Thursday, July 7, 2016

When Are You Finally "Represented"?

I am a graduate of the Massachusetts Institute of Technology, with a bachelors degree granted by the Department of Biology in 1973.  My degree still hangs on the wall in my office, at least when it is not packed away as it is now in anticipation of the sale of our house.

Although slightly over 43 years have elapsed since I was handed that certificate, there is certainly a lot I recall about my time at M.I.T., or the "Institute" as we more formally called it -- or just the "'Tute", as we more typically referred to the place.

One of those things I recall is a term used then, called "underrepresented minorities."  The term was employed then by the Department of Admissions and the leadership of the school, to refer to certain ethnicities that were thought to be at risk of failing to be "represented" in proper proportion in entering classes.  I assumed at the time, immature and cubbyhole-lacking as I was, that "proper proportion" meant relative to their percentage of the population of the USA.

I did not ponder at the time that the entering classes at M.I.T., then about 1,000 or so (maybe 1,100 now), were of too small a size to represent the population at large in any meaningful, relevant way.  In other words, if 2% of the USA's population were, say, Native American, we would not be able to look at a given class and know if we were properly balancing the admission of Native Americans.  A thousand people is too small a sample.

I also did not ponder the more mature question, which is this:  If you are trying to represent everyone properly and taking proactive steps to do so, then when are you done?  The question is more elaborately expressed as "At what point have you sufficiently eliminated overt or covert discrimination in admission to the point that you have achieved blind admission and no longer have to take any action, including even collecting statistics?"

Finally, I did not ponder the fact that the point of comparison, the USA population, was a completely erroneous reference.  Not only, for example, was M.I.T.'s entering class size too small to draw any evaluation, but the pool from which an elite science and technical university draws is not the entire population.  In fact, the pool consists virtually entirely of high school seniors, and even then only those in perhaps the upper 5% of the nation's seniors -- and in fact, the pool is only those who actually apply.

So if you were to tell me that:
(A) the upper 5% of the USA's high school seniors were Martian, but over a 10-year period only 2% of the applicants were Martian, and only 1% of those offered admission were Martian, than any Statistics professor from good old Course XVI (Mathematics) could tell you that maybe something was wrong -- but it was that Martians weren't applying.  If, on the other hand,
(B) a representative 5% of the applicants were Martian but only 1% of those offered admission were Martian, we might want to examine if a lot of unqualified Martians were applying,  Finally, if
(C) a representative 5% of the applicants were Martian and 5% of the offers of admission were to Martians, but only 1% of the entering class was Martian, then we might want to examine why they didn't want to go to M.I.T.

Why do I even bring this up?  Because that was 1973.  Let me throw some numbers at you -- these are the members of the class that just entered the school, the Class of 2019 at M.I.T.  When indeed they enrolled last fall, the freshman class was 51% white, 32% Asian, 10% black and 14% Hispanic (the total is over 100%, as some self-identified in multiple categories).  Check out the source of the statistics in this website from the school if you'd like.

Yet they still use the term "underrepresented minorities" even though they're no longer underrepresented!

I don't have to look up the actual number relative to the population of the USA to know that outside of being disproportionately Asian -- including cultures that admittedly stress scholastic success -- the entering class seems pretty much in line with the general population.  If M.I.T. Admissions had been working hard to be blind in evaluating applications, it can pat itself on the back.

And yet ... check out this page from the Admissions site.  You will note that 43 years later, the term "underrepresented minorities" is still in popular use.

That is my frustration with the whole process.  You are familiar with the notion that "if you don't know where you're going, any road is equally good."  It's the corollary here.  M.I.T. Admissions appears never to have decided where it's going, so it doesn't know when it got there.  Any rational person would look at this and feel that if they have achieved a racial profile within a set of reasonable variances over a reasonably long period of successive classes, it has won the battle.

And the rational person would say that they had achieved that end and could stop leaning on the scales.

Unfortunately, by continuing to use terms like "underrepresented minorities" even after decades of documented "success" in not discriminating, Admissions appears to have decided that it will simply continue to operate with an expectation that it will be biased.

Why don't they simply "declare victory" and do a re-check every ten years or so?  How about a press release that says something like this:

"The Office of Admissions at M.I.T. is proud to announce that after 50 years of diligent record-keeping and self-policing to ensure that our admissions offers are made without undue regard to race or religion, we have determined that we have "won the battle."  Admissions has achieved so many consecutive years of admitted classes that adequately reflect the population of American secondary school seniors and those who apply to M.I.T., that we are confident that our practices will, over any reasonable and statistically-significant period, confirm that we operate without bias.

It is important that actions that are implemented to achieve a particular corrective state are suitably discontinued when the remedy has been shown to be effective.  Therefore, M.I.T. will immediately discontinue the analysis of the racial and ethnic makeup of our entering class and will cease reporting on it.  We will continue to collect such data; however, we will only perform decennial, cumulative checks to ensure that our normal, unbiased efforts are continuing to be effective.

M.I.T. will proudly decline any future inquiries for such data with the following cheerful reply: "The Institute no longer provides data which may map our students on racial or ethnic profiles.  We were able to show, over a 50-year effort, that we could create -- and have created -- an admissions policy that produces classes which, over a reasonable period, are comfortably in line with American high-school senior applicants' racial identities.  Having defined our goals and succeeded at them, we no longer need to prove our ability."

Thank you for your applause for our victory."

I really think it's time they did that.  And then every single office in Federal and state governments should require their "diversity" offices to do the same thing.  And if they can't, fire the lot of them, hire a new group, give them a goal and a deadline and then terminate the office when it is achieved.

I'm grinning.

Copyright 2016 by Robert Sutton
Like what you read here?  There's a new post from Bob at www.uberthoughtsUSA.com at 10am Eastern time, every weekday, giving new meaning to "prolific essayist."  Sponsorship and interview inquiries cheerfully welcomed at bsutton@alum.mit.edu or on Twitter at @rmosutton.

1 comment:

  1. You would want to expect better from MIT. They should be the front-runners in making declarations like "We fixed the problem and are dismantling the apparatus." They solve problems out there, right, they should be proud!

    Also think they will be always having their stats skewed by the number of ethnic Asians being way high. They're great students, the Asians and Asian-Americans that go there, so you can't punish them, but that's going to mess up the numbers for the other groups. So just decide that at MIT, 35% Asian is NORMAL and live with it. Good reason to stop counting by race. "Under-represented", my foot.

    ReplyDelete