There can also be unintended consequences.  For the personality assessment, this is obvious and open to scoring maltreatment among the admissions committee.  Per the briefing data, it does move the needle in these cases for determining admission outcomes. 

The defendant’s (my Alma Mater Harvard) case suggests that there are too many variables to assess diversity, and in such a way that it is difficult to prove discrimination against any particular applicant.  For example, if we look at near-perfect test scores as the only admission’s criteria, then you could populate each class size, many times over.  So difficult choices need to be made among many diversity factors.  Not only on admission, but also on recruitment. 

The worst type of diversity is the one where there is no empathy to discuss it.

They can point to matters in their favor, such as the large fraction of applicants who choose not to provide their race/ethnic data, or the many random differences they find in the concentrations (majors) and other national population data, split among race/ethnicities.  In court they will likely defend their personality score, including adjustments for athletes and legacy students.  Good luck with that...  

Personality is important.  These are also 17-year olds.  How can anyone really know what they are supposed to look for, when scoring this?  They can't.  This case center on whether personality is filling gaps in the admissions process, and if there are downsides. Academic performance and extracurriculars also matter, so it will be of interest to see the public discourse over what new information the personality variable provides. 

So yes there are many moving parts and Harvard will have a chance to explain this in court.  This might allow for a fresh perspective to be recognized, from all parties.  At best, they will look absurd.