How Much Weight Should a Candidate’s School Have?
“The next time you do a job search break your committee into two groups. Have one group evaluate the candidates without reference to the institution from which they graduated and have the other evaluate the candidates with all of the institutional information included. I can almost guarantee that the short lists will not be the same. And I believe that anyone who is honest with him or herself about how this process works will agree. If you think I’m wrong, try it. Or at least try it as a thought experiment.”
Mitchell Aboulafia (Manhattan College) suggests that an emphasis on ranking schools leads to us overemphasizing the importance of candidates’ schools in making hiring decisions, at his blog, Up@Night.
Good topic. Thanks.
While I completely believe this is true, I think this thought experiment is pretty general. Pick any attribute that you want from an application package, and you’ll get the same story. I highly doubt that applicant rankings are stable against these kinds of perturbations. (They’re probably not stable across possible compositions of search committee for this reason.) Imagining that search committees can do one thing that will strip out bias and then the process produces a nice, stable, objective ranking is silly. There isn’t going to be one. We just have to accept that this is in part a stochastic process with different sorts of biases that creep in at different points. The best that departments can do is seek to have different biases than other departments, and maybe if they’re disciplined about it, make sure that search committee members each have different biases than each other. So, maybe randomize the order that you read different parts of the application, but leave that ordering stable for each search committee member.
No matter what, plenty of deserving people will get jobs, and plenty of deserving people won’t get jobs. Let’s keep that in mind, and not pretend that the search process can ideally seek out the ‘best’ candidate.
There was some interesting discussion of your title question at Certain Doubts a few years back:
http://certaindoubts.com/the-evidential-value-of-pedigree/
and
http://certaindoubts.com/the-evidential-value-of-pedigree-ii/
As I commented on the second post there, institutional pedigree can serve as a kind of “meta-evidence” that there are [perhaps yet-to-be-manifested] first-order reasons to favour the candidate from the more selective school. So the more completely informed you are about the first-order reasons, the less room there is for such “meta-evidence” to be indicative. Conversely, the less info you have (as when you’re earlier in the search process), the weightier this indication will — quite reasonably — be.
Also posted at Aboulafia’s blog:
Divide the search committee into two groups and give them exactly the same information. I bet they will still not come up with the same short list. There will be overlap but also quite a bit of difference. But if that is the case I don’t think your experiment will show much on its own.
Shouldn’t pedigree, all else being equal, count *against* a candidate?
Suppose two candidates have essentially indistinguishable publication records, letters of recommendation, etc., but one comes from a noticeably weaker graduate program than the other (by whatever criteria you use to measure such things). Would not the one who achieved this from a weaker program, and therefore presumably having gotten a lower quality of training, have to be the *more* talented of the two?
There was a thread on newapps a few years ago on exactly this point. I’m a bit too lazy to look it up, but I think it was posted by Lance or Schliesser.
Here’s something very easy for SCs to do. At the very least, read writing samples anonymously. Other fields have anonymized hiring practices. This seems a simple, small step we can take that would allow us to be sure that we are reading samples accurately. It’s hard to anonymize a CV, since the pedigree appears in other ways (who one has worked with, etc.) But anonymizing other portions would be a help.
I agree with Jeff Heikkinen that, everything being equal, pedigree should count *against* a candidate when measuring their philosophical ability. However, I want to suggest a modest proposal that will, in our status obsessed profession, probably sound only slightly less outlandish than eating babies.
In the case of shortlist-worthy candidates, where we are likely judging among candidates of really great and fantastic quality and all other things are equal, we should give more weight to their moral deservingness than their philosophical ability. For three reasons: 1) we don’t harm our departments by hiring the least great of a superb bunch and so aren’t morally *obligated* to hire the very top candidate, 2) most of us aren’t top philosophers at top departments and so we don’t necessarily have any moral or meritocratic *right* to the top hire, and 3) judgments of comparative quality among near equals are so unreliable that we will have a statistically equal chance of getting the top hire anyway.
So, in a brutal, unethical game that consists largely of moral luck, where the playing field is not only uneven but pretty much divided into separate playing fields with some players never let out of the box, we have a moral obligation to let moral desert override criteria of quality whenever we can. I think this would include, in some cases, favoring the student from the weaker department, because they have have often had to overcome greater obstacles to be in the running. (Of course, such judgments of moral merit are difficult and not always reliable, but so are hiring judgments about philosophical ability.)
I have argued repeatedly that pedigree should never be a factor in ranking job candidates since pedigree-snobbery rewards the well-pedigreed twice over. First they get an excellent education which enhances their opportunities to produce good work and then they get rewarded all over again for having had those opportunities.
Suppose we have two candidates Sophie X with a PhD from Boondock U and Yolanda Y with a PhD from NYU.
Case 1) They are both two years out from their PhDs each with five publications in top twenty journals plus good teaching evaluations. Is it rational to choose Yolanda over Sophie on account of her pedigree? Obviously not since Sophie has managed to do as well as Yolanda without having had her opportunities and is probably the smarter of the two. So in this case pedigree should not be a factor.
Case 2) They are both two years out from their PhDs with good teaching evaluations but Sophie has five publications in top twenty journals and Yolanda only has two. Is it rational to choose Yolanda over Sophie on account of her pedigree? Obviously not since Sophie has managed to do better than Yolanda without having had her opportunities and is probably the smarter of the two. So in this case too pedigree should not be a factor.
Case 3) They are both two years out from their PhDs with good teaching evaluations but Sophie has two publications in top twenty journals and Yolanda has five. Is it rational to prefer Yolanda to Sophie partly on account of her pedigree? Obviously not since although Sophie has not managed to do as well as Yolanda she has not had her opportunities and may still be the smarter of the two. The right decision may still be to hire Yolanda, but it would be wrong to use her lack of pedigree to diminish Sophie’s chances still further. So in this case too pedigree should not be a factor.
Conclusion. Pedigree should not be a factor in discriminating between candidates whether the well-pedigreed candidate fares as well, worse or better on other criteria than her less well-pedigreed rival. Hence pedigree should not be a factor in discriminating between candidates.
For more on this see my posts
at Leiter on the-job-market-and-the-stale-phd-issue-once-again.
at New Apps on academic-pedigree-and-exclusion-is-going-for-pedigree-racist-and-classist
and
at Philosophers Anonymous on the PGR Challenge.
I’m not convinced that just because Sophie went to a program with fewer resources (in the many forms resources in grad programs can take) is likely the smarter of the two in cases 1 and 2. I don’t think we should aim to determine who the smarter of the two is. We might want to evaluate who we think is *doing better work* and is more likely to do good work in the future. But these are not the same as being “smarter” than another. I’m not sure I can gauge smartness in a candidate; I feel more able to gauge who’s doing good work.
Conciseness can cause confusion. By ‘smart’ I meant more talented at philosophy. X can of course be smarter along some dimensions than Y but less talented as a philosopher.
I think that there are often some very bad inferences going into using pedigree/grad program prestige as a reliable proxy for quality of candidates. Chief among them is the idea that all candidates would apply to a “top” program (or would if they could). The inference is thus that people only go to the best program they could be admitted to, and this lets us use program quality (“prestige”) as a proxy for candidate quality.
But this is specious. Many people apply to the programs that they do for other reasons, such as family, proximity to friends/family, finances, not thinking about “prestige” and job market considerations, and many other reasons. It’s simply not the case that those who go to a non-“prestige” program only did so because they couldn’t get into a “prestige” program.
A few points:
1. I agree completely with Rachel. I know reputation was not even in the top 3 reasons I chose my current (non-philosophy) PhD program, and I’m sure that’s true for a large number of people in philosophy.
2. When you place that much emphasis on pedigree, you are basically deferring to the admissions committee at some other school. You are assuming they are better judges than you of an applicant’s quality.
3. I think a useful mental exercise when evaluating things like this is to imagine you are a historian a hundred years from now reading about the philosophy profession now; wouldn’t you find the focus on pedigree ridiculous, the same way we see time periods when non-whites, women, Catholics, Jews, non-gentry/nobility, etc., were closed out of educational opportunities or had it exponentially harder?
I think that a candidate’s school should only matter if it can be demonstrated that there is something unique about the PhD program in question that has consistently provided its graduates with skills that make them more appealing candidates. If all candidates from a given program have had exceptional preparation as teachers, opportunities to do empirical or interdisciplinary work, or just really rigorous coursework and comps prior to the dissertation phase, then that’s a good prima facie reason to believe that a particular candidate from that program might be an exceptional teacher/researcher. Of course, any inferences about a candidate’s quality on the basis of her PhD program ought to be supported by evidence from her application file, or perhaps used as a tie-breaker between two equal candidates when no other information is available.
So, I wouldn’t say that a candidate’s school shouldn’t matter – only that it should matter for the right reasons…
hope I’m not repeating what someone else said here, but there *is* a difference, I think, in how much leiterrific programs encourage students to publish while still in grad school. so, the CV of a leiterrific ABD might have fewer pubs than that of the CV of an active, non-leiterrific ABD. At least, at the leiterrific program I went to, some faculty actively and explicitly *opposed* students’ so much as attempting to publish (though there was some disagreement about this).
This was in stark contrast to the advice my friends at less leiterrific programs got. Plenty of ABDs from my program who had no articles went on to land great jobs. *they then published a pile of great work they were sitting on.*
so comparing side-by-side 2 identical ABD cvs, each with same # of pubs, it *might* actually make sense to fa vor the pedigreed candidate, on the assumption that s/he has been dutifully following theadvice of her/his advisors to eschew publishing in favor of improving the dissertation.
a couple more points, in response to Drew’s (i’m anon from 1207)
“2. When you place that much emphasis on pedigree, you are basically deferring to the admissions committee at some other school. You are assuming they are better judges than you of an applicant’s quality.”
well… it’s a bit more complicated than this. once you get into a pedigreed school, you can accrue certain advantages which benefit your work (I’m *not* saying this is fair). You work with great faculty. You see your peers getting great jobs and anticipate one yourself, so you’re not living in (as much of the) constant fear of job-less-ness that some other grad students have to endure (again, *nooot* saying this is fair). So you can focus on your work a bit less burdened by that anxiety. Also, you are discouraged from publishing or at least aren’t positively encouraged to publish, so you can develop your original work in a more leisurely way, taking your time to rehash and rethink (i’m assuming that a lot pubs from grads will be relatively minor points or critical replies — i’m not saying these aren’t valuable, but that’s time that could be spent on your Positive Cool Proposal).
So, fair or not, pedigree has its advantages, which can add up to a more polished, mature project. You as a search committee member aren’t going to read the dissertation, you’re just going to briefly glimpse an itty bitty sliver of it. it may not even be in your area. So you *can’t* tell whether a candidate is sitting on a such a project. so yes, you cut corners and use the pedigree heuristic. (this is to echo the ‘meta-evidence’ explanation offered earlier).
“3. I think a useful mental exercise when evaluating things like this is to imagine you are a historian a hundred years from now reading about the philosophy profession now; wouldn’t you find the focus on pedigree ridiculous, the same way we see time periods when non-whites, women, Catholics, Jews, non-gentry/nobility, etc., were closed out of educational opportunities or had it exponentially harder?”
Ok, I *must* object. Being a woman or a jew is not like going to a leiterrific program. of *course* there is a classist and racist element in who gets in to what schools, both at the undergrad and grad level. to that extent, pedigree is just another unearned bit of status. But my own alma mater is an *extremely* not-selective, not-pedigreed place, and i still managed to go to a leiterrific place for grad school. I’m *not* suggesting that anyone could do this. I was not a single mother at 16 (or ever). I did not overcome phenomenal obstacles. I just went to a shitty college.
a few of my peers in my Phd started at community colleges or went to similarly non-ivy-covered schools. a couple actually *did* come from surprisingly non-privileged backgrounds. and yes, many did come from ivies — and I am aware of the worrisome correlation between undergrad pedigree and grad pedigree. But the ivy grads I went to school with were extremely hard-working and passionate about philosophy, approached arguments damn originally, and had scary amounts of brain power (that brain power may also reflect years of educational privilege, i will grant that). privileged? some of them, sure. but there was some combination of privilege, raw talent, *and hard work* that got them there. i’m pretty sure hard work never even so much as *contributed* to anyone’s being a woman or a member of the nobility….
Sum: blind reliance on pedigree is silly and lazy — I completely agree with that. but in contexts of limited info, it might actually be a good proxy for how prepared a candidate is (at least at the ABD/new PhD). of course, as time since the PhD increases, reliance on pedigree should be less important. and my impression is that it generally is, and that’s a good thing.
a couple commenters have suggested that it is “talent” or how “smart” someone is that should be what search committees aim at. A different commenter suggested that there is an an unfairness in looking to pedigree, because the non-pedigreed are disadvantaged twice over..
second thing first, search committees aren’t looking to remedy gaps in candidates’ past training. fairness is not part of what they’re after at all. they’re wanting a candidate who is already trained-up, a ready-to-hit-the-ground-running, sure thing. They want someone who is ready now to take up up his/her fair share of the teaching burdens, publish enough to stay on and not upset a dean, and generally make their lives as easy and pleasant as possible. That’s the SC’s prerogative. they’re looking for a peer they can rely on, not for someone who they can help out of a bind. and for them, that is an utterly reasonable and fair way to approach it, as an exercise in finding a colleague, not as a way to remedy past injustices.
first thing second, no one cares how “smart” you are when it comes to being a faculty member. they want you to be competent, able to write, productive, a reasonably good teacher, and adult and organized enough to make it to committee meetings. most philosophers are plenty “smart”. search committees want much more than this –a competent colleague. however they got to be that way is irrelevant. indeed, i’m sure most search members would not think twice in choosing between a “smarter” candidate and a competent one (not that anyone has suggested the contrary, just putting it out there). they want someone who will hand the grades in and time and be an active scholar. someone “brilliant” who can’t do those things will end up someone they regret hiring.