Via Bryan Caplan, Lauren Rivera at Northwestern studied hiring practices at top financial, law, and consulting firms and found some surprises:
[E]valuators drew strong distinctions between top four universities, schools that I term the super-elite, and other types of selective colleges and universities. So-called “public Ivies” such as University of Michigan and Berkeley were not considered elite or even prestigious… In addition to being an indicator of potential intellectual deficits, the decision to go to a lesser known school (because it was typically perceived by evaluators as a “choice”) was often perceived to be evidence of moral failings, such as faulty judgment or a lack of foresight on the part of a student.
I’m not sure what those four schools are, but they exclude some pretty good undergraduates:
You will find it when you go to like career fairs or something and you know someone will show up and say, you know, “Hey, I didn’t go to HBS [Harvard Business School] but, you know, I am an engineer at M.I.T. and I heard about this fair and I wanted to come meet you in New York.” God bless him for the effort but, you know, it’s just not going to work.
And don’t neglect those extracurriculars:
[E]valuators believed that the most attractive and enjoyable coworkers and candidates would be those who had strong extracurricular “passions.” They also believed that involvement in activities outside of the classroom was evidence of superior social skill; they assumed a lack of involvement was a signal of social deficiencies… By contrast, those without significant extracurricular experiences or those who participated in activities that were primarily academically or pre-professionally oriented were perceived to be “boring,” “tools,” “bookworms,” or “nerds” who might turn out to be “corporate drones” if hired.
All this stuff sounds bizarre to people outside the world of corporate recruitment. And it is natural for academics like me to read this and silently congratulate myself on our superior methods of judgment. But surely there are things about our process which would seem just as irrational and counterproductive to people outside of academic mathematics. What are they?
It might make more sense to concentrate on graduate recruitment as against tenure-track hiring, since then both we and the financiers are talking about recent BAs with little track record in the workplace.
(Linguistic note: “Counterproductive” is surely a word that people would deride as horrible managementese if it weren’t already in common use. But it’s a great word!)
(Upcoming blog note: At some point soon I’ll blog about Michael Lewis’s The Big Short, which I just finished, and which is the reason the credentials of financial professionals are on my mind.)
Quomodocumque 
I’ll bite.
The big one is having all the searches on the same time-table. When a department is doing a tenured search and a big name is on the market, it distorts the rest of the job market and makes it much harder for people in the same research area. Also, the use-it-or-lose-it situation when a department has to hire this season because it doesn’t know the next time a search will be authorized.
And then there’s Weil’s law: “first-rate people attract other first-rate people, but second-rate people tend to hire third-raters, and third-rate people hire fifth-raters.” I think Weil gets it a little wrong. The error is one of category and not quality. It is as if a concert violinist fails to join any major city orchestra but then has to cobble together a career by joining a number of regional orchestra. Except those regional orchestras don’t play music but rather engage in a sport in which players hit each other with their violins.
I don’t see this as a problem, but one thing that some other academics regard as extremely odd is how broad our postdoc and tenure-track searches are. In a lot of other disciplines, the job ad is narrow enough that you might only get 20 or 30 applicants instead of the 500+ one gets in a math search. You can make a reasonable case that it’s a lot easier to fairly evaluate the applicants in the narrower search, and the way we do it leads to applying a bunch of coarse metrics before drilling down into the remaining files. The flip side of narrower searches is that any given applicant might only have 3 or 4 jobs that they can even apply for in any given cycle.
At the graduate level, I’m not sure we’re completely immune to strong distinctions in terms of undergraduate institutions in terms of graduate admissions at the very top schools like Harvard and Princeton, just because they admit very few people and so have no incentive to take a risk on someone graduating from a school they’re less familiar with. However, if you drop to the level of, say, the University of Chicago, they have take enough risks on jokers graduating from the likes of Oregon State that the system self-corrects.
Oh, there are plenty of bizarre practices, starting with preeminent status of certain fields. I recall an enlightening conversation with one faculty responsible for hiring at a highly prestigious but small department (under 15 faculty).
He: We must hire somebody in Number Theory. Our [resident expert] just retired. We must replace him.
Me: Why? Who is she/he going to work with? You have such a small department with such an uneven distribution across many areas.
He: Yes, that’s unfortunate. That’s why we need a number theorist.
Me: But why not consider all applicants and choose the best? Why would that be better than hiring somebody in growing fields such as Mathematical Biology, Combinatorics, Geometric Analysis?
He: No, no, no… I tried talking to the faculty. Not even Geometric Representation Theory works. We must have a number theorist. Everyone feels that we would not be taken seriously without one.
Or Iowa State, for that matter :-)
Nathan’s second point was my first thought too. I wonder how representative the “elite” firms are. Are there other financial firms that are playing the role of Chicago/Berkeley/Michigan vis-a-vis the Princeton/Harvard places that they talked to?
I’m sure the financial firms would find it odd that there’s nearly a year between applying for jobs and actually starting to work.
That’s true. You wouldn’t be taken seriously without a strong number theorist. Number theory rules.
I have both gone through and been involved in the hiring process for a financial services firm. I have also gone through the graduate admissions process.
First, the comments quoted above are not consistent with the process at the firm I was at (or the ones my friends were at). In general, the process at financial firms are agile and meritocratic. In fact, I believe that most finance firms do a significantly better job than many (math) graduate programs. Perhaps, my biggest complaint with hiring in finance (that isn’t very present in math) is the frequent use of nepotism.
In comparison, I think many math departments do a mediocre job at Graduate admissions. To start with, the departments make no attempt to talk to the applicants (in person or on the phone)! When I went through the hiring process at a finance firm, I went through 3 rounds, which probably included a dozen separate interviews. When I was admitted to graduate school, the committee relied entirely on my undergraduate transcript and letters from a few faculty members who had little information beyond my scores in their courses.This may be a decent way of comparing me to other students from my school, but I can’t imagine that it is very useful in comparing me to students from other programs.
I believe an interview/problem session with a candidate would provide much more color than a letter from faculty member X who has access to little more than the candidates linear algebra HW and exam scores. Of course, the two are not mutually exclusive: by not doing interviews departments are saying that they would provide NO additional information that isn’t captured in the application materials. In contrast, many firms are using competency and problem solving testing during interviews (DE Shaw is famous for this).
Another big difference is that finance firms engage in active recruiting. They try to go out and find the best students and entice them to apply. A math major friend who one a national prize got a dozen cold calls from finance firm recruiters, but not a single one from a university. Bridgewater recently paid a group of top undergraduates who didn’t apply to their firm $100 to sit in a fancy hotel suite, eat gourmet popcorn and tell them why they didn’t apply. Graduate admission committees tend to make little effort to attract anyone who doesn’t put an application in their mailbox. Even more so, universities do little (at the graduate and junior level) to compete for top talent on compensation and perks. Regardless of promise, everyone admitted gets the same deal (or one of a few tiers).
Of course, a fundamental problem is the incentive structure. When I worked for a finance firm, the people who I were going to work with every day had the ultimate call to hire me. The better I was the easier their lives would be and (potentially) the more money they would make. In contrast, in my current department it seems serving on the graduate admissions (or any facet of administering the graduate program) is considered a chore. Statistically, the chance of any particular person working with a member of the committee is low. Thus there is essentially no (direct) impact on the committee members if I end up being a star or a dud. Yes, the quality of the graduate program has some indirect impact on the committee members, but it probably isn’t a rational computation for a committee member to spend an extra 10 hours on admissions over pursuing his or her research program.
More generally, as you look across finance firms you see a spectrum of hiring and compensation models. Human capital is the largest resource input into finance and as firms compete for the best resources you see all kinds of innovation and creativity. This is healthy.
In contrast, essentially every research math department works on the exact same model.
In some lab sciences they do interview prospective graduate students, perhaps twice as many as they intend to admit. In that case, though, a student is going to be attached to a particular lab, i.e. be an employee of a particular professor, and thus its more natural to have a job type interview.
Personally, having participated in graduate admissions from the faculty side at three universities (Harvard, Caltech, Illinois), I’m a little skeptical of the value of interviews for math PhD programs, in part because there isn’t much evidence that interviews actually work even in the business setting. Also, it turns out that one can actually get a lot out of the letters of recommendation, indeed they’re the part of file that I usually start with since I find them to be the most “information dense” component. Part of this is because one can use certain rule of thumb equivalences, e.g. “best student this year at Uni A” = “best student in 3 years from Uni B” = “best student in 10 years from Uni C”. Also the letters usually help to put the course into a broader context — even though I was head of admissions at Caltech it’s not like I know what Math X at University Y actually entails…
Does the business world have anything like the GRE? I find it patently insane that all the grad schools in the U.S. grant monopoly power to a private company to administer these standardized tests (at great cost to the students) which, seemingly, play 2nd fiddle to recommendation letters anyway. We’re the ivory tower, the pinnacle of civilized knowledge– and we require all our future members to take tests from an unaffiliated private company? lolwut?