Letter to the Editor: Concerning Statistics, Ethics, and Journalism

On Friday October 13th, The Quest rather aptly published Mr. Declan Bradley's article,  "New York Times Ranks Reed in Bottom 8th Percentile for 'Economic Diversity' Among Selective U.S. Colleges [italicization added]." The article, under the dubiously labeled "Insight" column (the author at one point explicitly states he is expressing an opinion), argues for a reorientation on the part of its readers with regard to the statistical model we use when assessing economic diversity. The article compares two different models for measuring this sociological trend: one offered by the Times College-Access Index (CAI), and the other offered by the Quest, the Simpson Index, usually used to measure the amount of different species in an ecosystem, though here applied to measure the amount of times a Pell Grant recipient will see a non-Pell Grant recipient, and vice versa. Mr. Bradley's article often puts the term "economic diversity" in scare quotes, though only with regards to the Times understanding of the term. Once the article applies its own metric for economic diversity, which it presents as that of the Quest's in its entirety (which I suppose, given that the author is the current editor with the most seniority, is fair enough), these scare quotes go away. While we must apply some doubt to the Times' metric, the Quest is here to bring light to the masses, with all the "news" that’s "fit" to print. 

So one must wonder what is gained by the article's discarding of the Times' model and the introduction of the Simpson Index from the pastures of biological and ecological inquiry. On one end, it appears a rather fruitless endeavor, as the article admits that, if one were trying to save the Reed Institute with either metric, they would only gain a measly percentage point of advantage with the Quest's different model. Considering that the Quest released its own rankings according to this new statistical model, one must instead assume that the point here is to question how we model and define economic diversity, though one of these is a foregone conclusion for Mr. Bradley's article. The article states, "While the percentage of Pell-eligible students at a college is an interesting way to assess that institution’s accessibility to low-income students, it is not, in the opinion of this writer, a compelling measure of 'economic diversity." On its face, this is indeed an interesting premise. Not all working class students are awarded Pell Grants, nor does lack of receipt of that grant mean you can afford, for instance, a $66,710 tuition bill every year. But that is not what Mr. Bradley's article means by economic diversity.  The article interprets this term rather literally, as the implementation of the Simpson Index suggests: under the Quest's model, what is favored is an even distribution of rich and poor, rather than a model which favors low-income students. 

The issue with this literalist interpretation in Mr. Bradley's article is that it eschews any social understanding of the term "diversity." Sure, we may say that a number set with no real world referents is homogenous and non-diverse when it contains only the number 3, though we may be reluctant to say the same when a board of directors contains only women. That is because, in the world of people, statistics become significant according to their relationship with structures of power. Colleges materially benefit from accepting more rich students, as their tuition payments help to increase the operating budget, the endowment, and the college's finances. They do not benefit in the same way from accepting Pell Grant recipients or, if you would excuse me speaking a bit more bluntly, from accepting working class students, which is what this measure is really about. There is no ethical imperative to help the rich gain entry into an institution in the same way that there is to help the working class. Only one of these groups has been routinely denied the opportunity. That is what underlies any social understanding of economic diversity. To treat Pell-Grant and non-Pell Grant recipients as equally important while measuring diversity is not only an insulting misinterpretation of the term, but also fails to say anything meaningful about a college's relationship to systems of wealth and power. It is an interesting math problem, maybe even a fun assignment for a statistics class, though it is not telling us anything about how humans and institutions work. 

According to the Times CAI, the Reed Institute only had 11% of Pell Grant recipients in the freshmen class of 2020-21, while having an endowment breaks down to about $553,000 per student, despite 9 out of the top 10 schools on the listing having a lower endowment per student than Reed. That is the relevant statistic here. That is what we must reckon with. I hope in the future Mr. Bradley and his fellow editors take this into account.