Roshanda and Harvard

I don’t really have any well-formed thoughts on this yet, so I’m just throwing it out there. Maybe someone can help me decide how comparable these two phenomena really are.

I had a job interview last week. During the scheduling of the interview, I was given the name of the person who was going to interview me, as doesn’t seem to be uncommon. Being the curious sort, I found my way to this person’s LinkedIn profile the night before, and I noticed that he had attended the Phillips Exeter Academy. I had heard of this school before, but I didn’t really know anything about it. I suppose I fixated on it, though, because of the awfully highfalutin’ name. So in the spirit of killing time, I went to the above Wiki link, where I learned, among other things, its graduates are known as “Exonians”. “Hmm”, I thought, as I often do when filling the nooks of my brain with useless trivia from Wikipedia.

The next day, this trivia emerged from my mental ether at the interview, and I brought it up as small talk during that brief, slightly awkward period between the time the interviewer comes in and when the two of you “get down to business”. He mentioned that any resume with that school’s name on it always catches his eye. This isn’t surprising; I would take special notice of the name of any one of the unremarkable schools that I attended. However, my interviewer’s alma mater was obviously a different story. A resume bearing its name would’ve garnered additional attention from almost anyone because it’s an elite institution. By sheer coincidence, while I was waiting for my interviewer, I was staring at my phone reading this article, about disappearing work opportunities for all but the most skilled of workers. Hence, sitting in my brain’s cache was all of the talk over the recent years about the emergence of a bifurcated society where all the lower-skilled jobs are outsourced or (ultimately) obviated by automation and the few remaining high-skilled jobs command a wage premium and are dominated by an upper class of high-IQ, super-educated Ivy Leaguers who transmit (through genetics or environment or both) this good fortune onto their children, creating a new kind of perpetual, STEM-trained royal lineage. His remark linked up with these ideas and so lingered a little longer in my brain.

Anyway, the upshot of this meandering tale for my present purposes is that, regardless of political affiliations or beliefs about the above rising upper-class theory, I think very few people would criticize an interviewer, HR, etc. for paying special attention to a resume that claimed a Harvard, MIT, or Caltech pedigree. Even if you think that opportunities to attend such schools are unequally distributed and that all sorts of government interventions to change that are justified, you probably still think that it’s OK to “statistically discriminate” in favor of graduates of those schools. That is, most people condone this kind of discrimination because they believe that, on average, graduates of these elite schools really do perform better.

Thinking about resumes and discrimination reminded me of the now widely-known chapter in Freakonomics about black names. To quickly summarize, take identical resumes, put “Roshanda” on one and “Jane” on the other and send them out, Jane will get more callbacks. Suppose a similar experiment were carried out with college names, taking two otherwise identical resumes and putting “Harvard” on one and “LSU” on the other. Although I could be proved wrong with data, I feel comfortable hypothesizing that the Harvard resume would get more callbacks. With somewhat less confidence, I’d also theorize that the positive effect of “Harvard” would be larger than the negative effect of “Roshanda”.

Now, my guess is that a lot of people who would defend discrimination based on Harvard attendance would decry discrimination based on name ethnicity. Also, I don’t think it would be similarly well-received if HR departments defended themselves by saying that, on average, people with “urban” names performed more poorly.

Do you think I am right that most people would display this apparent contradiction? Is it indeed a contradiction, or is there a good reason why one should be OK and the other not, especially if the first kind of discrimination really is having bad sociological and economic effects?

Update

I re-read a post by Bryan Caplan the other day that asks the more general question that considerations like the above lead to. I originally considered making a similar post, but I thought it might be useful to focus on two scenarios to which the median person is likely to have extremely opposite opinions. The post and comments over at EconLong are worth reading.

Share

7 Responses to “Roshanda and Harvard”


  • Skipbeam Skanderhosen

    Easier to change your name than your alma mater, so I don’t really care too much about Roshanda’s plight. Besides, that’s nothing compared to Skipbeam.

    But I have also given thought to this social bifurcation/permanent underclass/Eloi-Morlocks idea. Like you, I am just putting some thoughts out there:

    If you can go along with the free-market economic premise that there is an unlimited demand for labor at a low enough price, then the issue becomes less about getting people to earn higher wages, and more about allowing for lower wage earners to live “comfortable” lives.

    I’m often in one of two states of mind: Either shock at how amazingly cheap so many awesome things are (a gallon of milk, lumber, LED flashlights, etc), or shock at how insanely expensive some very necessary things are (housing, medical anything, cars). One could live a king in terms of food and gadgets and entertainment, and then be totally bankrupted by a broken bone.

    It’s the government interference in the medical, housing, transportation markets that fucks shit up. Zoning and building codes force people to buy much more expensive housing than they otherwise might. Trying building a plywood shack to live in while you save money for something better. A workable structure would cost a couple thousand dollars. In very few counties in the USA these days would that not be totally verboten. Same with car regulations; consider that a 1990 Mercedes whatever-class would be illegal to make and sell today because it won’t meet government safety requirements. It’s probably worse than that – a 2000 model is probably illegal too. And medical regs are in a different class of fucked-up altogether.

    So, although I’m sure it sounds facile and beyond clichéd to lefties everywhere, I say get rid of all labor, building, medical, and safety regulations would basically make things ok for even, or especially, the ditch diggers among us.

    • Distinguishing between applicants based on merit is different from doing so on the basis of race.

      If Roshanda and Jillian had identical resumes (both got the same SAT scores, both went to Harvard, same college GPA, etc.) but Jillian got more calls back for interviews, presumably the discrepancy has a race-based explanation. And, not unreasonably, people have an issue with discriminating on the basis of race, not on the basis of merit.

      The trouble with parsing these issues is that (1) there are subconscious factors at work when people make these decisions, hence the need for carefully designed studies, (2) people make their decisions on less than complete information, making it more likely that they will be susceptible to their biases, and (3) the empirical studies on such issues need to control for many factors to establish a genuine correlation (e.g. making it clear that the applicants are not legacy admits or affirmative action admits, etc.), and this can be hard to do.

      But bottom line, when one has incomplete information on applicants, if they make their decision based on merit (even if this includes societal trends like race), this does not as problematic, but if in the presence of complete information, a decision is made primarily on the basis of race, that would (to me) be much more problematic.

      • Distinguishing between applicants based on merit is different from doing so on the basis of race.

        Is going to Harvard an issue of merit, though? Especially if all other things are equal? If you say yes because you think, statistically, people who got into Harvard are more likely to be good employees, then how is that different from someone arguing that, statistically, people with “urban” names are less likely to be good employees.

        • Scenario 1: We know nothing at all about X and Y, except that X went to Harvard but Y went to a lesser institution, and one decides to choose applicant X (based on statistics).

          Scenario 2: We know nothing at all about X and Y, except that X is named Jillian but Y is named Roshanda, and one decides to choose applicant X (based on “statistics”).

          What distinguishes Scenario 1 from Scenario 2? If there are statistics to back up the assertion that people with “urban” names are less likely to be good employees, and we didn’t know anything else about X and Y, then there is no distinction between the two.

          But… the scenarios above are so idealized that they bear no resemblance to how such decisions are made in real life. Indeed, the very premise inherent in your question, that when all other factors are held equal, people with “urban” names are less likely to be good employees, has no basis in any legitimate studies that I’m aware of. (I commented in my previous post about how such a study would have to be conducted). To my mind, evocations of idealized scenarios like the ones above are just tools to shield ourselves (usually from ourselves) from the fact that our decisions are largely shaped by our biases. We can tell ourselves that our decisions are based on logic and stats, when in fact logic would suggest that such decisions would never occur in the kind of vacuum described in the scenarios above. Having less than complete information about people is not the same as having no information about people.

          By the way, I’m not disparaging the value of thought experiments, which are a powerful tool in moral philosophy. I’m suggesting caution in the inferences we should derive from them.

          • Indeed, the very premise inherent in your question, that when all other factors are held equal, people with “urban” names are less likely to be good employees, has no basis in any legitimate studies that I’m aware of

            I agree with this. However, isn’t this also the case with college attendance? Maybe I’m wrong, but I can’t imagine there many studies out there that provide evidence of Harvard grads being better employees than Arizona State grads when controlling for all other observables.

    • If you can go along with the free-market economic premise that there is an unlimited demand for labor at a low enough price, then the issue becomes less about getting people to earn higher wages, and more about allowing for lower wage earners to live “comfortable” lives.

      I’m sympathetic. There are a lot of issues tangled up in this statement. One is the old issue of absolute poverty versus relative poverty and how much we should care about the latter (if at all). For example, people like to talk about how in the 50s/60s/70s/whenever people could live a middle class life with only one earner but now two earners are generally required. However, I’d wager that it’s still easy for a single-earner to provide a 1960′s middle class life. It’s made harder by the regulatory issues that you mention (among other things probably), but it still seems possible.

  • Right. So in both scenarios one is drawing a conclusion about a prospective employee based on mythical statistics. But (to my mind) in scenario 1, the misunderstanding of statistics is driven by Ivy League bias. In scenario 2, the misunderstanding of statistics is grounded in racial bias. And people are more forgiving of the former than the latter.
    BTW, your post above on the drug war (a war that I think is completely asinine as well) is the funniest thing since your picture-post of N. Korea’s new “Dear Leader.”

Leave a Reply

Anti-Spam Quiz: