Faculty hiring dominated by graduates of elite institutions
aaas.org
aaas.org
The top 25% are not what I think most people would call 'elite', though maybe there's a different understanding within the industry (i.e., in academia). Is the top 25% considered 'elite' in anything? Top 25% of software developers? Chess players? Playwrights? Restaurants?
Elite means to me the top 6-10 schools. I don't know how many are included in the top 25%, but I'd guess over 250 (anyone know?).
I'm surprised it's the top 25%; that's actually pretty broad-based and seems appropriate. The professors are supposed to be the 'elite' among students. If they lack the academic ability and drive to even attend a school in the top 25%, do I want them as a professor? (Of course there are exceptions.)
I think the problem with the statistic is that it points to tenure-track maybe not being meritocratic.
From [1], “placements of nearly 19,000 tenure-track or tenured faculty, among 461 North American departmental or school-level academic units, in the disciplines of computer science, business, and history (see Supplementary Materials and table S1).”
The 25% is from 461 universities, so probably 115 or 116 universities is included in that number.
Edit: On further reading of the supplement, there is a table that says they investigated CS faculty from 205 universities with 87% "doctorates in sample" meaning 87% of faculty having doctorates from those 205 universities being investigated. Business has 112 unis with 84% of doctorates in sample and history has 144 unis with 89% doctorates in sample.
It's not clear to me how many institutions of the "doctorates out of sample" are represented.
There is a stat in the supplement showing 50% of CS profs have doctorates from 18 universities, 50% of business profs have doctorates from 16 universities, 50% of history profs have doctorates from 8 universities.
It's possible the 25% of institutions could be closer to 51 for CS, 28 for business, and 36 for history.
> The 25% is from 461 universities, so probably 115 or 116 universities is included in that number.
That seems to confuse the school whence the professor graduated, and the school at which they teach. 25% refers to the former, 461 to the latter, afaict. Or am I the one misunderstanding?
If I understood the supplement to the paper correctly, there were only 205 CS departments investigated and fewer departments for business and history. The 461 would refer to the sum of universities across the CS (205) business (112), and history (143) fields, which is where faculty are currently teaching.
The supplement had a statistic saying 87% of the profs in CS departments had “doctorates in sample.” Which meant 87% of the profs came from those 205 schools. I couldn’t find how many universities made up the remaining 13% of faculty.
The supplement says 50% of CS faculty came from 18 universities. I believe the 25% is referring to the 205 CS programs (which is both where the profs are currently teaching but also where most of them earned their PhDs from - the 87% doctorates in sample statistic) which would be 51 universities producing most of the faculty at the 205 universities in the study. At least I think…
For all I know, I am misunderstanding the supplement, and further muddying the waters with poor explanations.
I also graduated around the 2008 financial crisis. The grads from MIT (and other elites) were scared about getting jobs, and applied to schools way below their "station". The geniuses at said departments didn't interview non-elites and focused on trying to recruit the prestigious candidates. End result was that the elite students didn't accept these lesser offers (duh!) and about 2-3 cycles of non-elite but strong grads were not able to get academic positions (or had to suffer the pain of post-docs).
Academia is just such a stupid system.
It was especially disappointing when I found incredible research being done in "less elite schools", which I was explicitly told to not apply to or "I would have no chance getting to tenure track interviews." But, guess where those professors at those "less elite schools" came from? Those same 6 schools! It was absurd.
At my own B-tier undergrad school, which I went to because it was affordable and offered me a lot in scholarships, all of the new hires were also from those schools. The whole system was absurd.
Things working against me when I was finishing up (early-mid 90s), included being at a lower tier school, the massive overproduction of Ph.D.s in physics, combined with reducing NSF and related budgets, and the fall of the USSR, which resulted in extremely senior faculty decamping from Russia and taking every possible post in the west.
This was the time of the Young Scientists Network in the late 80s through mid 90s. Many other soon to be and freshly minted Ph.D.s learned of interminable post-docs, getting your first tenure track when you turned 40 or so, and not getting a tenure decision until you were 48 or so.
I skipped the post doc, and went into industry. Means I didn't get to do much science (occasional collaborations here and there). But I was able to support a family. That latter part was more important to me than anything.
As my daughter shortly decamps for grad school, we are telling her how it really is, and what to expect. My simple guide for her was do what you love, but have a realistic plan to be able to support yourself at the end of the process.
No one ever said that the system is fair. It isn't. But then, neither is life.
As you say, it is unclear what the actual 'product' of low tier schools actually is - yes, many will graduate with a PhD funded by public and/or private grants, but ultimately low tier PhDs aren't valued much by either the academic or industry job markets. One possible explanation is that the 'product' of the low tier PhD programs is to further the existence of the PhD program.
Edited for clarity
"The observed placement rates would imply that faculty with doctorates from the top 10 units are inherently 2 to 6 times more productive than faculty with doctorates from the third 10 units." At least looking at top CS conferences the authorship of accepted papers ARE in fact this correlated (it seems to me).
I expect the number to be more like 100%. Just like I expect lawyers and doctors to have attended Law and Medical schools.
I went to Stanford and cal (undergrad and master’s respectively). I can tell you academia is plagued with elitism. There’s almost a class system that parallels the idea of an Ivy League or an elite institution. People who get in, feel special, that they are better and worked harder than the others, and therefore want to keep the others out.
This extends beyond academia where a lot of hiring decisions boil down to who went where. It’s ultimately a private club where you’re not invited. Cal and Stanford grads can get away with it in Tech, but the minute you go higher up the ladder, suddenly Harvard (mgmt) grads make you feel you’re on the outside.
It's not just professors either. Look at any group where academic credentialing matters (and even groups where it's not supposed to matter).
Most of the YC partners are from elite institutions. Many of the YC founders are too (to YCs credit this has gotten a lot better recently, but a while back it was either Ivy League or Berkeley/Stanford/Washington/CMU for the most part).
And I don't think YC does this intentionally nor many of the other groups where it happens. Academic signalling is just so ingrained that it happens subconsciously.
If you submit an article to Nature or Science from, say, Oxbridge it has much higher chances to get in than if you submit from a good university like Southampton or Exeter.
Professors from elite institutions are really well connected to editors and funders. Actually, many editors come from the same institutions. So while reviewers might treat you fairly if you submit from a less prestigious institution (I think this is also questionable), the editorial board will be less likely send you article for review.
So, yeah, I can see that. I graduated from (at that time) an R2 and was applying at R1s. Nope.
I will say it again, that means ALL RESEARCH (which informs government policy) IS TAINTED by the views appropriotated by these people. That means assumptions, starting and ending points, methodologies.... ie. drug addiction, mental health, treatment of aboriginal populations, behavioural, gene, medical, tainted. Anything, tainted.
Tainted.
Just like journalists, scientists can select which ideas to research, which ideas to promote, which ideas get funded, etc.
So if all of science comes from people with one homogeneous ideology, you're going to get homogeneous ideological research.
If journalism can be captured by certain ideology's why wouldn't scientific research be able to be?
What percentage of the top schools is needed to fill all the positions? Top 5%? Top 10%?
I'd be highly surprised if it the top 25% of schools were necessary to fill all spots.
1) The scenario where signalling matters above all else. So there's too many highly-credentialed candidates for the openings at the institutions they graduated from, so they apply to jobs at lower-tier institutions.
2) The scenario where these institutions are producing truly better candidates and the interview process is robust. There's still too many great candidates for the "same-tier" positions, so they apply to lower-tier.
This rules out a few scenarios such as: "the interview process is robust BUT these institutions produce no-better candidates" but it doesn't tell us much specifically about the interviewing process or the candidate quality because of the candidate pool size.
A few interesting followup data points would be:
1) what ratio of graduates go into faculty positions for each tier of school. It must approach 0% at the lowest tier, but how quickly does it fall off?
2) what does industry hiring look like? Do you see a similar distribution, or are faculty positions preferred by enough PhDs that top-tier industry companies have to "settle for" a broader set of schools?
25% seems like a pretty wide pool to me.
I don't see how it could be any other way?
The article starts:
“ Faculty careers are shut off to all individuals with Ph.D.s except those from a small number of universities,”
I’m wondering what % they think would be expected if it wasn’t ’shut off’ ?
Should it be <1%? E.g. equal to the number of elite graduates? What would this outcome presume?
In regards to FAANG, it will help but it's way more fluid than some law firms that only hire Ivy League.
That's quite a self-serving narrative for us on HN. Do we have any evidence that the degrees are harder to get or that tech is more meritocratic?
> Tech is also significantly less white and full of immigrants (because of the meritocracy)
Tech's diversity record isn't evidence of meritocracy. Compare women in law, business, and media and those in tech, for example. It also may be that law, business, and media require exceptional cultural communication and understanding, which may be hard for a new immigrant to learn.
Later in the career though, all bets are off. You only have to get into an interview room and solve 5 questions on a whiteboard (6, if you include the phone screen). If you're in Silicon Valley and have "software engineer" on your LinkedIn profile you're guaranteed to get at least one shot at each company.
[0] https://www.wallstreetoasis.com/forums/8-years-old-with-no-f...
The expensive school to Oxford pipeline is definitely a thing, but it's much more prevalent at undergraduate level, at least if we're talking about expensive British schools...
So this isn't too say you're wrong - there is definitely something about better students go to better schools and there should be some correlation there, but this is way bigger than that.
That's how it works in every single industry for better or worse.
> 25% of the institutions produced 71 to 86% of all tenure-track faculty
which means that the vast majority of PhDs who got their degrees from the other 75% of institutions never got on the tenure track.
ratio of professors to PhD grads = top percentile who get those jobs that want them
Ooh, why not make a hypothesis and test it?
Anyway, where the hell did the headline come from if "we can't say yet what causes ..."?
Personally I see this actually as a fairly good thing as a student - when I was in less good schools I was always happy to know when a teacher/professor had gone to a top school themselves. You want to be taught by the best.
Academia is really hard. There are very few spots, and there's a wide spectrum of talent that exists. The top 25% is a very large spread already (I'd suggest the bottom of that 25% shouldn't really be called elite), and you will easily encompass a huge portion of the best in it just by virtue of pareto distributions. There will possibly be some weaker talent that might get a job that could have been had by an equally weaker talent from a non-top 25% school, but I seriously doubt that's a general rule. There's just not enough slots, and the top talent is so much more obviously capable than the bottom talent.
For reference, I've personally attended a wide variety of schooling in the US (public & private pre-grad, community college, lower-tier state college, upper-tier state college, and elite private graduate) and seen the differences in the population makeup. There has been top talent everywhere, but the percentage at the school has varied drastically (from <1% to >50-90%).
I don't know why we pretend that all people are created equal, from equal backgrounds, and should be expected therefore to have equal results... and if they don't, then bias must be inherent! This is to me ideals of what society ought to look like forced onto messy reality where it's patently not the case. Not everyone is an A+ talent, and in fact, those are rare by definition. The percentage of A+ talent is going to drop as you go down the totem pole. That's just life.
I think the number of people who truly believe that are low. But you'll find people all over the map in HOW they disagree with it!
For instance, if you point out that some people have unequal backgrounds, you might get labeled "woke," especially if you push for certain policies to try to mitigate that!