I'm an academic and I can point out a few factual clarifications:
1. There's R1 (e.g., Ivy+, Stanford, MIT, ...) and not-R1 (typically liberal arts) institutions. Tenure-track professors at R1 institutions are hired to do research and the teaching is incidental. It's flipped at not-R1 institutions. I _believe_ tenure-track professors teaching at R1 institutions is mostly a historical accident that got cemented into every part of the tenure/promotion/grant process so it's basically impossible to change at this point.
2. There's also STEM/not-STEM. As a concrete example, the top-4 CS programs don't hire adjuncts in the way you think about them (e.g., Andrew Ng is adjunct at Stanford but this is just to keep him affiliated with the University). This dynamic is very different in non-STEM.
3. In STEM, the University puts in substantial resources to grow tenure-track faculty (startup packages go from $500k and way up).
4. Poaching does happen a decent amount. You just don't really hear about it much because tenured professors have it very nice and there's no need to complain about the processing.
I'm also happy to share some non-factual thoughts but those are speculative.
Basically on the right track. But it isn't colleges who do the hiring; it's mostly departments. Colleges can veto departmental hiring decisions, but they don't pick the candidates. Departments obviously have slightly different incentives than colleges do (even more prestige-driven, significantly less budget-driven, less financially even very informed or aware).
I'm very confused by your not mentioning post-docs (which are the norm in STEM). In effect, these are temporary positions with a moderate teaching load (or perhaps a pretty heavy teaching load, or, if you are lucky, no obligatory teaching load at all). I don't know what data your graph is based on, but it's rare now for people to get tenure-track positions (at research universities) without doing a postdoc at all.
As for poaching, my impression is that it does happen a moderate amount, but usually only when the professor in question is upgrading to a more prestigious institution or wanted to move anyway. It is also common for professors to get an offer from an outside institution and then ask their current institution to match it, which they often will.
Adjuncts surprisingly shooting up several years into their career is a rare thing, not least because their teaching schedule (6+ courses a year not unusual) leaves them little time for serious research. (Of course, in lab-heavy fields, they face even stronger headwinds.) I think that's the main source of the vicious cycle you're talking about. Journal editors, AFAIK, don't usually gatekeep much; a non-famous tenured professor and an adjunct with an .edu address have about the same chances of getting their paper in all else being equal. (Really big names sometimes get the "even if it is unreadable, there must be wisdom in it" treatment.)
1. lots of people go on the tenure-track job market the year they get their PhD, so you get a big burst. Then are people who delay this for a year or two or three — it's a pain to put together a TT application, but postdocs and one-off teaching jobs are often offered with just a cover letter. So you have two distributions; a delta function at "same year", and then an exponential decay over the subsequent years — this is marking the time at which the applicant finally makes a first TT application.
Notice that this explains away the apparent "bias against experience", simply by virtue of the fact that people choose to go on the market at different points.
2. People who are out of the market longer are seen to have a hidden variable (not being as good) that is serving as a signal.
For TT people, in general, they do not move from university to university very much; the standard thing is get a good job when you start, and then negotiate when you are up for tenure by getting outside offers. It's rare to go on the market, and generally indicates someone who is at a low-ranked institution who wants to move up; those applications are less likely to be successful, since being at the low-ranked institution is itself a signal.
More generally, people would rather take someone who looks like Harvard would hire, than someone who (even if they look like someone Harvard would hire) definitely Harvard didn't actually in the end hire.
As to the question why do people hire from outside when they do take people with experience: my guess is that academia is ridiculous and snobby, and it's very hard to get someone to see an adjunct as a TT. They're excluded (for example) from certain meetings and decision-making, and have less ability to negotiate for what they need, and people use that as a signal of actual quality. But it's much easier to say ha, Berlin totally doesn't realize that their guy is super good, we can get him, let's do it.
An inside hire does occasionally happen (my institution is smart, and does this — it's a great way to test the waters with someone), but it's also quite clearly flagged as the plan ahead of time (we think this person is good, hire them as a VAP, but there's a job coming soon...)
A similar phenomenon happens in the non-FAANG computer programming world. Junior devs get hired at something like $60-$80k, then within about two years are worth double that. But the company that hired them will never promote that dev to a $140k salary because it would upset the apple cart of all the devs making $120k wondering why they don't get a 100% raise in two years, or anything close.
So the junior dev has to go elsewhere for that first big pay bump. Meanwhile the company hires an ex-junior dev to replace them, and happily pays them $140k. But no one leap-frogged anyone or got an out of place raise, so the apple cart remains placid.
Academia is filled with mediocrity & a lack of imagination
PhDs that are above average get quickly snapped up in tenure track while others have to cycle around various adjunct positions . Experience can only help so much if your initial base itself is low
Some things you need to include in your investigation:
1) In reality, a lot of the teaching is done by graduate students. Even when it is not explicitly, it is very common that professors will have a teaching assistant that will do most of the menial work of teaching (grading, conducting discussion sessions, etc.).
2) Connected with the first point: in high prestige institutions, they usually hire for the research potential (which includes ability to attract grants, which is arguably more important than how promising it is), with teaching as a not particularly close third place (second is DEI nowadays). Adjuncts and post-docs are never hired back because they are expected to be looking for a tenure-track anyway.
3) You are correct in assuming a career gets shut down if you are denied tenure. It is very rare one can recover from that, usually involves some level of nastiness that happened for tenure to be denied in the first place, something other institutions will look over. But in most cases it is an indication of lack of impact research-wise. And when it happens, why hire someone who is also tenured (thus, more expensive) or already flamed out elsewhere when you can hire a new student? It is similar to sports, you don't run it back with veterans when you failed, you draft the potential next big thing because it is cheaper and more likely to make an impact.
4) Which leads to the final point. There is a lot more graduate students than there are jobs. Like, by a mile. A majority might never even try (myself included). So you don't need to go for experienced teachers for adjunct jobs that are temporary. There are plenty of people jumping from a one-year appointment to another for a few years until they give up and go work somewhere else. There is no short supply at any point. And when people have no experience, you can pay them less.
One advantage to only hiring outside of your own institution is that it enables you to select for ideological purity more easily. And academia is extremely ideological. If you make it possible for local academics to be hired for tenure, and a promising one is passed over, it leads to questions about why this person wasn’t picked. In many cases that may be because the academic wasn’t ideologically pure enough. If you make it the default to hire from other schools, then this doesn’t happen, as you can merely dismiss the non-ideological fits at other schools without needing to explain yourself, and only select the ones that do fit ideologically. Anonymity neuters outrage.
For what it’s worth, my personal mental impression is that this IS happening in other jobs. At least, the ones that also confer a lot of white collar status.
Anecdotally, I’ve heard people say a similar thing about lawyers (many more law firms hiring people as paralegals who they would have hired as lawyers; far fewer people being offered partner than in the past).
I don’t know how it is in medicine, but it certainly seems like, whenever I have a medical complaint, I can take my pick of a 100 RNs or PAs, but man, finding an actual MD with an opening anytime soon seems REALLY tough.
Here’s an interesting aside - writers and bit part actors are currently complaining as much about the job market as anyone. Supposedly, mega-star salary stays the same, but the people on the lower rungs - who in past times would have had a stable mid-paying tv writer job, or consistent living as a recurring guest star on a smaller sitcom - claim it’s becoming financially untenable to exist in these smaller jobs in huge industries.
Two immediate thoughts here: white collar jobs have prestige that people might be more willing to compete for.
In my experience, applicants to this jobs also are more likely to have wealthier parents who can help pay their way if they end up as adjuncts. Part of me wonders if plumbing would be able to get away with milking the “apprenticeship” period way more if they could get away with it. If the apprentices were more likely to be able to pay their own way for the first couple years, would salaries show the same pattern?
As an academic, I believe it is mostly the case of the superstar theory. There appears to be an extreme Pareto principle / power law at play, where the best researchers are monumentally more impactful than their only slightly less talented peers. Partly this is because of funding and publication dynamics, where a successful researcher gets exponentially more money/opportunities to continue being successful (see also Lotka's law on publication count). Given this, it makes sense to take a shot at getting the next big thing, more than having a wide pool of moderately successful mid-career researchers.
Colleges (at least R-1 in STEM fields, see diddly's comment) do not hire all low-commitment adjuncts as you suggest because the startup packages are substantial and they invest a lot in making sure new hires can be successful. But there is a similar process in the usually 7-year tenure review: professors are given 7 years to "prove" their research output before either being given permanent tenure or being fired. It tends to be a very stressful 7 years in a young faculty's career. So in some way there is a bit of a "free trial" for even tenure-track faculty that serves the same purpose as an adjunct pool would.
Maybe every academic is faking their data, you want to hire someone who is happy within the first year to start faking data because someone who is reluctant is dangerous.
And you want people to move around because otherwise fellow adjuncts will notice their not very brilliant has suddenly made great discoveries and the simplest explanation is fraud.
Probably overly cynical but I think it ties into a world where only 17% of authors provide data when it is requested.
The top research universities do compete against one another and do bid up salaries, bonuses (aka summer salary), and other perks for the top people. At the same time, there are norms against the ratio of top to bottom salaries within rank -- typically top tenured faculty will receive 3-4x the bottom in the same discipline and rank. Top academic salaries now run in excess of 400k for 9 months.
One of the things I have long espoused, in the sense of thinking it but literally never telling anyone, is that the United States Congress should establish a system of National Teaching Universities that do exclusively undergrad and non-degree training. Recruit early- and mid-career academics with modest conventional prospects by using "Moneyball"-style strategies on the assumption potentially great teachers are being underutilized. Recruit outright non-academics, such as senior enlisted military personnel, who have an extremely deep reservoir of meta-knowledge about how to train people that is underutilized in the white-collar world for the crasset possible class reasons. Completely exclude research from the main career path and consider imposing outright guild-like rules preventing your hires from even publishing research while they're employed with you, and perhaps for six months afterwards. Require everyone to go through extensive, demanding paid training from which many drop out, and make sure all the training materials and training apparatus is built from private-sector and military people uncontaminated by the schoolmarm-industrial complex and its fraudulent "education research."
Basically, the bargain here is just that if surplus academics are willing to compromise SOME of their self-regarding and comically left-wing "ideals," move to some underserved working-class community, and actually make themselves useful, while abstaining from the corrupt academic system somewhat in potentially controversial ways, right-wing taxpayers should and reasonably will open their wallets and hand out genuinely life-changing compensation. It would require legal cover because the academic bums have written their special privileges into the legal system itself in deep ways, but that's why you have a legislature. You can literally just change the law to fix things. Conservatives control the Supreme Court and it would be pretty perverse and surprising if they actually sabotaged a program like this by enforcing special rights for academic dweebs who want nonsense guild rules enforced as a matter of first-rank constitutional principle.
Nobody is going to do this because conservatives fundamentally do not believe it is possible to fix anything and do not care. Their idea is just simple-minded buffoonery involving a rotating cast of villains, always tied to modernity, diversity, and social change in some vague shifting yet obvious way, who are thought to be corrupting the otherwise pristine and workable system through secret collusion. So even though solutions like this are actually kind of obvious and plausible (defuse social tensions by paying surplus academics enormous amounts of money to educate people who could actually use an education to do something, as opposed to just to sustain the education system) you will never see a Republican legislator ever propose something like this. Because their social base is the rural grandees and local bigwigs who claim to speak on behalf of dispossessed working class folks while privately colluding to cheat them so they can send their kids to Harvard as they always have.
"Acceptance rate for tenure-track positions by how long it’s been since the candidate has gotten their PhD."
The chart doesn't show the rate, it shows the number of hires. That's very different: It depends on the number of applicants, and five years from PhD is a very weird time to be applying for tenure track jobs (given that you're close to tenure, but probably not close enough to be hired with tenure.) And it's from the American Historical Society, which is not likely to be representative of academia as a whole.
“Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?”
This does happen. First as post docs in many scientific disciplines. And then you are hired w/o tenure on a 2-4 yr contract, hopefully to get tenure at year 6-9 (varying on field and institution).
Experience isn't always good. Fresh faces bring enthusiasm, attention and a joy for life. If the numbers really matter, bring in the old hands. If it's more of a wishy-washy social thing, out with the old and in with the new.
Plus, in the 2-tiered system, do professors really want a similarly-experienced, worse version of themselves challenging their authority, angling for their job, or to feel sorry for?
Plus plus, and this is the big one, in the current system, students only see academic success stories and aspiring successes. Hiring old adjuncts is putting the losers of the process on display, (potentially) discouraging students from entering the meat grinder themselves.
“The value of the #1 brightest new PhD is that she has a 5% chance of becoming a future superstar ... Once you’ve been around for five years, colleges can see your track record, satisfy themselves you’re not the next Einstein, and lose interest.”
That’s a large part of it. So much hiring into the entry-level tenure track is based on buzz, and any university with pretensions is hoping they will lock in (even for a few years) the next superstar. They would rather take a chance on that than hire a solid performer 5 years out who would do a fine job but probably not become a household name. Some of this becomes a self-fulfilling prophecy; someone who starts their career teaching four courses a semester at a second-rate place with little or no support is trapped. Publishing your way out of that (no grant writers helping you, poor lab facilities, few good students or research-active colleagues, paltry travel funds) is like climbing Everest.
At the senior level there is poaching, but again, only for the creme de la creme. Otherwise I think there is (1) age discrimination, and (2) the fact that new PhDs are (relatively) cheap and senior faculty might not expect big bucks, but more than entry-level pay.
> Here’s another question that confuses me even more: Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?
There's another, more mechanistic explanation for why adjuncts who transition to tenure track faculty do so outside of their own university.
If you were to rank the university positions by prestige you might get a ranking which looks like
1) First Tier Tenure Track
2) Second Tier Tenure Track
3) First Tier Adjunct
4) Second Tier Adjunct
Adjuncts who get promoted one level up either get more prestigious adjunct positions or get tenure track at a lower ranked university. Getting tenure from the place you are adjuncting at is harder because it requires getting promoted two rungs up the ladder.
"Supply + demand" and "weird path dependence due to historical circumstances."
Maybe the most interesting book on the topic is Louis Menand’s *The Marketplace of Ideas: Reform and Resistance in the American University* (my comments are at https://jakeseliger.com/2010/01/21/problems-in-the-academy-louis-menands-the-marketplace-of-ideas-reform-and-resistance-in-the-american-university-2/). Basically, the academic job market worked reasonably well for would-be professors up until around 1975, when the number of undergrads fell (baby boomers worked their way through the system) but the number of PhD-granting institutions and programs continued to hum along. Since then, in most but not all fields, the number of PhDs has way outstripped the number of slots. When supply exceeds demand, strange things happen.
The most interesting question is why people keep starting PhD programs that are designed to produce all-but-dissertation "students" who can teach classes for very little money, and adjuncts. Granted, I was one of them, at one point.
Maybe academia is the brainiac's version of gambling in Vegas: likely to lead to losses, but fun in the moment.
Assoc. Professor at major R1 "brand name" university here (~12 years experience). This is mostly right. A few things:
1. STEM is very different from non-STEM. Title is the same but there is little actual overlap in what the job is like day-to-day. I am in STEM and run a lab with 20 graduate students and post-docs. This just doesn't happen on the "other side of campus".
2. Superstar effect is a big explainer. People want Steph Curry, and are willing to fight over potential candidates that have that potential. Experience does not turn [insert middling player] into Steph Curry.
3. You shouldn't think of it like you are locking someone in to a career path at the first blush. They have gone through (often) 10-12 years of college/grad school/post-doc by the time they even try for TT. Like it or not, potential is seen as at least somewhat baked in by the time they finish their PhD or post-doc. Remember, they are often nearing 30 years old at that point. In other industries this would be like a promotion of your leading 30 year olds into a management or leadership track.
4. Teaching is a small part of my job. Not going to give an effort/time percentage, but in practice it is low. This is why there are effectively "two markets" for TT and adjunct.
5. Adjuncts are treated horribly. Thankfully we have very few here (remember, "name brand" university). It just kind of sucks, and is very similar to "winner take all" dynamics happening more broadly in the economy.
Here in Germany, 1) you have to change universities - "to avoid intellectual incest". You can not simply do your Master at Humboldt-Uni, then your PhD AND then get tenure at the same place. Some faculties solve this by close cooperation with one other department at another college/Universität - swapping their presumably best and brightest. 2) There are now some laws to stop colleges to give their cheaper teaching-staff just one 2-year-contract after another - now it's either end the co-operation or give them a long-term contract (not tenure, but quite secure).
I was never going for that rat-race (way too dumb for it, too). Still, in my early career I got some very fine&well paid teaching jobs at the DAAD (German academic exchange service), though still kinda inexperienced - but 5% of us turned out fine PhDs or even professors (id est: valuable alumnae). When at 40 it was obvious to them I just wanted to continue being a good teacher, they closed the door. Which is fine; grateful for 11+ years on German tax-euros.
Here is another question, not addressed here. Universities rarely fire people. I know they cannot fire tenured people. But many of the people hired to just teach are really bad at the teaching part, but it is very uncommon (compared to the private sector) to let someone go. Whereas on the tenure track there is real accountability for grants and publications, off it there is almost no accountability for teaching quality.
This seems very simple to me. You've worked for me for several years without getting tenure. You are likely to continue to do so. If I am going to offer someone tenure, it should be an individual I believe that I would not be able to entice without that offer.
For one thing, football and basketball teams are way more visible to way more people than all but the very best researchers, and frequently not even them. And you can't really 'watch' researchers, even excellent ones, do their research in competition with others.
And in my discipline (Political Science) about 98% of the research is just stone stupid. (I can say that. I'm retired.) Irreproducible, statistically incoherent, and generally of little practical consequence.
Having said that, I've had friends do good, solid work because they enjoy the topic, but the topic is not 'sexy' so it just disappears. I have worked on some very good stuff but even that has pitfalls if it doesn't adhere to whatever the current narrative is.
Anyway. I was a "late bloomer" and got a full-time adjunct position that paid very well and had great benefits. But I negotiated that first contract before being forced to join a union. So, on average I got paid about 3 times what my fellow union members got. (Thanks, AFT.) Also, I was qualified to teach courses that were in high demand. Very high compared to supply.
After about 5 years (I was then getting 3-year contracts) I found that I had too little time to do both good teaching and good research (I contributed causal modeling on the work of others) plus I got asked to do something I considered unethical, so I dropped the research and spent the rest of my career just teaching. When an appropriate tenure track positioned opened (3 times) in my department I was asked each time to apply, with some assurance that I would get the position. By then I liked what I was doing, so I refused. Why do relatively worthless research just to gain tenure? I had the equivalent by being very, very good at my job, and a lack of easy replacements.
Look. We really need a two-track tenure process. One for research, and one for teaching. I know more than a few good researchers who aren't that good in the classroom. Why not reward both more appropriately? I also know that that currently tenured loathe that idea.
Also, remember that I was at a second-tier state school. Maybe around #150. We were good at what we did when we concentrated on it. But we were NEVER going to get a #1-100 new PhD. Sometimes they would apply, but they just used the interviews to bounce to the next one at a school they liked.
So. Why did my school try to hire the same way? Because the top admins are just as socially conscious as the top at Top 25 schools. Plus, to the extent that they can hire adjuncts cheap, it looks like they save money. (I'm not convinced of that, but that's a whole nother story.
I do feel bad for those who take adjunct positions in the hopes of landing a tenure track job. But several of our ABD adjuncts have made that happen, and one full-time adjunct colleague landed a tenure track job at a decent private college. But it's kind or rare.
We also pump out new PhDs for lots of bad reasons, and at least in my discipline we've been told since at least the '70s that a "wave of retirements" was coming and there'd be lots of jobs. But we study politics, so you should expect lots of lies.
Also, many "intellectuals" seem to enjoy having lots of indentured servants around them. That's a psychological problem that I'm not competent to help with.
The analysis in part I is spot on. As for why they only hire inexperienced people, I would add that adjuncts have experience in the wrong job. They have experience teaching, but excellence in teaching is not what is being hired for. They don't have experience in research (because adjuncting does not provide the time or resources to seriously pursue research, in general cases). Plus, there is an inverse signalling problem. You see that someone is an adjunct and you conclude that they were not good enough at research to get a tenure track job. So experience as an adjunct has zero actual value, and negative signalling value. And the more time you have spent as an adjunct the more negative signalling value this has. (`Jones was so hopeless he couldn't get a tenure track job even after ten years of trying!). Although practically speaking, the negative signalling value of any time as an adjunct at all is so strong that spending accepting an adjunct appointment is basically a kiss of death to chances of landing on tenure track.
In my fairly limited experience, it's less the case that teaching-track faculty never get hired for tenure-track jobs at their own institutions than that they rarely get hired for tenure-track jobs at any comparable institutions.
Academia is very hierarchical, and pay, teaching workload, research resources, and prestige are all very different if you're at a top-20 institution, a top-100 institution, or a non-R-1 institution. In my field (economics), pay and teaching load for permanent teaching-track positions at top universities are not too different from pay and teaching load for tenure-track positions at lower-tier institutions.
When I was on the job market, I was considered for tenure-track jobs at some top-100 research universities and some liberal arts colleges. I was also considered for teaching positions at elite universities. If I'd taken one of the teaching positions, done really good work, and gone back on the job market, it would be reasonable to expect that the jobs I'd be competitive for would be in the same basic ballpark as the ones I was competitive for when I went on the market the first time. It would require a pretty dramatic change in my productivity and quality to go from someone that the 100th best institution in the country would consider to someone that the 10th best institution in the country would consider.
As for why academia is so hierarchical, I'd guess that it's a combination of the fact that the most skillful/productive academics like to work with other very skillful and productive academics, plus the fact that outsiders (i.e. the people funding the work through their tuition, donations, or grants) don't know as much as insiders about what constitutes good work, so rely on academics giving each other seals of approval to know who to trust and who to give money to.
As for why superstars are not so mobile - there is a key difference to sports stars, in that sports stars are typically single and in their twenties, whereas tenured professors are middle aged, typically with working spouses and kids, and this makes them significantly less open to geographic moves (and very seldom are there two really good universities in the same geographic area). Also, for experimentalists, moving means recreating a lab from scratch which is a multi year undertaking (in which time you are not doing any research and thus losing prestige), whereas even for theorists you have developed a system that works for you, which presumably involves some web of local collaborations and so forth, and moving means tearing that up. People do move certainly, just the friction is greater.
There's also a second-order effect here: once grad students know that the odds of getting tenure-track positions are highest in the 12 months after they go on the job market, their incentive is to optimize all their work in progress to maximize the amount of publications, etc. that they have done as of when they go on the market. So they often have relatively little incremental publications that will occur after that 1st year, because they've focused on getting as big of a python lump of work out at once, rather than establishing a steady stream of incremental publications after that date.
As for your closing question: universities can and do engage in bidding wars, trying to outbid each other for superstars. However, there are a couple of factors constraining this (1) morale issues (plus pay transparency, which is the norm at universities) means you are somewhat constrained in how much variation in pay you can have (you can and do have factors of a few, just typically not orders of magnitude). (2) entering a bidding war and losing has reputational costs (university X tried to get Jones and failed, guess X sucks), which makes them reluctant to enter a bidding war unless they have reasonable costs of success. (This then gets compounded with principal-agent considerations. Whoever pushed to try and hire Jones personally loses a ton of political capital if Jones declines).
Why can't you get hired at the institution you got your PhD in ?
That's mostly to ensure that institutions at least try to hire good candidates rather than their own students. It also help with intellectual cross fertilisation - people from outside can bring in knowledge and methods that are not available in the department.
Why do they only hire inexperienced people ?
It seems like something particular to either the US or social sciences. In my field (maths) and my country (France) you typically only get a permanent position after a couple of postdoc positions (hiring committees get suspicious if you have not been hired after 5-7 years so being to experienced is not good either).
A lot of this agrees with stuff already said by others, but:
1. The pay difference between adjuncts and research/tenure-track faculty is largely about "research success is what drives prestige", as you say.
2. The "negative premia" for experience is basically just observing that people (mostly) go on the academic job market when they are completing their PhD, some succeed and get TT jobs (or in fields where postdocs are ubiquitous, good post docs that sometime after hopefully lead to TT jobs), others fail to and get adjunct roles / leave academia. For an adjunct after that point, there is a massive negative selection problem, plus as you say it's hard to do good research while teaching 6 classes a year.
3. "But then why do they only hire inexperienced people?" In fields where post-docs are ubiquitous, obviously people have more experience when applying for TT jobs (but if a post-doc counts as experience, why doesn't a PhD? Isn't the last few years of a PhD relevant job experience which people are judged on for TT hiring?) But ignoring post-docs, there's an initial clearing market of new PhD grads, some get TT job offers, others don't. (See 2). Beyond that, obviously they don't "only hire inexperienced people". If a TT professor is successful at lower-ranked university X, higher ranked university Y will poach them, but with a promotion to Associate or Full Professor. That people hired at Assistant Professor level tend to be inexperienced is just synonymous with the job title.
4. "Why only people from outside their own institution?" This is sold in part as being about avoiding nepotism. And in part its a "go into the wild word and prove yourself, and then if you are successful we will hire you back". In equilibrium it often works, especially seeing 99% of the time the school someone did their PhD at is higher ranked than their AP placement.
5. "Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?" You could say this is what post-docs are. You could sort of say this is what PhD students are. You could also say "yeah, but why do the top universities bother to have Assistant Professors at all, rather than waiting to see who becomes a star and then hiring them?" The latter is an interesting question (especially given that a large share of TT faculty at top schools don't get tenure). My guess is part of it is historical tradition, and part of it is that faculty will tend to show a decent amount of loyalty to their current employer, so you get an "unravelling equilibrium" in that the most promising candidates are poached early, because it's much harder to poach later on (who actually wants to upend their entire family life to move city for a modest pay bump at a slightly more prestigious university, especially if they are already successful. Just get current university to match the pay bump!). Part of it might also be that senior faculty like having productive junior faculty around. And another part might be that a lot of the most influential research is produced by faculty while they are fairly young, and the departments might care about this (even if it's too inside baseball for it to influence "prestige" in the public's mind).
I think in modern academia the incentives are so misaligned that it's hard to really model it in normal economic terms.
Students are mainly there for signaling purposes, not to actually learn stuff. And the tuition is being paid for out of subsidized loans. But also, the institution is nominally a nonprofit, in many cases literally run by the state. But also, the departments are usually in charge of hiring. The main department stakeholders are the tenured faculty, not the people nominally in charge of the university as a whole.
IMO this makes academia really susceptible to fads and trends. There's not really any strong competitive pressure to impose discipline. I don't necessarily think there's any "rational" reason for it to work how it does. There are definitely *some* structural elements in play, like budget cuts. But the structural economic elements dont on their own determine all the specifics that develop in a relatively small and close-knit community.
this is a huge topic but there's a few things that make easy analysis different.
1) as many have pointed out already, STEM is very different from non-STEM. I'd further subdivide things - fields have really different hiring patterns, adjunct use, etc. and that has a big impact on the way the market works. Comparing the market for a law professor to those for an English professor to an engineering professor to a chemistry professor is at least several kinds of apples, and probably a mixture of fruits and vegetables. Then there are multidisciplinary things (like public policy schools that hire economists, sociologists, political scientists, etc.) where things are even weirder b/c multiple strands come together in a weird mix.
2) schools differ a lot - someone has already noted the difference between R1 and others, but it is even more differentiated - the very top private schools behave very differently in hiring than do large state flagship schools even if both are R1s. There are whole separate "teaching institution" markets, the 2nd tier (which no one would admit to being in) research schools, community colleges, non-flagship state schools, liberal arts schools, and so on. Each submarket operates differently.
3) put those two things together and you have a large number of submarkets that operate very differently in hiring, use of adjuncts, etc.
Then we get in to allocating slots for TT faculty at a university. Some disciplines are going up in enrollment (Data Science); some teach huge "service" courses (English) rather than / in addition to large numbers of majors; some are going down (Classics). When a university hires a TT faculty member, that's a long term commitment to funding a slot. Some slots are cheap (English), some are really expensive (STEM). Administrators want flexibility in adjusting staff - which leads to more reliance on adjuncts and/or "APT" or "Academic Professional Track" professors, who generally don't write much and teach more, serve on committees. These folks get contracts from a year to longer. But they allow adjusting head counts. And most university central administrators know the demographic cliff is coming in terms of students for many places and want to keep flexibility to adapt if their institution's enrollment drops (as it will for many schools).
And then there's the need to bring in adjuncts to provide "real world" perspectives. Not much of a need in, say, English for that, but there is in law, business, politics (sometimes), etc.
There's more too - the academic job market is really a lot of submarkets which behave differently. Many of them are quirky at best, dysfunctional at worst. But you have to look at the institutional details of each to tease out why they behave the way they do (at least to some extent).
If I'm reading that graph correctly then it absolutely does *not* reflect my field (astronomy). Almost no one gets a tenure-track job within one year of earning their PhD. Nearly everyone does a postdoc, which in the US usually lasts 3 years. I don't have data in hand but I would guess that the median tenure-track professor in astronomy is somewhere around 3-5 years after getting their PhD. I think physics is very similar. Is this really that unusual?
One advantage may be that Universities are both consumers and producers of new PHDs. If the only way to get into Tenure Track was to suffer through years of adjunct teaching, would the PHD > Tenure Track career route be as popular?
As an outsider it seems that people going for PHDs in the hopes of staying in Academia are playing the odds that they are the lucky few that get a first year Tenure track posting, and if not either leave Academia or are stuck as an Adjunct.
From the University's perspective, the value of not shifting senior adjuncts to tenure track is that the tenure carrot can stay open to as many hopefully PHD students as possible, many of whom will end up spending years on a PHD with no tenure track to show for it. At that point some have no better option than to teach under whatever terms they can get.
I'm a former theoretical physics postdoc in the UK. My wife is currently a postdoc in biosciences at a top UK University. Here is my two pence:
1) Einstein didn't work in a lab. I get the point, but I had to say it! Sorry.
2) In my experience universities don't usually differentiate their workforce like that. Rather, most experienced people get forced to teach, early career people not so much, and only the very few (maybe <10%) get to avoid it (usually not by negotiating with universities but rather by getting grants which bring money and prestige and happen to force them to spend their time not teaching). And usually the more experienced and well-paid you are the less time you get to spend doing research rather than writing grants, supervising, and teaching.
3) Having talked to people on the inside during hiring processes, I don't think it is accurate to describe hiring people out of their PhDs as trying to find the people with the highest chances of being the next superstar. People usually try to screen for some measure of quality, but typically if you meet some minimum requirements (like having published in decent journals and/or having graduated from decent institutions) then what they try to optimise for is complementarity of research interests. Basically, they care more about how your work will fit with the work of other people in their groups than about whether you are a bit more likely to be the next Einstein. If you would be a good collaborator, who has a track record in lines of investigation which don't overlap too much with the ones they have going but which have potential for interacting with them, then they'll want you. (Bonus points if your niche area of expertise happens to be hot at the moment, where "hotness" often lasts less than the average PhD.) Which makes sense. Why hope for Einstein when you can more reliably aim for a multiplicative impact now?
4) Even the people in the dream superstar jobs are usually earning not that much compared to industry - or even non-academic techy jobs in the public sector.
I don't have data on this, but there's a similar phenomenon in the publishing world. A debut author with a book that sounds exciting and topical will have a decent shot at a six-figure book deal, and those books frequently go to auction. But once you're published, if you're among the majority of authors whose books don't turn into juggernauts, your advances drop dramatically, with some previously well-paid authors grateful to get $20k or $30k for a new project.
This is a "shiny new thing" phenomenon - an exciting new author has enormous potential... until their book hits shelves and they don't hit bestseller lists. My initial guess is that hiring committees are susceptible to the appeal of the "shiny new thing" and get so excited by the boundless potential of all the newly-minted PhDs that the adjuncts who've been slogging for years now just don't stand a chance.
After a couple of disappointing books, an author will often change genre or age category, choose a pen name to publish under, and go to a new publisher to leave behind the stink of their sales record. This can be a way to reboot their career, re-invigorate the marketing department (who will have given up on them at their previous publisher), and try to capture some of that "shiny new thing" energy again. It's interesting you mentioned that successful teaching- or adjunct-track professors will go somewhere else for a tenure-track position when they make a breakthrough, and feels related to this.
A little concerning that an industry like publishing that acquires talent almost entirely on *vibes and sparkle* appears to have similar patterns to academia!
As someone who's spent ages 15-41 in academia: you're *very* far from wrong, but it's also not as drastic as someone reading this might fear. Academia isn't sharply divided between "Einsteins" who teach only occasional upper-level seminars, and poorly-paid adjuncts who teach huge numbers of intro classes. There's also a large "middle class," e.g. all the non-Einstein tenure-track faculty, which both does research AND interacts regularly with undergrads, sometimes even at the same time (at least in CS, lots of undergrads get involved in research). Maybe that just helps confuse prospective students all the more! :-)
(Also, agree with the other commenters that postdocs are now a huge part of the story.)
Sports team analogy is probably even more apt here than Scott realizes, especially salary capped and contractually strict ones like the NBA. Just like in academia, in the NBA a rookie who has the chance to develop into the superstar is far more valuable than an established veteran who has proven themselves to be a reliable above-average player. This is true even if, on average, the rookies do not become above-average players. This is reflected on the trade market, where the #5 pick in the draft (which becomes a first-year player in the league once drafted) is valued more highly in a vacuum than solid rotation veterans, even if "solid rotation veteran" is actually probably the EV of what you get from the #5 pick.
There are additional wrinkles and dimensions to this, as rookies are also on a more affordable rookie scale contract, NBA players are locked up into defined contract lengths than is difficult for the player to exit early for other teams, etc. The limited free agency explains superstar inertia in professional sports - there are typically only a few defined windows in a player's career where they truly are pure free agents. But "superstar inertia" I don't think is very difficult to explain for most tenure-track academics. Unless you are in the true 1% of academia there's a lot of disruption to uprooting your life and your intellectual networks to go somewhere use and the upside needs to be very high to entice you to do so. Given the natural geographic distribution of academic employers it's not even like other working professionals job-hopping to go to the competitor firm down the street. And it seems to me that you do see the true top 1% of tenure track academia, celebrity public intellectuals etc., move around, though perhaps my perception is shaded by the fact that these moves are naturally going to be more newsworthy than random tenure track faculty moves.
Some of what's going on is historical, going back to about 1950, when the Feds started throwing money at universities to do research. The NSF was founded c. 1952 b/c everyone realized that science had won WW II thus needed serious funding. The question was HOW, and we got a bastard version of the German system and our old tradition of colleges being for educating elite youth. Remember that many of the important scientists of WW II were German. There, all universities were research universities--there was no UG education as such--and their goal was to produce research and prepare scholars/scientists. They were supported by the German state, and departments were given budgets they could spend as they chose (and there was, basically, just 1 fully tenured "Ordinary Professor" per department, and he ran it, often like a fief. That was what US scientists envisioned after the war--fund the best scientists directly to produce the best science. But this met howls of protest from most US colleges & universities who feared all the money would go to the usual suspects, Harvard, Yale, Illinois (go Illini!), etc. And it was unAmerican. States urged the money go to states for allocation to their schools. The scientists howled -- nobodies at places like the Southern University of North Dakota at Hoople would get, and waste, scientific research dollars. Bad Science would be done!! The compromise is what we have today is the result, anyone, even a nobody at SND@H, could submit a grant, and if it was, in fact Best Science, it would be funded. So, virtually every school in the US tries to be a research oriented school, not just for the prestige, but because of the sweet overhead money that comes in, helping fund student-attracting facilities like salt water pools and paying big money to sports coaches.
The adjunct vs. research professor divide reflects the old teaching-college vs. research institution divide in the US. If you read what the poor adjuncts write, they generally thought of themselves as teachers expanding and informing the minds of their students and did not want to be researchers. They suffer for pursuing their dream. Between that and high course loads, they don't do research and so can't transition to the main tenure track. They'd like to be paid well and have job security to do what they love, teach.
Top stars get hired (note rise in numbers for older faculty) when a university wants to set up a new field of study or some "center." Then they go for an established name who will attract bright and ambitious folks quickly.
I'm an odd duck in this case. I wanted to be an UG teacher back in 1974 (my mentor said, "I suspected that of you."), before the Federal money started streaming to my field, psychology. But there were no good jobs available then, so I took one at a comprehensive university that then got ambitious and shot up to R1 status remarkably quickly. So I had to get with the program and do research, and did get tenure, though I could not today, as I never got grants.
"colleges have decided they only need a certain amount of prestige before they stop caring and fill the other teaching positions with warm bodies" - No, it's more that they buy as much prestige as they can afford, but (in part because its hard to hire good research faculty while requiring them to do lots of teaching) that isn't enough in order to meet all their teaching obligations.
"Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?" - This is what the best universities do - they hire "tenure track" professors, but very rarely grant them tenure - instead, their tenured faculty are maximally prestigious people poached from elsewhere. This works for the best universities, because being tenure track at Harvard (with the expectation that you probably won't actually get tenure) is a very good deal for the very best new PhDs. It doesn't work at other universities because being tenure track at Podunk College, with the expectation that you won't even get tenure out of it, is not an attractive offer, and so good people don't accept offers at lesser schools if they expect it won't lead to tenure (assuming they have a better option elsewhere).
Note that Devereaux is writing about the *history* job market, not all of this applies to academia in general. Like, I've never heard of the whole "experience actually hurts" phenomenon in mathematics.
Also worth noting is that in math and various sciences, it's typical to do a few postdocs (which doesn't fit into any of Devereaux's categories; these are neither tenure track nor teaching track nor adjunct) before getting a tenure track position. Devereaux mentions this but *mostly* ignores it because he's mostly focusing on history.
So I don't know if the market is as broken as in other fields.
As for history, though -- I do think you may be making a mistake in assuming that people are necessarily doing things for good reasons rather than due to biases. Devereaux explains the "experience hurts" phenomenon via a bias for hiring from prestigious colleges; I think that answer is definitely worth considering.
I think academia and performing arts have a lot more in common than you'd think. A record label and an academic department both want to sign someone that will HIT. But you have no actual idea what prospects will do it. For recording artists, you can give a thousand advances and if even one hits it will pay for the rest. But you can't do that for academic departments. So you make choices based on what you can actually observe (a degree from a top 10 program) and hope that will give you a better chance at a hit.
Why do departments only hire people with no experience? The same reason that there are no 35-year-old breakout pop stars. Labels and departments buy based on POTENTIAL. Someone with experience has no upside.
> What process carves off 30% of professors to get good pay and benefits, but passes over the rest?
You're neglecting the influence of unions and worker activism. The institution, like most employers, is coordinated and can apply power effectively with ease. It has most of the power. Workers can only exert pressure by leaving (and going to another institution that is likely similar) or striking (difficult to coordinate, pisses people off, risky, etc.).
Often, the low-paid teaching staff are contractors, have little clout in the union if they're in it at all, and may well be gone in a year or two. The union is incentivized to do the most for its long-term, prestigious members, and so it uses its power mostly for them.
You end up with a scenario where the uni would like to pay everyone crap, the workers would like everyone paid well, and the way the power imbalance/application of political capital shakes out is that some small amount of workers are paid well and the rest are paid crap.
I don't think your explanation about trying to trick students about who is teaching them works, or at least it's not complete. Disclaimer that this is based on my anecdotal experience in astronomy.
Teaching professors are a relatively new phenomenon but colleges have existed and have been teaching students with something like the current model in the US for decades. Why would they only now start trying to offload teaching responsibilities onto a different class of professor behind students' backs? Also, the teaching professors that I'm familiar actually have the title "teaching professor" right there on the department webpage, so if they are trying to hide this from students they are doing a poor job.
Also, in my field, adjuncts are people who have a primary appointment elsewhere but who then get a secondary affiliation with some college or university, usually so they can directly advise grad students, and sometimes as a way of funneling research funding through the university. They aren't paid positions and don't come with any teaching or department service responsibilities. I think one needs to be careful about differentiating between very different positions with the same or similar titles when doing this type of analysis.
One element you might be missing is the nepotism angle. My partner was a an academic who recently left academia, and one thing they claim is an open secret in academia is that most new tenure track hires have parents who were tenure track hires before them. There could be reasons other than direct nepotism that account for this (the children of tenure track people are more likely to go into the field and be suited for it, and also they know the most common pitfalls to navigate), but however you parse it, it makes sense to me that in a system that has a nepotism component, you see a lot of people being snatched for high-prestige jobs the moment they become eligible for those jobs, not the moment they have a proven track record (because nepotism generally works counter to meritocracy).
It would be interesting to do a deep comparison of the academic and entertainment industries, because entertainment is a famously nepotistic industry where lots of people toil in competition for low-paying, shitty gigs and a few people rise to the top, and those people are often related to other people in the industry.
"Why only people from outside their own institution?"
There's a strong social tradition of avoiding 'academic inbreeding.' Hiring your own graduates is seen as both an obvious form of nepotism and a recipe for intellectual stagnation. Not that it doesn't happen, when a favored student can't find a job.
Adjuncts are de facto on another planet, and tt assistant professors are by design junior, i.e. you typically hire them immediately after their PhD or a few years of postdoc. In many countries, only associate and full professors are public servants like judges and diplomats, i.e. not on a contract.
Teaching is incidental, and can easily be done by graduate students or postdocs. What matters is research, grant funding, and prestige. In my experience, poaching exists but up to a point, especially in Europe where there is less room for negotiation.
Hypothetically, a person doing a job doesn't look like they can do more challenging work, at least if management is sufficiently unimaginative. So they prefer shiny new people.
This seems to import the assumption that "Einstein/superstar" researchers are a thing outside of STEM. I don't know if this is true -- I'm STEM myself -- but it strikes me as counterintuitive in a field like ancient history (Bret's field) where research is driven more by laborious study than clever ideas. What is an Einstein-equivalent in history or English literature *doing* to be so valuable?
How do you discover someone is a good researcher while 1) soaking up their time teaching and 2) not giving them any significant field-appropriate resources for researching?
I think the explanation mostly is institutional capture of the hiring methods and standards by the tenure-track profs. They were hired into the prestigious TT in this manner, and they do the hiring/control the hiring of TT positions, and if the methods or standards are questioned that would question their entitlement to the prestigious TT jobs they have.
Scott: Your question "Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?" has an easy answer.
Once you're hired as an adjunct you really have no hope (except in the most unusual non-STEM circumstances) of getting a good TT job. You have to start producing amazing research/scholarship almost immediately, and the adjunct grind just makes this impossible.
I am a non-tenure track research professor (meaning I don't teach classes) at an R1 institution and have been an adjunct professor at a different college. I can share some insights into hiring decisions.
There is a bias toward hiring young people for tenure track positions for several reasons:
1. They are trying to hire people who will be making a name for themselves at that institution. Hiring committees do look at a persons career trajectory in addition to accomplishments.
2. Faculty want to make sure their department is in on whatever the current trend is in order to appeal to grad students and postdocs. Naturally, the best way to buy in to such trends is to hire young people.
3. There is an extreme level of concern about diversity. Since diversity is increasing, the most diverse subset of candidates is the youngest group. This leads to a strong bias toward hiring younger people to faculty positions. (This is probably one of the most important factors and I have known cases where it was explicitly stated that no white men would be hired. That clearly ruled out a lot of the more established candidates.)
Some of this post also assumes some popular stereotypes that are not true. One is that research and teaching are incompatible. Almost all the time a person who is good at research is also an excellent teacher. Besides which, students like having a professor who is active in their field. Researchers generally like teaching, in spite of the stereotype that researchers wish they could avoid teaching so they could focus on their research.
Universities hire adjuncts because they're cheap labor, not because they're expert teachers. Don't read too much into it. Teaching at the college level is really not difficult for someone who understands the subject. I think Scott has a post somewhere showing that most innovations in education are just showing selection effects. Faculty know this, even if they do not usually admit it. So, there is not much value to be gained from having people who are hired for teaching abilities.
A mildly snarky take on not offering TT to adjuncts in-house:
As a successful academic on the hiring committee you've successfully jumped through a large number of hoops, starting in high school. If you've hired someone as an adjunct already, your department's process has confirmed this person as a 2nd tier researcher. Giving them a TT position would imply a flaw in your hiring process (if they're good enough to be TT, why would they take an adjunct position in the first place?), the same hiring process which has confirmed you as a successful academic & we can't just go around calling that into question, now can we?
I'm a department chair at a state university (with unionized faculty), and have experience with both tenure-line and adjunct hiring.
When we hire a new adjunct, the pool is generally limited to people who are already in the area and open to more employment. Usually, in an adjunct's first semester, we can only offer them part time work. This is partly because the collective bargaining agreement requires that we offer classes to any qualified current adjuncts before hiring someone new. For example, if someone retires, gets sick, or goes on sabbatical, I have to offer their courses to any qualified current part-timers who would like to increase their teaching load. There might be a course or two left over for a new adjunct to start part-time. It's essentially impossible to offer a full-time position for a new adjunct.
In this situation, we can't cast a wide net for candidates. We aren't offering a level of compensation or security worth relocating for, and in many cases the need arises last-minute. So, we hire the most qualified person we can find locally, among people with the right credentials who neither got a tenure-track nor took a full-time job in industry. It's often difficult to find anyone.
For a tenure-track hire, by contrast, we get a large number of applications from across the country and internationally. If one of our current adjuncts is in the pool, they will get the same consideration as everyone else. But the odds that they will happen to be the best fit out of that large pool are very low.
An important point to consider is that in many fields (e.g. economics) adjuncts typically do not have PhDs, only masters degrees. As such they cannot move into tenure track or teaching track jobs. This is often raised as a point of contention, that adjuncts are not considered for such jobs, but they are also explicitly unqualified for the job. Adjuncts in these fields are in effect below even PhD candidates, as they are not even working on their dissertations, and more like regular grad students.
That isn't to say anything bad about adjuncts, just that simply going by job titles isn't putting like with like.
I should also point out that when I was in the job market it was common advice when applying for a "normal" tenure track economics position to not even admit that you like teaching. I use admit intentionally there, as it was seen as an admission that you were not serious about research. This is important, as most schools wanted researchers and offered extremely low teaching loads (something like 1 course a semester, maybe 2), one even promising TA support such that "you don't even have to interact with students, ever." Accreditation is easier if you have more courses taught by PhDs, however, so depending on how it is counted (number of courses taught by PhDs vs percent of faculty with a PhD) teaching track roles are becoming more valuable to counter balance the number of research only faculty.
I have a couple of other thoughts here that have been mentioned before, so I won't belabor them, but one that seems to be left out: networking.
My professors have explicitely told me that at some level, whether I can land a post-doc is a luck of the draw for how many of their friends have recently taken post-docs. From undergrad I moved to a grad school in large part through my mentor's contacts, and the assumption is that from grad school I would move to a post-doc largely through my advisor's contacts. This is a pretty common story in math, which seems to be a particularly small and close-knit community as opposed to many other STEM fields. While I am in graduate school, and for a year or two out, my advisors will be concerned with my success, as it reflects on them. After a few years, they will have other students to go to bat for, and I imagine that they would reasonably focus their efforts on landing their current students' jobs. The longer I'm out, the less contact I have with my advisors, and thus the harder I expect it will be for me to land a job.
DISCLAIMER: I think a lot of merit goes into getting grad school/post docs/tenure track jobs. I just also think that there's a lot of networking in there too; you sort of need both, or overwhelming amounts of merit.
I think Scott's basic explanation is correct. I'll just add tweak that I thinks clarifies the matter further.
As Scott notes, there are two distinct functions: research and teaching. In a model of perfect competition, salary (Y) would form a power law when plotted against the ratio of research-to-teaching (X), with the long tail extending toward 100% teaching. However, there is not perfect competition, because bureaucracy. For budgetary reasons, hiring and salary must be legible to administrators, so they segment the curve into distinct roles: tenure-track, teaching, and adjunct, resulting in the present system.
The power law in competition for researchers is also congruent with the preference for superstar researchers and the disfavoring of older adjuncts. But it's the bureaucratic dimension that prevents this from being rationalized into a system of free agents.
My experience is in the humanities, and it's also about 30 years out of date now. I had a tenure-track job right after I finished my dissertation. My husband was a lecturer for two years before getting a tenure-track gig at a different school. (There's another distinction for you -- lecturers make a real salary and get bennies, but they're one-year renewable hires. So better than adjuncts, but worse than TT faculty).
My observation is that all the universities assume that the best candidates will get snatched up right after they graduate. If they didn't get snatched up, well, then I guess they weren't the best candidates, right? Almost anyone you ask in academia in a one-on-one conversation will agree this is stupid and incorrect, but in committee meetings it's kind of the default assumption.
I'm female, and I also felt like there was a similarity between fresh, virginal candidates/fresh, virginal young women and slightly used, been-around-the-block candidates/been-around-the-block less-young women. Do I have any social science to back this up? Nope. It was just a feeling I had.
I eventually left academia so that I could live in the same state as my husband (who also left academia), and we both got higher-paid, lower-status jobs (lawyer and IT manager). In the same town! Mirabile dictu.
I guess I should also mention that humanities folks don't usually do post-docs, though there are exceptions.
I think you should question your approach of asking why colleges don't do the thing that would be more fair, rather than asking why any organization ever does the fair thing. No explanation is needed for colleges behaving unfairly. An explanation /is/ needed when one sees an organization behaving fairly.
As Adam Smith said, What needs explaining is not why so many nations are poor, but why some are rich.
Colleges aren't for-profit corporations, and my experience with organizations of all types has been that organizations that aren't run for profit, don't incentivize their people to be fair, or to do their jobs well in any respect. The only exception I've encountered are religious organizations, volunteer organizations, and Dept. of Defense front-line organizations, which all attract only people who are already incentivized to do the mission they're signing up for.
(You might think that the non-random behavior of colleges shows they're incentivized in some way. That's correct; but when white-collar workers aren't incentivized to do their job correctly, their personal preferences take over. These are highly correlated between different people in the same job, even in different organizations; so the behavior across organizations has a similar distribution.)
Substack tells me this comment is too long, so I will be splitting it up into two; the second half is in the reply. (Edited to remove some potentially identifying info and to soften some things I think I was too harsh on at first.
I was recently on the CS academic job market and successfully got a job. A few thoughts -- this is mostly focused on the top 30 CS departments in the US. Once you get out of there I don't know how much this holds anymore.
1. First of all, your caption on the graph doesn't actually match the chart. Assuming the chart label is right, that's not acceptance rate of people being hired, that's the *number* of people being hired at each rank. (Yes, these numbers are low, which matches my expectations for a specific field. See the annual Taubee survey for CS hiring stats in the US and Canada -- it's typical to hire about 200-400 professors in the US per year: https://cra.org/resources/taulbee-survey/ (last available for year 2021). I would be shocked if the acceptance rate for recent PhDs is 90 and the acceptance rate for people 1 year out (including postdocs, people who did a year of research prof, etc) was so much lower -- but the graph is just raw numbers as far as I can tell. The fact that the "Within Previous Year" bar is so high does not indicate that the acceptance rate for recent PhDs is higher. It just indicates that more of them were hired, which is not surprising considering it's by far the most common time to go on the job market.
2. Speaking about top R1 schools (let's say the top 30 computer science departments -- and everything I say here counts double for the top 10): The reason it becomes much harder to get a tenure-track job after becoming a teaching-focused professor (either teaching-track or adjunct) is that the application for tenure-track jobs is almost exclusively focused on your research. Aside from the allure of the "full package," I'm not sure why the charade is maintained that research is just one of three equally-weighted considerations in a tenure-track job (research, teaching, service). In reality it seems to be about 95% research, and as long as you don't stab a student or something you'll probably keep your job. (I will say: some places, including the place I ended up going, did actually have one teaching faculty interview me as a tenure-track faculty. This is rare and a good sign that this institution/department actually treats its teaching faculty well. But I think aside from that, I got asked exactly zero questions about teaching at R1 interviews. Yes, I had to provide a teaching statement that described my teaching philosophy. Did anyone actually read it? I have no idea, but I know they read my research statement, because I got questions about that.)
And if you are a PhD student or a postdoc, your full-time job is research. Most PhD students and postdocs don't work a day job, have only minor teaching responsibilities, and have collaborators and peers who are also doing research. Plus they (theoretically) have guidance on which well-scoped problems to work on from their advisors. So, PhD students and postdocs are basically in the best state to spend the 5+ years of run-up to a tenure track application. Doing research as a PhD is way easier than doing it on your own time as a hobby, if only because your university will pay to get you access to all the electronic library resources you could want, lab space, equipment, administrators, conference travel, seminars to watch, administrators, etc, plus you don't have to hold down another job while doing that. This is the biggest pool of people who want to become tenure-track professors, so this is everyone else is competing against.
Teaching as a teaching-track professor or adjunct is a full-time job. My understanding is that a typical teaching-track load is 2-3 classes per semester. The advice I've been given is that it takes 4-8 out-of-lecture hours to prepare materials (lecture, homework, "fun" activities, etc) per every hour of lecture -- not even including grading, office hours, helping students who want to chat, dealing with students who decided they hate you, dealing with cheating, paperwork, interfacing with the rest of the department, etc. As far as I can tell, it is one of the hardest jobs ever, and it doesn't pay nearly as well or have the prestige of being a research professor. But, if you want to be tenure-track faculty, you will be evaluated on very little of this, and realistically if you're teaching 2-3 classes per semester you will not have time to conduct the kind of research agenda that will help you compete with tenure-track candidates. The competition is the PhD students and postdocs (and maybe research professors) who *have* been spending full-time jobs on research, so even if you do a great job in the <25% of your day remaining to you after teaching, that's a *really* uphill battle. I've spoken with several teaching faculty and asked why they did that, and I got two reasons. Almost all of them: they love teaching! Which is great! One person said he liked academia and was thinking about the tenure-track life, but didn't want to be busy with the 20 jobs of being a tenure-track professor, so he went with a teaching job so he could be overloaded with only 10 jobs instead. I didn't talk to any adjuncts because I don't know any, which probably says something about how much they are overlooked in academia in general.
Maybe a more charitable way to look at this is: Research, most commonly done during a PhD/postdoc, *is* the experience they're looking for. Everything else, including teaching, is a "distraction." (I don't honestly know if it would be better if everyone gave up on the charade and had tenure-track faculty just stop teaching. More on that in another post later, perhaps.) And people who do research-focused jobs, especially in research-oriented academic roles (see next), or occasionally from particularly paper-publishing-focused industry labs, do occasionally switch into tenure track.
On the point of not hiring from your own institution I have seen one senior professor in economics claim they would do this specifically for purposes of diversity in intellectual thought. They argued that on principle their doctoral students should get experience somewhere else and that they should try and get people from elsewhere so that the intellectual climate didn't get stale. No idea if this if this value system is strong enough to override more base incentives, seems unlikely to me, but could be part of an explanation.
As a STEM PhD, in a very qualitative way I feel like I can tell who will be hired and who won’t pretty easily. The truth is you have to be exceptional to be hired in my field (not bio) and it’s easy to tell which phd grads will be chosen based on how many papers they published and who they worked with in graduate school. They are the real go getters and they are zealous.
Hi Scott and thanks for this post. I'm an associate professor at a European R1-like university and I think there are relatively straightforward economic explanations as follows (with tounge partly in cheek).
As an academic in a research position, my main job is not to do good research, nor to teach classes. It is to bring home research grants. Overheads from these grants keep the lights on for everyone else. Therefore, the university needs to try to give tenured research positions only to people that can be expected with high confidence to earn more than they cost.
Next part of the explanation: a successful grantmaking career is a trajectory that needs to start almost immediately after getting a PhD. I am not sure whether it is rational for funders to allocate funds in this manner, but it is broadly perceived that this is how it works: more money comes to those that already have it, and if you don't get on the escalator from the start, it quickly gets much more difficult to climb up.
You ask why universities do not hire lots of people in a low-commitment position. But in a sense this is what the tenure track is. The salary is not high and you get kicked out after a few years if you haven't established a track record of external funding. Low risk with high upside for the university.
You also ask why colleges aren't more like sports teams, trying to buy each other's stars. This does happen to an extent. But a closer analogy might be to senior partners in law firms or management consultancies. If conditions are better elsewhere they may leave, but transaction costs are high, and most of the money anyway comes from what they bring in on their own.
Thus, I think the economics of a dependency on external grants goes a long way towards explaining the phenomena you describe.
I'd point out that the incentives at top institutions to care about superstar researchers run deeper than prestige. Top-tier profs pull in grant money, in many cases millions of dollars for departments (indirect costs on grants, i.e., the percentage of a total research grant that a uni takes in for its own operations rather than a prof's research is around 50%). The difference between A tier and B/C tier profs in this ability to pull in grant money is likely substantial in many fields that are heavily dependent on research funding given the many biases present in professor brand name recognition, exponential growth dynamics in citation counts, and knowing how to work the grant system, let alone differences in actual talent (which from my observations are substantial).
"But why aren’t colleges more like un-salary-capped sports teams, trying to outbid one another for their rivals’ superstars?"
Institutions do this (we always heard that many top-tier Stanford profs were paid in the low to mid six-figures). However, departmental research budgets are finite with high potential upside but definitely not startup go-to-the-moon upside (how much grant money can one prof pull in? How can you predict who will win a Nobel just from a track record early in someone's career? patent potential is rarer) Switching universities is a huge move and as always, you have reputational risks if you e.g. move around too frequently and lose research productivity. What motivates academics is mostly not money, but freedom to do research. This means freedom from teaching/departmental responsibilities as well as the ability to collaborate closely with other leaders in your field (different schools have different specialties and you may not be able to find similar collaboration opportunities at peer institutions no matter what they pay you). Academics (in the hard sciences/engineering fields at least) could generally get paid way more out in private industry.
There's only so much grant money to win. It could be an interesting question whether unis are optimized for winning grant money based on faculty counts (whether they're over or under hiring)
Some hires appear to simply be political appointments, like Noam Chomsky at University of Arizona. I know they never replaced Ed Abbey. But he was the eco terrorists' darling. Really? Noam Chomsky for Ed Abbey? I don't get it.
Chomsky merely mumbles his way through his monologues; Abbey had a goal in socializing, you might say.
Yitang Zhang is a fascinating case: he failed to get a tenure-track job because his early work had no real successes, he bounced around as a lecturer (and even fell out of academia for a while), and then he produced a monumental advance on the twin primes conjecture and was instantly given tenure at his university (and naturally, he was poached the next year). https://en.wikipedia.org/wiki/Yitang_Zhang
I’m a tenure track faculty member at an Ivy. From my experience, I think this analysis gets almost everything right, except:
1) take Scott’s observation that every hour spent in the classroom is an hour not spent doing research together with the observation that the greatest research contributions tend to come earlier in people’s careers, and it’s no big mystery why universities don’t hire everyone as adjuncts and wait for the superstar researchers to rise to the top — if everyone had to teach a moderate amount, those future superstars wouldn’t have enough time to do the research that would make them superstars and everyone would be less productive as a whole.
2) I don’t think students are really clamoring for famous thought leaders to teach them. With maybe a few exceptions at the very top (eg. Steve Pinker), few undergrads have any idea who the field-famous research stars at their university are and almost none of them know when they are being taught by a research star and when they’re being taught by an adjunct who isn’t doing much research at all. That might leave us without an explanation for why the titles are so confusingly similar, but I’m not sure that needs much special explanation. All titles are pretty confusing to outsiders. How many people know the difference between the Dean of the College and the Dean of Student Affairs? I’m not sure they’re less confusing outside of academia, either. Is the difference between a Software Architect and a Software Enginneer really less opaque?
This looks like a multi-armed bandit solution. If the variance in outcomes among inexperienced hires is greater than the variance in outcomes among experienced hires, and you are trying to maximize your likelihood of making a great hire, you're going to make a substantial number of offers to untested but possibly great new hires. If they don't turn out to be great, you can decline tenure later.
I don't think it's a mystery why the job market (in humanities) penalises experience. It's a bit like the dating market, desirable candidates get snapped up quickly, and if you're single for five years straight there's probably something wrong with you.
It's only slightly different in the sciences, where there's an expectation that you need to be a postdoc for a few years. This really just means the clock starts a few years later.
Universities hire based mostly on prestige rather than ability. Whether that's because prestige is an easy-to-measure proxy for ability or whether they genuinely want prestige more than ability, that depends on how cynical you are. But you don't become more prestigious just by sitting around on the shelf. There's a chance you might, in that extra year, do something that makes you extra-prestigious, but it will need to be something pretty special.
> But why aren’t colleges more like un-salary-capped sports teams, trying to outbid one another for their rivals’ superstars?
Because there's not enough benefit to doing so which would offset the huge cost of a bidding war. The fifth best basketball team can become the best basketball team by hiring the best player in the world, but no single hire or handful of hires is going to loft U Chicago above Harvard. Bidding wars cost a lot of money, and they also piss off your existing staff who are annoyed that Professor Big Shot down the hall is suddenly earning four times more than them without doing anything meaningfully better.
This is what I’ve been told: universities like to hire young PhDs for tenure-track because they want to make sure you can be productive for as many years as possible into the future. If you’ve had a PhD for 6 years already, the university fears that’s 6 less years of productive output and once you get tenured, the university is basically stuck with you. That’s what I’ve been told by profs who are / have been on hiring committees.
acoup is generally on point about the structure of the academic profession. Basically, much of teaching is done by specialized teaching faculty who get little money and are looked down upon by "research" faculty.
With that said, we have major disagreements.
First, I don’t think universities can easily save a lot of money. Universities spend huge sums on administration not due to incompetence but because all stakeholders are demanding more IT support, more support for student mental health, more support for finding jobs, etc. There is spending I see as clearly wasteful, but is generally spending that pays for itself – fancy buildings requested by donors, luxury dorms that attract out of state rich kids who pay the bills, conditional government grants, etc.
Second, it isn’t clear to me that spending more on adjuncts is the most compelling priority for any money that can be freed up. Adjuncts are highly educated adults repeatedly choosing to take crappy wages in exchange for a nominally cool job and a chance of a better job. Compare that to, for example, undergraduate students who are often literally children when they commit to spending in the low six-figures on a college education that is now a barrier to entry for a middle-classhood. I'm not sure I'd spend extra money on the adjuncts over just reducing tuition.
Full disclosure – highly remunerated tenure track faculty at R1.
1) Grants. An Einstein doesn't just bring in prestige, they bring in millions of dollars of funding, which can then be used to hire more people. As a result, there's pressure to find people who have markers of future grant success. But grant funders are cautious, they want to make sure that their millions of dollars are going to be put to good use (which is exactly the kind of strategy that en masse makes research worse, as moonshots almost never get funded), so another vicious cycle develops where they preferentially fund people who have already had grant success, or certainly look like they will. This then interacts with....
2) Hyperspecialisation and illegibility. Grant decisions are often made by non experts. Job searches often feature someone from management or outside the department. Even within departments it can be hard to assess the quality of someone's work if it's not your exact area. As a result other markers (E.g. outputs in ranked journals) become the lowest common denominator. So if youre a few years post PhD and don't have these markers, it makes you look much less likely to have future success compared with the newly minted PhD who already has a few but hasn't had much opportunity to fail yet either
3) Relatedly, the pressures for more presige, publications, and funds point towards 'collaboration' and interdisciplinarity - it's easier to tell a future grant success story if you can say you've teamed up with Einstein at another university, and universities that work together can feed off each other's prestige, and if you add your name to their paper while they add theirs to yours you can get more outputs. So new hires who have connections with other instutitions, or whose supervisor was Einstein, can be a genuine asset, and have an advantage over someone local who only knows everyone you know.
4) It's also important to remember that the Adjunct market isn't normal. Departments preferentially give jobs to their current grad students so they get the experience. As you finish your PhD, there's a new batch of grad students needing work, so it gets harder and harder to justify giving the job to you even if you are more qualified. This means that as you get further out adjuncts often have to find non-academic work which makes it difficult to research and get those other markers.
Regarding why there's some poaching but not a huge amount: the people are involved are all quite intelligent, and can recognize Moloch traps, even if they're not entirely immune.
Overall, the academic job market is like all the worst parts of asymmetric dating markets. I think literally you can just pick the worst dynamics from dating markets, and there'll be an academic equivalent, and vice versa.
On that note, maybe one of the reasons it hasn't gotten fixed is that you just don't talk about it. The impression I get is that it's airing dirty laundry in public, and erodes trust in the Academy at large. I've been in a discussion with some academic friends about Bret Devereaux's blog, and they were distinctly unenthusiastic, and I think one part of it was the dirty laundry aspect.
I don't get why my nice tenure-track job exists. I teach political science in a largely undergraduate program (no PhD students, sort-of-related Masters program) in an institution that excels in STEM. Nobody is expecting truly great research out of *anybody* in my liberal arts college. Anybody in it who does do great research moves on to a more prestigious university (especially for social science) in no time. So, why even have 30% of the faculty in my department (and college) on this much more expensive tenure track. The university's overall reputation is made by scientists and engineers. I think there's a lot of path dependence and a lot of prestige to publishing research even if almost all of it published in my college is... fine.
My wife is tenure track research faculty. Based on her experience, I would say the elephant in the room is grant funding.
Essentially, research faculty are as valuable as the grant money they can bring into the university. Universities get an additional 30-50% of the value of the grant paid directly to them for overhead costs (and there is little accountability for how the university chooses to spend that money). Also, research faculty are expected to fund a portion of their own salary from their grants. The percentage varies, but my wife funds 50% of her salary from her own grants, so as long as she has grants then the university only pays half. If you are research faculty with funds you are an asset. If you are research faculty without funds you are a burden. Anyone the university considers making tenured faculty needs to prove they will not be a long term burden.
The best NIH grant you can get is an R01. If a researcher has that, then your lab is funded securely for the next five years. Once you show you can write an R01 grant and get it, universities are willing to make you tenure track. If you write and get an R01 as part of your post-doc work, then you can apply to tenure track positions immediately and are likely to get one. Then all you have to do for tenure is get a second R01. However, if you finish your post-doc without an R01 then you need to find a non-tenured position and scrape by until you get a big grant.
In short, there are no superstars. There are professors with a track record of bringing money into their university and professors without that track record. From the university point of view, if your people keep getting grants it means they are probably rising to the top of their field. Or maybe a university could poach a superstar, but the actual value added is the money they bring with them. A 'superstar' that has no active grants is no longer one for the purpose if hireability.
Academia is basically one of the "dream careers", in that there are tons of people who dream of doing research and/or teaching at the college level, and so you have a huge field of folks fighting over limited space. It doesn't help that by the time they're looking for tenure track jobs, your professor folk have spent virtually all of their adult life (and a big fraction of their overall life) inside academia with its peculiar ways.
"Maybe colleges really do want “superstars”, not just moderately good researchers. The value of the #1 brightest new PhD is that she has a 5% chance of becoming a future superstar; the value of the #100 brightest new PhD is that she has a 1% chance of becoming a future superstar. Once you’ve been around for five years, colleges can see your track record, satisfy themselves you’re not the next Einstein, and lose interest."
Others have said this for specific other fields above, but my main reaction to this was just: that graph of decaying chances to get into good jobs looks *right*. It looks like the way the world should be if it's functioning properly. These are all PhDs, which means that the academic world has already kicked the tyres of these people and determined which ones are roadworthy. On the first pass, most of the good ones should get swept up into the good jobs. On the second and third passes, there are fewer good ones who escaped notice first time round, so there are few hires into the good jobs.
There are presumably lots of ways in which academia is messed up, but I don't see how that graph can be one of them.
My experience is that academia is not a closed system. It's competing with private industry. And industry will always have more openings and bigger budgets.
I did my master's at a research university. Not tier 1, but well regarded in a few specific fields. My lab was swarming with tech and defense companies looking for specialists. Tenured professors would consult with or at least exchange interns with these companies. For a fresh PhD, then, the BATNA wasn't settling for a different track or a lesser university. It's getting a six-figure job offer in Huntsville and leaving academia entirely.
If industry and academia have similar demand for fresh PhDs, but industry's demand grows faster with experience, I would expect industry to win. That could generate the weird curve that Bret showed.
a) a lot of the most groundbreaking research in an academic's career happens when they are relatively young. Since you mention Einstein, it's worth going to take a look at how long he lived vs. his age when his most famous theories were elaborated. So this means that, to paraphrase Pulp Fiction, academic's minds don't age like fine wine like they think, but instead slowly turn to vinegar. Hence older academics who haven't become famous are not very valuable.
b) Tenure committees are probably incapable in most cases of distinguishing between candidates based on objective quality criteria (the few true megastars everyone knows about and get hired quickly, leaving a bunch of grinders who are more than capable of doing job but don't quite have the "it" factor). In that context as you say hiring the young guy who less proven either way provides more upside, but conversely if people have been on job market for years without getting tenure anywhere that also becomes an easy way for the committee to ding them. Sort of like a house that for whatever reason sits too long on the market, you start wondering what's wrong with it even if you can't see any obvious issues.
c) am skeptical it's politically acceptable to start paying professors in a differentiated and market-reflective way. If professors with similar job titles make vastly different amounts of money that's going to create tons of jealousy/infighting, and if they get truly paid free agent type money the marxist types all too common among student body are going to be picketing the president's office to fire them (like they regularly do to try to cap pay for the finance people who manage the university's giant endowment programs). Happy to be enlightened otherwise, but would be very surprised that universities compete on individual-specific pay packages in a meaningful enough way that professors would be tempted to regularly jump ship.
I’m a professor… some additional clarifications. 1) Adjuncts are sometimes Ph.D.s (particularly in the humanities), but in other fields are frequently simply people who know a lot about the field (practitioners). Practitioners don’t do research, therefore they are not candidates for tenure track positions. Adjuncts with Ph.D.s are almost always out of the research game almost entirely. 2) Teaching faculty don’t have much time to do research. A full course load in many schools is 4 courses a semester, at others it’s 3. Grading, teaching, dealing with student issues, etc. all take up a great deal of time. Research is not a side-activity — it takes an incredible amount of time as well. Anyone on the teaching track is likely to not have much time for research. 3) In fields supported by grants / labs / access to technical equipment, universities frequently must fund the up-front costs in the hope that the professor will obtain grants to cover future costs. In Engineering, a startup package for a new tenure track faculty member might be in the hundreds of thousands of dollars, for example.
Overall, however, the analysis that universities have a “prestige” track and a teaching track; that the teaching track has a surfeit of people willing to go into it, and that it is deliberately obfuscatory so that the “advertising” can be about prestige is spot on. To speculate a bit, I think that one of the reasons that teaching is easily undervalued (not just at the university level), is that it is so difficult to measure teaching quality. Clearly some teachers are better than others at getting students to learn the material. However, our instruments for measuring teaching quality are really terrible. By contrast, everyone can count papers, citations, and grant dollars and can assess the quality of journal that the paper is in (but assessing actual impact is very hard as well, which is why that also isn’t a big evaluative component).
Once you are on the teaching track or an adjunct you spend way more time prepping new teaching materials and applying for jobs the next year so you don’t get research time. Changing universities or being given new teaching to do every year has an enormous marginal cost compared to teaching your same courses year on year when you are TT. In some professions the journals are corrupt and they will publish the work of people who already have good prestige signals so when you get a little prestige it compounds very quickly
Working on the research side... I think it's even more dysfunctional.
There's research and academic staff. Research staff typically have to chase their sallary from grants.
Academic staff typically get a steady paycheck and don't need to chase their salary and have more teaching duties.
Many research staff enjoy research and want a quiet life but almost universaly hate the constant grind and uncertainty of chasing grants.
Many research staff at prestigious universities will at some point move into an academic post in a less prestigious university for more money and more stability.
Almost any particular field is a small world where everyone knows everyone and networks, collaborations and connections have their own value.
I was one of those PhD students who graduated with a much better teaching CV than research CV, so I was considering applying for university teaching positions when I graduated. (I ended up at an tech start-up instead.) I think I had very different impressions of the market.
(Context: American R1 institution, graduated a couple of years ago, in engineering.)
My impression was that the academic job market was primarily for research, and if you mostly wanted to teach, you were basically looking in a different job market. Traditionally, some of those people would be job-titled lecturers, and there was some grumbling among teaching-focused faculty about being treated as second-class citizens. But there was some optimism about change, as some institutions came round to treating teaching more seriously. Some departments in some research universities had started hiring for tenure-track teaching faculty positions—"with voting rights", one of my professors explained to me—with it explicit in the job title, say, "Assistant Professor (Teaching)". It wasn't super common, but those departments would've wanted to build very strong undergraduate foundational programs. Or, you could go for teaching-focused colleges, like liberal arts colleges, or engineering colleges like Harvey Mudd or Olin. But this was still functionally a different, and smaller, market to the "real" academic job market.
In the real job market, people seemed to think that hiring committees claimed to care about teaching, but didn't _really_ care. And as a student, every faculty member I knew was research-focused (and possibly loved teaching, but it wasn't their primary role), or if they were teaching-focused you knew it from their job title. So I never really felt that there was any attempt to obscure the distinction between teaching and research roles, more just that teaching roles… weren't that big a thing.
That said, I admit I can't really provide any insight into how the adjunct thing works, other than to say I didn't think it was about teaching.
The question Scott asks is an excellent one (Why Is The Academic Job Market So Weird?). I've often wondered this, and it's weird in so many different ways that answering it touches on almost every aspect of academia and higher education. It's perhaps pointless to write when there are already 300 comments (!), many of which are excellent (esp. noting the existence of postdocs in STEM), but here are a few notes from the perspective of a tenured faculty member in a natural science (Physics) at a large U.S. public university. These comments here are fairly minor; perhaps at some point I'll attempt an actual answer to the question, which I think involves (i) the existence of multiple competing, incompatible, and often cryptic goals and (ii) the mismatch between these goals and where money comes from.
1 Poaching. As many have noted, in the sciences it is essentially hopeless to go from an adjunct to a tenure track position, mainly because the former is (usually) based on teaching, and the hiring the latter is based largely on research. A more puzzling question is why there isn't much more poaching of tenured professors, since these people have *demonstrated* that they can succeed, and postdocs newly on the job market have not. I think the answers are (1) a somewhat irrational desire for young blood and (2) the higher cost of more senior faculty, especially in terms of lab startup funds (in most STEM). Also, the lack of immediate incentive for tenured people to move leads to attempted senior hire processes dragging on forever (> 1 year) and often being unsuccessful.
2 Research income. "Colleges want two things from their professors." Actually three, the third being grant funding, from which the university takes a cut (at least 30%). This also relates to the "superstar" part of the essay -- grant success isn't uniformly distributed, and superstars can bring in a lot of money.
3 "colleges will enter a bidding war to get you" -- there is remarkably little variation in salaries, and often a lot of institutional constraints on this. There are, however, bidding wars in terms of STEM startup costs. In Physics, Chemistry, and Biology, these are often ~$1M.
4 Postdocs. As others have noted, postdocs exist, which is in some sense a trial period pre tenure-track. However, schools generally don't hire their own postdocs, and this makes sense: that postdoc's research area is probably already represented by the lab he/she is in, and hiring the postdoc as faculty would create questions of nepotism, independence, etc.
Devereaux's post is also excellent, though I do think he underestimates the role of the willingness of people to take terrible jobs in encouraging the existence of terrible jobs.
It's honestly very difficult not to read Devereaux's post as mostly being an argumentum ex butthurtum over not having the tenure-track job he admits he wants. I don't think there's any real mystery about why there are many adjuncts and few tenured professors: adjuncts are cheaper and they don't have tenure so you can fire them. And as you suggest, the fact that they're teaching (a necessary evil) instead of doing research (what academics actually want) means they can be pretty crap – and since they don't have tenure, you can fire them if they're *too* crap.
And Devereaux not having a tenure-track job also seems like just a good decision by his admins, since he's rampantly ideological to the point it hurts his academics – for example, he freely admitted that he wrote his terrible "exposé" of alleged Spartan incompetence pretty much because he hates laconophilia and the ideology of those moderns who admire them, but flat out admitting it doesn't make it any better. Academia may be grossly ideologically skewed, but "being wrong on purpose to own the cons" probably still isn't what even very liberal faculty want in a researcher.
I think there are two equilibrium tenure cultures: a strong presumption of tenure, and a strong presumption against tenure. Anything in between leads to too much politicization and infighting. In legal academia (where I work) there is a strong presumption that a tenure-track professor will get tenure. Tenure denials are rare and traumatizing events for law faculties. One consequence of this is that there is much more pressure on the entry-level market hiring decision, because you are basically hiring this person for life unless they are good enough to lateral somewhere better.
These days hiring in law is mainly focused on how exciting/interesting/productive you are as a scholar (and secondarily on diversity and course offering considerations). In prior generations there was less of a focus on scholarship, and the model was just to hire the smartest recent law graduates (many of them supreme court clerks). But now there is basically an expectation that you will have a PhD and two or more full published articles. Adjuncts in law faculties are on an entirely different track - they are either explicitly time-limited fellowships of one or two years (with the idea that you will go on the market for a tenure track job), or they are longer-term teaching position with 0 expectation that you will ever move to the tenure track (e.g. legal writing, clinical professors, etc.).
The emphasis on scholarship for tenure-track hiring (as opposed to teaching) is, I think, a product of institutional culture rather than incentives. Faculties are self-governing and (at least at most law schools) concerned with producing interesting/important ideas. I'm not sure how much the law students themselves care about how good their professors are as scholars (I suspect very little), and the new US News rankings basically don't weight faculty scholarship at all.
Within my immediate circle of current and former colleagues there has been pro-athlete style poaching of researchers by universities and non-university academia. At t=1, t=3, through t=25 years from completion. Within Engineering, Physics, and CS.
My advisor was poached from R2 to R1, months before I joined his newly forming group. 50-ish guy then. My sister’s advisor had the same - her first advisor had been recently poached from a DOE lab to R1 university. Both cases were university corporate donors who committed to funding specific research expansion, then deans etc headhunted - not existing grantors coming along with professors/researchers deciding to go looks for greener pastures. Full disclosure: I quit mine, my little sis quit hers.
Within CS there are 10+ professors or PhD’d researchers I know personally who have been approached by approached by at least one university dean/board member/prestigious alum with an offer of “come join us, we want you, and have funding lined up + new space for you + (perks)”.
But since it’s CS, and decently talented folks with recognized achievements start at $1m-$2m base salary + $xM I’m stock at any one of hundreds of tech companies, universities have to compete to play for that talent.
As an academic drop-out I do want to mention that in my experience the tenure track profs are sometimes Einsteins, and often just people who know how to play the funding/publication game.
My advisor was a "meeting post" type. He published a paper in the 80s that sparked a bunch of futurist speculation in the media. The research itself did not end up being seminal in the field, but it happened to catch the public's imagination at the time.
He leveraged that fame/prestige into setting up a self-reinforcing cycle. He has access to a lot of funding, so he attracts a lot of smart people into his lab. A small subset of those students end up doing interesting work. He takes those students and parades them through conferences and such, which secures him more funding.
So cynically, universities mostly compete for prestige. And prestige has little to do with meaningful contributions to human knowledge, at least in my experience.
In terms of colleges being good at predicting research success, and virtuous/vicious cycles, this is what things look like from my perspective as a math postdoc.
There are 50-100 top universities that hire full research professors. When I talk about tenured positions I will be talking about these research positions, despite the fact that many other positions are also tenured with some level of teaching. You can determine if a professor is a research professor or not by looking at their teaching load. It varies by field, in math the lowest teaching load at top universities is typically 1,2 which means you teach 1 class one semester, and 2 classes the other. A position with 1,2 teaching load is a research position. A 2,2 teaching load is mainly a research position as well. 3,3 is a teaching position with maybe a little research on the side, and anything more like 4,4 is entirely teaching. This is just a fact of life because there are only so many hours in the day and time spent teaching is time spent not doing research. (these numbers are for math. For chemistry/physics and other sciences the teaching loads are lower, say 0,1 or 1,1 for research positions. This seems to be because everyone takes math classes in college, but many students never take a chem or physics class so less teachers are needed, also likely because other sciences bring in big grants for labs and universities like this. Math is a bit weird in being a STEM field which requires next to no funding or grants to do good research.)
Research professors are hired entirely based on their research prospects and teaching is essentially irrelevant. Tenured Research positions tend to be very comfortable, and have many more good applicants than there are positions. You can tell how impacted/competitive a field is by looking at how many years people with recent tenure track offers spent in postdocs (short term research positions that are not tenure track) before they got that offer. In math people typically do ~3 years of postdocs. In theoretical physics I hear it is more like ~6 years. In econ I hear it is often ~0 years so many people are hired right out of grad school. There are people who get tenure track offers in math right out of PhD but these people are exceptional even among people who make tenure track and were successful enough in their PhD research that they were at basically no risk of not getting tenure somewhere if they did a few years of postdoc.
I see 3 reasons for the empirical fact that if someone does not get a tenure track research position after the standard number of years in their field plus one or two years, they are likely to never get one. (So if you have been a math postdoc and not gotten a tenure track research offer after 4-5 years you likely won’t get one.)
First: People have different levels of aptitude in research. Some people are strong on some combination of intelligence/hard-work/choosing good research problems/having a personality well suited to research. People who are sufficiently good at this are usually apparent by the end of PhD or a few years after. People get tenure track offers when it becomes obvious they have this, so if you haven’t gotten an offer after doing a PhD and a few years of postdoc, then you have spent a lot of years doing research without displaying this. Universities want someone who has a chance to do crazy good research. After you have spent 6 to 8 years doing research in PhD and after, the chance of you suddenly getting a bunch of way better results is not that high. Skill caps are very real in math research and the top end is extremely high. I was pretty successful in PhD and have a Postdoc at a top 20 school, but I knew several grad students who made me look like a lazy idiot. Think 5 times as many papers as me and their best papers are orders of magnitude better than mine.
Second: The more promising you appear as a researcher, the more likely you are to find postdocs with no teaching load or very little teaching load. This means you have more time to focus on research than the people who looked a bit less promising, and the gap widens.
Third: If you have done obviously good research in the past then people are more likely to reach out to you about exciting questions in the field because they know you might be able to help solve them, and people are more receptive to working with you on projects you are excited about. This means if you have better research than someone else at the end of your PhD, then you will likely collaborate with better researchers in the following years and have more/better research problems to choose from.
Another fourth thing that is not very relevant in math but is relevant in experimental sciences is lab funding. Someone with more promising research gets a postdoc at a top 5 lab with a bunch of cutting edge equipment, and their friend with less promising research goes to a top 50 lab with decent equipment. There are more opportunities to do exciting research with better lab equipment. I work in math so I don’t know how big an effect this has.
Getting a tenure track research position many years later than usual does occasionally happen but it is quite rare. Examples of this look something like Yitang Zhang, who spent 22 years after finishing his PhD in several positions including lecturer and accountant, then surprised everyone by proving an extremely impressive result related to the twin prime conjecture. He won a bunch of prestigious prizes for this and got a tenured position at UCSB 2 years later. This is an extreme case, but the general idea is colleges want research professors who get results that are so good they attract people to the college. For people early in their career, universities speculate a bit based on the quality of your previous research. If you want to get a tenure track position more than a few years later than usual, you probably need to get a result that big before you get that position. That late in your career colleges will not speculate on you.
Maybe graduate students are the missing link in all of this? Prestige doesn't just come from top-level professors, it also comes from students which find immediate success upon leaving the school. Job guarantees are a big recruitment tactic, after all. So, tenure track positions are given to recent graduates, in order to pad the school's success rate. But, hiring too many of your own students is suspicious, so colleges dance around this restriction by hiring externally, weaving a sort of web of academic interdependence. This probably wouldn't be feasible in a different industry, but with all of the collaborative research and critique going on, I'd wager that tenured professors from a broad range of colleges are fairly well connected.
When asking the question "What process carves off 30% of professors to get good pay and benefits, but passes over the rest?" it's very important to note that the process is not in equilibrium. Tenure track roles used to be a much higher proportion, but that proportion fell in the past few decades and is seemingly continuing to fall.
You have to think of the "superstars" thing a bit like venture capitalists with Unicorns. They're gambling on "potential", but the potential they're hoping for is even more rarely realized than you're thinking: literal Einstein (or at minimum Weinberg, who according to legend demanded to be paid more than the football coach). Everybody else, they're trying to find a really big story really early. So what you get is kind of the same dynamic where people invested in Theranos rather than a comparable startup that wasn't run by a charismatic Stanford dropout. (It's also like Theranos because of the reputation cascades: the top people like this person, so they must have potential, so we bid them up, etc.)
Now, does that mean they're unhappy with superstars they hire who are not literal Einstein? No, because everyone else (grant agencies, philanthropists, heck on a "what is worth covering" level science journalists) is trying to make the same gamble. So as long as a sub-Einstein/Weinberg superstar can keep seeming like they could end up as Einstein/Weinberg later, people will keep throwing money and status at them, and the university won't need to spend much in the way of resources to keep them around.
(Disclaimer: I'm saying this from the perspective of a (sub-)field that in general does not do the adjunct thing. I know zero people in my field who became adjuncts, and a lot more who left to do data science or machine learning or finance.)
I'm a professor at a "teaching college" in Europe. I don't mean that research doesn't happen here, but (1) it's not supposed to be a priority and (2) what research does happen is applied research, i.e., industry projects.
Even so, research sounds so sexy to the school administration. I am a teaching professor, and always have been. A few years ago they put in the requirement for me to...publish? At first, I cheated by publishing pop-sci articles. Now, with only a couple of years to retirement, forget it. By the time they notice my paper deficit, I will be gone.
What I find hard to understand is this: Educating our young people is a hugely important task. Why does the school administration not understand that? Why are they suddenly trying to compete with research university that have different staff, nationally funded laboratories, etc, etc? We cannot compete in that space, and wasting effort to do so distracts from what we are supposed to be doing.
Finally, allow me to offend a lot of academics: There are too many low quality PhD students, and low quality universities awarding them degrees. Almost every university has a bunch of PhD programs, which they want to fill, because PhD students make for cheap labor as TAs and even teachers. Most of those PhD students don't do useful research, but they serve their purpose for a few years and get their doctorate as a reward. Then they join the ranks of adjuncts, driving the pay even lower.
> But why aren’t colleges more like un-salary-capped sports teams, trying to outbid one another for their rivals’ superstars?
There are three big pieces missing in this analysis:
a) The "superstars" have their own ambitions about what kind of places they want to work in. Adjuncts who become "superstars" move to more prestigious places, but they can be offered better positions where they currently are too. See e.g. Yitang Zhang, who solved a big maths problem and then, quoting wikipedia: "He served as lecturer at UNH from 1999 until around January 2014, when UNH appointed him to a full professorship as a result of his breakthrough on prime numbers. Zhang stayed for a semester at The Institute For Advanced Study in Princeton, NJ, in 2014, and he joined the University of California, Santa Barbara in fall 2015."
b) It's not so much about money - working in a "top tier" place with "top tier" people and better research funding (often from external grants) seems to be more motivating than getting more money yourself. The people who want to make money are already pre-selected against, or they make money from side-hustles.
c) Such poaching *does* happen, but is usually more subtle. See e.g. KAUST, a Saudi university which pays enormous amounts of money for researchers to be affiliated with them (and sometimes also work there). It's working moderately well.
The idea is that deficient guidance leads to sticky unrealistic expectations causing adjuncts (and to a lesser extent other "trying to break into academia" players) to engage in what to outsiders seems like self-harmful and irrational behavior
I cannot claim I read all 351 comments carefully, but one thing I have barely seen mentioned, which is the big thing, is that most of these tenure track jobs are held by people who have been in their job for years and likely decades. As they retire, many of these tenure track lines are simply abolished instead of being filled by someone else. The 30 percent mentioned by Devereaux is likely to shrink a lot further, and many subfields are basically dead as tenure-track already. Tenure as it was once known is likely to vanish in our lifetimes. (And this is completely separate from any political attacks on the concept. DeSantis et al. are simply trying to hasten the inevitable.)
Lawyers are traditionally elite jobs, but there's just too many credentialed professionals to admit them all into the elite, and on the other hand too much demand for their work to limit the number of jobs available to them, and on the third hand there's little incentive for the current elite to forego their own professions' elite status. So the elite status remains available, but only for a select few - some actual superstars, but mostly the current elites' progeny and proteges. The rest remain commoners with elite job titles.
I think this is right. It’s all about money but also about who controls the money. University’s make money from three ways: student income, research grants and philanthropy. They are all proportional to prestige which is proportional to research metrics of top researchers more or less so hiring them brings in more student dollars. Once students are committed the teaching just needs to be at a certain level and you’re good so no need to compete on teaching just hire bare basic there and cut it to the bone and use the rest of the money to subsidize research so you can max out support for top researchers. And I agree make the titles confusing so everyone thinks they’re being taught by the top researchers. Prestigious researchers also bring in more research money as they can put in more competitive grant apps because they have the metrics and the money for a big team who are all providing them with data and help writing grant apps. Those people can then make demands about how they want things to be as otherwise they’ll go and take the money with them. Also it’s the prestigious researchers and their friends who all decide who gets the funding as they are the ones who review grants and who funding agencies and journal editors seek advice from. So they structure the system to benefit themselves. This is natural and unsurprising. You might think it’s ok if the system were actually good at picking excellent researchers but really it’s not, it’s good at picking people who can play the game well. Which means making friends with the right people to get good reviews getting invites to the right conference etc. These tend to be very outgoing people who are socially very skilled but often scientifically pretty mediocre. Also they tend to be a lot older as relationships tend to accrue. Same reason the leadership got real old
In the Soviet Union. Like it takes forever to get high up in academia to a position where you can actually influence things as you have to be a senior researchers which means old. Like it’s almost structurally impossible to lead anything in academia unless it’s in a field where you can bring in large amounts of money from a rapidly growing new industry. Like in comparison to the age of ceos of breakthrough companies it is kind of silly how old academic leadership tends to be.
I haven't managed to look at all the comments, so sorry if this repeats someone. I've spent time as a part-time teaching lecturer (in the UK, may be a bit different), and one thing is that you spend your time preparing and teaching – there isn't much time to do the research/publishing stuff you need in order to appeal to a TT job recruiter. Perhaps a fresh PHD is [a] current in all the required new stuff a university needs (both academic and cultural – i.e. progressive, whatever, doctrine), [b] fresh and ready to do research as that's what they have just been doing. [c], possibly, young and malleable to be moulded for the politics of the institution. (No, I'm being frivolous...)
The usual path for research STEM faculty is to get a PhD and then do a postdoc, after which the best researchers have had time to develop and often have stronger records than people with much more experience. Some people are late bloomers and do a second or third postdoc and succeed after that, but being hired as a teaching adjunct usually means failure to get a good additional postdoc.
I was hired as tenure track at MIT after being a postdoc at MIT and doing a PhD that involved working with MIT professors. If a college really wants you, they'll be happy to higher you for tenure track even if you were already there. If you perform above expectations along the tenure track, other colleges will try to higher you away and offer large salaries. If you perform below expectations you might be denied tenure.
I haven’t read all the comments, so apologies if this point has been made.
I think a hypothesis has been skipped. Departments (who do the vetting of tenure-track candidates) have a pretty good idea of who can be a thought leader based upon their graduate level research and pedigree. A candidate who attends a top tier institution and has a thought leader as an advisor has a better chance of becoming one of the prestigious research professors of the future. So, the promising candidates get hired out of grad school and are retained by the institution that hired them. Of course there are surprise outliers, adjuncts who publish important work while maintaining a heavy teaching load, but this is uncommon.
There are also errors in the other direction, hires who get to tenure then rest on their laurels, but this is also uncommon.
I’m not defending the system, but noting that the data can indicate a type of market efficiency.
> But then why do they only hire inexperienced people? Why only people from outside their own institution? Here I’m even more confused, but a few guesses:
Surely some degree of this is the woke purity tests?
A young person is likely better at being for all the required beliefs, like supporting newly trans rapists being legally considered women, have a shorter history on social media and head just has exploded at the rapid swapping of beliefs where they confuse the politics situation of 6 months ago as the current beliefs, maybe they don't remember bird flu at all so when corona happens they are a blank slate for whatever medical claims are acceptable this month(masks dont work, we lied about masks for front line workers, masks outside 50 feet away from anyone are saving lives) or that obama bombed civilians and so can believe that if a black man was in the white house Americas forien policy would be magic.
A far away institution probably has less knowledge of what someone has said in private.
I'll echo what others have said about this post, and the relevant source, not generalizing well to STEM, confusion as to why postdocs were not mentioned at all.
To Scott's final confusion: Why not hire everyone in some low level capacity?
I can only speak for STEM fields, but my impression was that this is what Assistant profs used to be. There were many openings and it wasn't too competitive, but getting tenure was quite hard. This is essentially placing low-stakes bets on many people, and seeing who becomes a superstar. Now it is flipped in that it is very hard to get a faculty job at anywhere doing high quality research: my department recently went through ~400 applications for a position and made no offer... but tenure (although not guaranteed) is much easier. I've heard that MIT and other places are still "old-fashioned" in this regard.
The confusion is then stated as: why did things change? I don't have a great answer. Perhaps it is related to funding structures. I'm not convinced that Scott's idea would be any better than what we have now, because most people don't end up becoming superstars.
Another subtlety: faculty in some of the sciences can be divided into two categories:
1) Research faculty: these people win grants (from which the Uni takes a cut), advise many students, and teach ~a course per semester. Typically these would be more applied fields like quantum information, biological physics, condensed matter be they theorist or experimentalists
2) "Liberal arts faculty": Profs who can't win grants because they work on fields no one funds (String theory, quantum fundamentals), but the uni takes them on because of the "prestige" of having them. They aren't making the uni any money, and many of their students (if they have any) are supported by external fellowships. Dominantly theorists, and they all teach; most of their courses are hard! (A friend of mine took a course from a string theorist who handed out the final, said "I'm sorry" and left the room)
There are vanishingly few of 2), if any, in the biomedical sciences, a few in Evolutionary bio, and a bit more (but still few) in Physics/Math. Many don't properly separate these and get further confused when trying to understand why the responsibilities and hiring for some faculty can be so varied. The vast majority of real liberal arts faculty follow a model like this where support comes almost entirely from the university.
Does anyone understand the first "plot"? I've only known one STEM grad to get a faculty job directly after their PhD (and they were told it would be better to do a postdoc first!); most postdoc gigs are 2--4 years. Among Social sciences / Humanities I've known a few, -- but my impression is that there are far fewer openings per year in these fields. I'm very confused....perhaps only the ones who know they can get a job apply their first year out, and the applicant pool is just much larger in the 2-4 year range?
Article brings some interesting points, despite it has been deeply discussed in media and with professionals and organizations in the field.
As I am involved, I’d like to put here my perspective here more impartial possible.
So, at first, and been absolutely partial only in that subject, I don’t think everybody who was taking to this job, actually was looking for fame and money. Primarily, I dislike generalizations once they never will represent the whole, soon even if it is the minority, this shouldn’t be desmoralize or not listen because of the main majority.
FACTS:
Actually, particularly in this situation, the individual who was recruited and he/she didn’t look for or apply for the job.
He/she capability of production has to be measure by the amount of tasks, challenges and daily work and homework, which, maybe you doesn’t have this known, but the program is established in a progressive school, where students are launched. While the tests requires a great deal of memorization, (it’s all about “mind games”), it can be told it is not purely educational. It is about stressing in the maximum level students to make them learn focus and learning in the worst scenarios. In that way, they can learn how to unblock and access contain they even know it was there.
The main factor that some scientists disagree is about not to teach and develop the necessary skills before start the training. The program consist in make the student self-learning. Detail: it is require lots of technical skills.
Considering less than six months, I consider the improvement and increase of the individual very impressive.
But, to be honest and impartial, I will let the own scientists give you the results:
All these theories assume that departments are acting in their own rational best interest, so I'd like to propose a totally different explanation: the failure of decision by committee.
An adjunct who has worked in the department for 6 months to a couple years is known to the people on the hiring committee; their work, their attitude, their skill set. Some of the people on the hiring committee will like them and want to bring them on, but others will be looking for someone else (more research focused, less research focused, more interested in teaching a particular class, better at teaching certain classes, particular personality differences, etc etc etc). A new applicant is unknown, and everyone can convince themselves that "maybe this person is what I'm hoping for", and so gets more general support (or at least, less severe no votes).
I think I can answer Scott's questions but warning: this comment is long. I did my Ph.D in the natural and physical sciences at a top 10 institution and now have a tenure track job at a small liberal arts college where I teach several classes per semester and do some research. Some background first.
About teaching: Teaching is particularly time and attention consuming. When I'm teaching, especially large undergraduate classes, the top thought in my brain is usually about that. Research is also time and attention consuming, so even for talented people it's hard to do both. If I take on research that's too ambitious I'm likely to fall behind the rest of the field or get scooped entirely because I work more slowly since my attention is split with teaching. Research also moves fast, with new techniques being adopted over a period of a few years. Being out of a research area for several years, or just being slow to keep up because of a heavy teaching load, means you have a lot of learning to do if you want to try to ramp up your research productivity, on top of the time and difficulty of the research itself. Most people who try to be good at both solve this by working on problems that aren't very interesting to anyone else where it doesn't matter if they go slow. Or, they work at prestigious liberal arts colleges with lower teaching loads and more resources for research. Ruthless time management is also important, but not enough on its own.
About grants: Productive researchers don't just add to a college's prestige, they also contribute financially. They are expected to bring in millions of dollars in grants of which 40-50% goes to "overhead", where the college takes the money and uses it for whatever they want, and the grantee doesn't see any of that. Grant funds are also used to pay graduate student "tuition." This is another transfer from grant funds to college since on paper graduate tuition is as much as or more than undergraduate tuition even though for most of their Ph.D the student isnt taking any real classes, just doing research full time. The student never pays a bill, it just comes out of their advisor's grant funds.
About networking: Getting grants is partly about actually having good ideas and results, and partly about convincing reviewers that you have good ideas and good results. For research that is good but not outstanding it's a subjective decision who to fund and who not to. Since reviewers are other researchers in your subfield, there are high returns to networking and selling your work. People preoccupied with teaching don't have time to do this at all, while prestigious researchers spend a huge amount of time on it in the form of giving talks and going to conferences - my former advisor was often away for weeks or months at a time. Top 10 graduate schools are also full of networking opportunities, eg. the college pays for students to take visiting professors in their subfield out to lunch.
Ok, now to answer the questions:
1. In the natural and physical sciences colleges can't hire everyone into a low commitment job because just to get started requires a huge commitment, usually a startup package in the hundreds of thousands or millions. Colleges put this up because they expect to make it back later in graduate tuition and overhead from future grants. They can't do this for everyone because if the new hire flames out, they wasted a bunch of money on stuff that will be hard for others to re use like salaries for graduate students and postdocs, consumables, and specialized equipment or instruments that may not be useful to anyone else.
2. Colleges mostly hire new Ph.Ds or recent postdocs because they are current in their field and haven't fallen behind by spending all their time on teaching. If they are politically savvy they have also been networking with others in their subfield during their graduate and postdoc years which makes them more competitive both for grants and for getting the job in the first place. Colleges do sometimes hire experienced superstars by poaching them from other colleges, but usually you need to offer an established superstar a very sweet deal to get them to move, like their own building. Superstars bring in more grant funds but not that much more, so the decision usually only makes sense if you think trading money for prestige is a good deal, which it only is at top 10 institutions.
3. I'm not certain why departments have a cultural norm of preferring to hire from other institutions. My former advisor told me it's because they dislike the appearance of neopotism. Maybe, but I suspect it's also because cross hiring is good for networking. If your entire department was trained at your institution and mostly knows each other you'll be at a disadvantage when applying for grants or looking for collaborators.
These weird confusing things happen less in fields where research is cheap, like math, or where cushy highly paid industry jobs are readily available, like CS. In math Yitang Zhang went from adjunct to superstar by solving a huge open problem. This worked because he didn't need specialized equipment, worked on it slowly for literally decades, and the problem was so hard he wasn't in much danger of being scooped. In CS it's more common for senior people to transition back and forth between academia and industry.
1.) You really, really can't do good research and have a heavy teaching load at the same time. Your proposed idea of "hire adjuncts and see who turns out to be superstars" doesn't make sense. The adjuncts will not have the time (or, in lab science, the grant funding) to demonstrate that they are superstar researchers.
2.) As many other people have said, poaching absolutely happens.
3.) Maybe this has changed recently and my experience is pretty math-specific, but traditionally my impression is that academics fundamentally trust Academia's ability to recognize a researcher's quality, and trust that department rankings accurately reflect department quality.
The model is, "good people get jobs at top institutions right out of grad school; those who don't are probably not good."
Sure, there's general acknowledgment that there are more qualified people than there are tenure-track positions, so good people get rejected all the time, and there's an element of luck in career advancement. Everyone knows this. Everyone also knows that there are "who you know not what you know" elements and you're more likely to get hired/promoted if you have personal relationships with people influential in the department.
Somehow this doesn't seem to translate to an equally firm belief that there are underrated people who "got unlucky" on the job market but would be as valuable to the department as their luckier peers, or a desire to "scoop up" underrated applicants.
Losing, it seems, brands you as a loser...even though *everyone knows* that many excellent people must lose every year through bad luck alone, by virtue of too many people chasing too few jobs!
While a lot of the incentives here make sense, having spent most of my life being raised by/married to college professors who have been part of or run the hiring process for their departments, this doesn't match the process on the ground as I've heard about it.
I think it mostly fails by imagining 'the school' to be a unified entity with coherent goals and plans, rather than a conglomerate institution with a a bunch of heavily divided and often openly antagonistic factions.
To whit: my impression is mostly that colleges and universities have pretty strong unions for professors, which want to have 100% high-paid tenure-track positions. Who are in a constant state of cold war with the very powerful and goal-oriented administrators, who want 100% cheap adjunct positions.
(Or, if they're trying to be a top-tier prestigious institution, maybe want 10% tenure track all-stars and 90% cheap adjuncts. But if you're looking at national statistics rather than anecdotes from the ivy leagues, those few prestigious schools will get swamped by the thousands of community colleges and 'average' local colleges that make up the bulk of the numbers)
What I've seen is basically a slow shift in the power balance, with administrators ratcheting in few more adjunct positions every time there's a lean year of budget-tightening, and giving handouts to the old tenured professors who are about to leave anyway to molify the unions, while never actually adding more tenure positions back in when things are good again.
As for why there's a preferences for newer professors in tenure-track positions, I think there are three main factors I have seen in action:
1. Being an adjunct for 8 years and applying for a tenured position is sort of like being unemployed or working as a Walmart greeter for 5 years and then trying to get hired as a program lead on a large software project. Everyone knows that being an adjunct sucks, and there's an assumption that if you haven't been able to get a better job than that for many years on end, there's probably a reason why. It's taken as strong evidence against your quality as candidate.
2. Professors always want their department to tech exciting new ideas and use exciting new techniques that the field has developed sine the last time they were is school, but they almost never want to take the time (or have the resources) to learn enough about them to teach them themselves. Hiring the freshest graduate means getting the newest techniques and ideas and feeling a little bit less behind the times and a little more useful to your students.
3. Because the administration keeps shrinking the number of tenured jobs, the remaining tenured faculty form sort of a 'core' that serves as the heart and soul of the department, and ensures its values and character get preserved into the future. This is probably tacitly the #1 thing that existing tenured professors care about when they interview for tenure track positions, they are basically hiring the new pillars that will have to fight with them in the trenches and continue fighting after they're gone to keep the department healthy and independent, and they'd much rather get a new pillar that is young and will be there for 40 years than one that is experienced and will be there for 20 years. (And of course, the less experience they have, the more you can impose your own values and priorities to teh new hires who will have to carry them after you retire)
“Colleges want to pretend to students that the same people are doing both these jobs, because students like the idea of being taught by prestigious thought leaders.”
I get the 2nd half of this in the most basic sense, but outside of the most elite schools, I'm wondering what anyone's thoughts are on the percent of college students (or their parents) who even know that some professors do both those jobs, what the tenure system is, or what a research university is and whether or not they attend one, especially the further you get from the hard sciences. I was a philosophy major in the early oughts and don’t think I had a clue. I'm currently doing a PhD at a public R1 and would put the percentage very low in the classes I've taught in human development/ed psych- both undergrad and master’s level.
The worst part, is that industry professionals aren't in the mix.
Let's face it, for STEM degrees, the student is really earning a ticket to apply for a vocation. Let's give them a step from university to industry.
Who are the most impressive vocational instructors? Those who have spent ten years in the industry, not the bloke who couldn't really cut it in industry. What do graduating students really need? Internships, industry contacts ... the very things that vocational instructors who rotated out of industry can provide.
Most of even our high school teachers should be drafted out of industry for a year of teaching, then return. In my senior year, I saw the least motivated students say 'I'll just take the BA and become a high school teacher.' Oh right, that person is going to be a great motivator. [last sentence was sarcastic tone]
Suprisingly few mentions of the fact that this isn't a steady-state situation. It's adjunct*ification* that he's talking about.
The process is optimised for hiding problems more than anything else. It's easier to quietly shift more people to the 2nd class track than to make the experience of members of either track (especially the 1st class one) worse.
Honestly this would be an interesting situation to throw a marxist lens at. The preference for choosing the tenure-track ἐκλεκτοί as soon as possible in their career separates those who will make the decisions from those who will bear the brunt of them.
I'm reading B. Devereaux and thinking; this is just like the moral mazes (by Zvi), (middle management capture of the 'ropes of power' of the institution.) It's a piece of Moloch. What to do? Maybe teaching colleges and research universities. With only a smaller percentage (of students) going onto research. There would still be research at the teaching college, but the focus would be the students and not some new result... though 'new' stuff happens every where, all the time.
Oh and cut out the middle management, Profs have to run the place.
I'm in computer science, and think maybe what I say below is somewhat true of close disciplines like math or engineering, not sure about other fields though.
> Or it could be a vicious cycle - if hiring committees dismiss you, then grantmakers and journal editors will also dismiss you, and you won’t have the resources you need to do great work.
> Here’s another question that confuses me even more: Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?
The vicious cycle is more like this (and explains why the second strategy won't work): It is almost impossible to do good research without a job that grants you significant time to do research. Yes, Einstein was a patent clerk, but guess what... he's Einstein. The rest of us need much more time to think and devote to research if we want to get anything done.
Adjuncts have high teaching loads and little time for research, so it's rare for them to improve their research record while working that job.
To counter your claim about preferring newly-minted PhDs... actually more experience DOES help you get a tenure-track job. I was a postdoc for 6 years and got a WAY better job than I could have straight out of my PhD. With a few exceptions, most people we've hired in my department since I got there have several years of postdoc experience (or were tenure-track faculty at another institution).
Postdocs are somewhat similar to adjuncts in being lower status and having no job security. (In fact, their job is guaranteed to end after 1-5 years, whereas an adjunct can in principle stay indefinitely.)
But the key difference: since a postdoc's job is almost exclusively research (or say for math postdocs, have a low teaching load, so still lots of time for research) they can improve their research record in a way that's much more difficult for an adjunct teaching 2-3 courses per term. No one is as *single-handedly* productive in research as a postdoc: they have the research skills of a professor coupled with the free time of a PhD student. (Professors with lots of papers every year typically have lots of PhD students and postdocs working for them, but if they were on their own they'd publish less than a typical postdoc due to their extra job duties.)
I've never seen anyone on a hiring committee ding someone just because they are an adjunct. I've just never seen an adjunct with a strong research record apply in our department. Obviously there will be a counter-example someday, and likely there are counter-examples now, and I'm not making a claim about any individual adjunct. But the law of large numbers kicks in at some point to make the *average* postdoc much more research-productive than the *average* adjunct.
No, they aren't. Lots of poaching happens, as discussed by the other commenters.
But you really have to be a SUPERstar. If you're merely doing well, then that's typically not enough to get hired by a place even at the same rank as your current university, and certainly not enough to move UP. In general, it's much more difficult to get hired as an Associate Professor (the rank just after you get tenure) at a university than it is to be granted tenure (i.e., to be promoted from tenure track Assistant Professor to tenured Associate Professor) at that same university.
Here's my guess: Once someone has been an adjunct patiently for five years, you figure they'll probably still stay on even if you don't give them tenure.
Or roughly the same thing from another perspective, maybe people who care about money will leave academia within five years unless offered a tenure or tenure-track job. The rest can be taken advantage of.
"Now you may ask why anyone would take a job like that with poor pay (for a job that requires a PhD!), no job security and no benefits. And of course the answer is ‘because they have no other choice;’ leaving academia, even temporarily for a non-academic job is generally a career death sentence"
This is my question, and the answer is provided, though I consider it unsatisfactory. Someone with a PhD should be able to be hired into some useful, PhD-related position, unless the degree is in something purely academic, like history. One should, therefore, if one aspires to a career involving a university and cannot get on the tenure track, then one should hope to come to the university as a "professor of practice".
Anyone who can earn a PhD ought, by definition, to be intelligent. I still don't understand how the majority of adjunct professors would maintain that position and fruitless hope for a tenured position under those conditions for more than two years.
I have recently started as an Assistant Professor at an R1 institution, in biology/life sciences. I think the reason places try to avoid hiring internally is twofold: the first as Scott identifies is that there would otherwise be pressure to keep everyone, including more marginal hires, since you know them. But there is a second stronger force - R1 institutions are trying to hire faculty that will become the next superstars. And it is a LOT harder to become a superstar, with your own independent brand, if you stay at the institution where your previous mentors are located. Initially most people in the field at large will assume the ideas came from the mentors and you just executed them. There are exceptions of course - post-docs and trainees who are so good that they already have a lot of their own recognition in the field, and often these are the very few people who are recruited and retained internally.
Academic in mid-tier university in the United Kingdom here. I want to explain concretely how the graph makes sense (which others have done fairly well), but also to provide some context about the financial incentives to the university that may be UK-specific.
For academic teaching & research jobs (which may have a greater or lesser research focus depending on the university and department), one of the main factors considered by the hiring panel will be the applicant's publication record. What we are specifically looking for is high-quality publications in the last few years, rather than 20 spam papers. (High-quality usually means prestigious journal/conference, as it's not like we have time to read the papers ourselves.)
If you have just completed your PhD or postdoc, you will almost certainly have some good publications co-authored with your PhD supervisor or PI. We're prepared to take the punt that you'll be able to produce more, or kick you out after 2-6 years (depends on which university) if you don't.
If you've been working as some kind of adjunct (which for us, usually means fixed-term contract full-time teaching-only staff, although can be part-time in other universities) or as a permanent teaching-only lecturer, you probably won't have been publishing recently. So your number of publications will be the same as someone finishing a PhD or doing a post-doc, your publications per year will be lower, and you've probably already been rejected for teaching & research jobs when you applied for them with the same research record when you were younger.
I think that's sufficient to explain the graph in the post. But why do we care so much about publication record?
Scott is partly right when he says it's because of prestige. There's a specific mechanism by which that operates in the United Kingdom. About every 5 years, the government runs an assessment called the REF (Research Excellence Framework). This is unpopular, but is meant to hold universities to account for the quality of their research. Details vary from cycle to cycle, but each department has to submit around 3 papers per research academic (teaching-only staff don't count) to a panel of external academics in the same subject, who score the papers on a scale from 1-4. The score of the department is published and feeds into highly publicised university league tables, as well as affecting the amount of government funding the universities receive.
So it is about prestige, but also about money. Hiring low-quality researchers lowers your department's average, sending it down in the league tables and losing you government funding.
Of course, there's a more direct mechanism that makes hiring active researchers financially important. If you have a good publication record, you are more likely to be taken seriously when writing grant proposals to the government-funded research councils. Some of this goes directly on hiring new post-doc researchers or buying equipment, but maybe half goes towards the general running costs of the university. As ever, it varies by university, but in the top research universities, most of the income comes from research grants, not student tuition fees. (In Oxford/Cambridge, the university loses money on home undergraduate students.)
Once again you admit to basically being clueless about how an institution functions, only to then speculate brazenly. You don't have to endorse the structures, but if you really want to add something of value to the conversation you should ask a lot more questions before attempting to spread a bunch of guesswork.
Many very smart people have put a lot of thought into centuries-old norms such as tenure and non-incestual hiring. Such practices are fundamental to the mission of higher education, but are easily ridiculed with the help of facile libertarian assumptions emerging from a basic lack of interest in that mission. (Preferring to weigh the efficiencies of markets, status implications, etc, may be more in your wheelhouse, but without investigating how/why the structure was built it amounts to just idle disruption.)
Of course there's a ton to be fixed in academia, but it's like democracy: if your heart's not in it, you're contributing to its failure. Yet you settle for a cartoonish version of colleges writ large, personified as a businessman who "prefers" a game of pretense with his customers, while most colleges have in fact been wrestling tirelessly to retain many core, long-held values during this period of late-capitalism, where popular opinion treats them as if they were just glorified trade schools.
There are other institutions with this same two track system, for example officers and enlisted in the army or partner-track and non-partner-track jobs in a law firm or consulting company. So whatever is causing this structure probably isn't driven explicitly by academia, but rather some property of academia that it shares with both the army and legal partnerships.
"Second, they need them to do good research, raise the college’s reputation, and look prestigious."
The obvious problem being that there are entire fields where "good research" is literally impossible, because those fields draw no distinction between fact, taste, and opinion. Most of the original academic disciplines fall into this category (I'm looking at you Queen of the Sciences.) Nikole Hannah-Jones was insta-tenured, yes in order to give Howard prestige (among the right sort of people) but you're going to need a particular sort of metaphysics and a lot of hope to think she's going to increase the amount of knowledge in the world.
When a college wants to hire for a research position it's in their interest to consider all possible interested applicants and have no incentive to favor internal canidates.
Ok, so in that system how often would an internal canidate get promoted to a research position? Well, only when the college had an opening for a research position at the same time as an adjunct at that university had put together a publication record good enough to warrant hiring there but not so good as to let them secure a better job and hadn't yet been hired elsewhere...and the fates smiled on their application (lots of randomness in who gets hired).
Given that most academics on the market apply for multiple dozens of jobs the chances of this happening are just going to be really low.
--
Yes it does suggest that there isn't that much of a premium on staying put (academics prefer a TT job at Harvard than a TT job at no name uni even if it means moving) but that seems mostly correct.
Ohh and the difference with sports teams just mostly comes down to the fact that there is a greater variability in what different universities want. Universities often want to complement existing research teams or have views about what kind of research is valuable.
This makes the system much more like a matching market (eg medical residencies...but more so) than bidding on a commodity.
I think this post is misleading for a few reasons. Someone else probably already mentioned the points I'll make, but in case they haven't:
1. The chart in the post is based on data about historians, not all academics. Your post doesn't make that clear.
2. The bar chart depicts the *number* of historians hired N years after completing their PhD. Even if there was no preference for less-experienced applicants at all, the chart could look the same. Clearly, people are more likely to look for a job right after they get their degree than several years afterwards. There will be fewer history professor job applicants with 3 years of experience after getting their PhDs than with 0, because a lot of people with 3 years of experience are already in a TT position and will not apply for jobs. The chart doesn't correct for that.
I'm not denying that there's a preference for less-experienced applicants; I think there likely is, but that chart in particular doesn't really show this.
3. In a lot of STEM fields, it is vanishingly rare to be hired for a TT position straight out of a PhD; you need a postdoc first. Bret Devereaux mentioned this in his post, but your post makes it seem like getting a TT job right out of grad school is the expected career path in all academic fields.
Perhaps it's an instance of efficiency wages? According to efficiency wage theory, overpaying some employees can be profitable because it raises loyalty, morale, and productivity. https://www.investopedia.com/efficiency-wages-5206757
Tenured professor at a mediocre public university in the US here. I think a lot of what you say here about the separation between research and teaching, and why people fresh out of grad school are likely to get jobs, is probably right. But I think you also need to factor into the equation just how *bad* universities (and everyone else) are at predicting who's going to be the next Einstein. Including the case of Einstein himself, who in my limited understanding was not respected in his field at all until after he revolutionized it. So I don't think it's true that there is competition over the top few PhDs, because I don't think universities will have enough of a coherent set of criteria to even converge on the top few; once you get into the class of 'wrote a competent dissertation at a respected PhD program', within that class of people you're pretty much throwing darts at a dartboard. And it's definitely not the case, at least in my field, that there is a shortage of such people: there are far too many promising recent PhDs to fill the small number of tenure-track positions that come up each year. That said, I think 'able to secure large external grants' might be more predictable than 'will do amazing research in the future'. But I have no data to back that up; it's just a hunch.
A few mechanisms I've seen in play that I haven't seen mentioned in the comments I read:
- For impressive profs you don't want large research accumulation, you want evidence of ability to identify and lead important research. You can estimate pretty well from PhD (plus maybe post-doc years) whether people are doing important work per year.
- There's a huge pool of applicants you need to filter. Having been on the job market for many years is strong evidence that nobody else wanted to hire you. A cheap heuristic for filtering is therefore that lots of time on the job market indicates a candidate who is unlikely to succeed in your hiring/isn't valued by other hiring committees etc.
- Perhaps an uncomfortable point, but teaching university students and leading research are very different activities, and experience in one is not necessarily even helpful for doing the other (past a point). "Experience manifestly hurts applicants" seems very misleading, because they are experience in different jobs. Insofar as there is a problem here, it is with unrealistic expectations about a path from teaching-focused posts into research-focused posts (which might be encouraged unfairly by research institutions who want high supply of teaching).
Few thoughts. First of all, this is not just true of colleges. It’s also true of funded startups. I’d think of it like this: there are probably two Nash equilibria when it comes to getting talent at “low risk”. (A Nash equilibrium is a situation where no one can improve their results by deviating to another strategy.) One is let another institution shoulder the initial risk of fielding a farm team and seeing who shines and who washes out, then poach the winners. This probably seems less risky than fielding your own farm team because if true 1% stars are one in a thousand but your local farm team can only be 100-deep, odds that the true superstar will come from your own back bench are low. My guess is that this is priced in when poaching superstars, so you’re probably overpaying substantially for someone else to incubate them.
The second Nash equilibrium is probably “moneyball” where you take the pain up front and invest in becoming better predictors of talent so you raise the odds of somebody in your own incubator being a break out star. Because this requires patience and up-front sacrifice, we’d expect it to be less prevalent than the first Nash equilibrium, especially in a niche known more for playing the game than questioning the game, until the price of free agents is literally outrageous.
If my theory about the second Nash equilibrium is correct, though, it’s what somebody at one of Scott’s Bay Area house parties would call “ripe for disruption”.
If superstars are generally underpaid (to prevent e.g. "upsetting the apple cart" of other employees' expectations of no one making waaaay more than them for doing "the same job"), then they are from a hiring perspective much more lucrative and their allure to institutions totally makes sense. If they were paid what they are worth, there'd be no incentive to hire them over others. The perverse incentives created by underpaying stars hurts others rather than just the stars.
As an early career adjunct. I would also like to get an answer to this.
I'm an academic and I can point out a few factual clarifications:
1. There's R1 (e.g., Ivy+, Stanford, MIT, ...) and not-R1 (typically liberal arts) institutions. Tenure-track professors at R1 institutions are hired to do research and the teaching is incidental. It's flipped at not-R1 institutions. I _believe_ tenure-track professors teaching at R1 institutions is mostly a historical accident that got cemented into every part of the tenure/promotion/grant process so it's basically impossible to change at this point.
2. There's also STEM/not-STEM. As a concrete example, the top-4 CS programs don't hire adjuncts in the way you think about them (e.g., Andrew Ng is adjunct at Stanford but this is just to keep him affiliated with the University). This dynamic is very different in non-STEM.
3. In STEM, the University puts in substantial resources to grow tenure-track faculty (startup packages go from $500k and way up).
4. Poaching does happen a decent amount. You just don't really hear about it much because tenured professors have it very nice and there's no need to complain about the processing.
I'm also happy to share some non-factual thoughts but those are speculative.
Basically on the right track. But it isn't colleges who do the hiring; it's mostly departments. Colleges can veto departmental hiring decisions, but they don't pick the candidates. Departments obviously have slightly different incentives than colleges do (even more prestige-driven, significantly less budget-driven, less financially even very informed or aware).
I'm very confused by your not mentioning post-docs (which are the norm in STEM). In effect, these are temporary positions with a moderate teaching load (or perhaps a pretty heavy teaching load, or, if you are lucky, no obligatory teaching load at all). I don't know what data your graph is based on, but it's rare now for people to get tenure-track positions (at research universities) without doing a postdoc at all.
As for poaching, my impression is that it does happen a moderate amount, but usually only when the professor in question is upgrading to a more prestigious institution or wanted to move anyway. It is also common for professors to get an offer from an outside institution and then ask their current institution to match it, which they often will.
Adjuncts surprisingly shooting up several years into their career is a rare thing, not least because their teaching schedule (6+ courses a year not unusual) leaves them little time for serious research. (Of course, in lab-heavy fields, they face even stronger headwinds.) I think that's the main source of the vicious cycle you're talking about. Journal editors, AFAIK, don't usually gatekeep much; a non-famous tenured professor and an adjunct with an .edu address have about the same chances of getting their paper in all else being equal. (Really big names sometimes get the "even if it is unreadable, there must be wisdom in it" treatment.)
There seem to be two obvious reasons why:
1. lots of people go on the tenure-track job market the year they get their PhD, so you get a big burst. Then are people who delay this for a year or two or three — it's a pain to put together a TT application, but postdocs and one-off teaching jobs are often offered with just a cover letter. So you have two distributions; a delta function at "same year", and then an exponential decay over the subsequent years — this is marking the time at which the applicant finally makes a first TT application.
Notice that this explains away the apparent "bias against experience", simply by virtue of the fact that people choose to go on the market at different points.
2. People who are out of the market longer are seen to have a hidden variable (not being as good) that is serving as a signal.
For TT people, in general, they do not move from university to university very much; the standard thing is get a good job when you start, and then negotiate when you are up for tenure by getting outside offers. It's rare to go on the market, and generally indicates someone who is at a low-ranked institution who wants to move up; those applications are less likely to be successful, since being at the low-ranked institution is itself a signal.
More generally, people would rather take someone who looks like Harvard would hire, than someone who (even if they look like someone Harvard would hire) definitely Harvard didn't actually in the end hire.
As to the question why do people hire from outside when they do take people with experience: my guess is that academia is ridiculous and snobby, and it's very hard to get someone to see an adjunct as a TT. They're excluded (for example) from certain meetings and decision-making, and have less ability to negotiate for what they need, and people use that as a signal of actual quality. But it's much easier to say ha, Berlin totally doesn't realize that their guy is super good, we can get him, let's do it.
An inside hire does occasionally happen (my institution is smart, and does this — it's a great way to test the waters with someone), but it's also quite clearly flagged as the plan ahead of time (we think this person is good, hire them as a VAP, but there's a job coming soon...)
A similar phenomenon happens in the non-FAANG computer programming world. Junior devs get hired at something like $60-$80k, then within about two years are worth double that. But the company that hired them will never promote that dev to a $140k salary because it would upset the apple cart of all the devs making $120k wondering why they don't get a 100% raise in two years, or anything close.
So the junior dev has to go elsewhere for that first big pay bump. Meanwhile the company hires an ex-junior dev to replace them, and happily pays them $140k. But no one leap-frogged anyone or got an out of place raise, so the apple cart remains placid.
Academia is filled with mediocrity & a lack of imagination
PhDs that are above average get quickly snapped up in tenure track while others have to cycle around various adjunct positions . Experience can only help so much if your initial base itself is low
Some things you need to include in your investigation:
1) In reality, a lot of the teaching is done by graduate students. Even when it is not explicitly, it is very common that professors will have a teaching assistant that will do most of the menial work of teaching (grading, conducting discussion sessions, etc.).
2) Connected with the first point: in high prestige institutions, they usually hire for the research potential (which includes ability to attract grants, which is arguably more important than how promising it is), with teaching as a not particularly close third place (second is DEI nowadays). Adjuncts and post-docs are never hired back because they are expected to be looking for a tenure-track anyway.
3) You are correct in assuming a career gets shut down if you are denied tenure. It is very rare one can recover from that, usually involves some level of nastiness that happened for tenure to be denied in the first place, something other institutions will look over. But in most cases it is an indication of lack of impact research-wise. And when it happens, why hire someone who is also tenured (thus, more expensive) or already flamed out elsewhere when you can hire a new student? It is similar to sports, you don't run it back with veterans when you failed, you draft the potential next big thing because it is cheaper and more likely to make an impact.
4) Which leads to the final point. There is a lot more graduate students than there are jobs. Like, by a mile. A majority might never even try (myself included). So you don't need to go for experienced teachers for adjunct jobs that are temporary. There are plenty of people jumping from a one-year appointment to another for a few years until they give up and go work somewhere else. There is no short supply at any point. And when people have no experience, you can pay them less.
Other countries are available.
https://theprofessorisin.com/2017/05/26/academia-as-identity-a-ukus-comparison/
One advantage to only hiring outside of your own institution is that it enables you to select for ideological purity more easily. And academia is extremely ideological. If you make it possible for local academics to be hired for tenure, and a promising one is passed over, it leads to questions about why this person wasn’t picked. In many cases that may be because the academic wasn’t ideologically pure enough. If you make it the default to hire from other schools, then this doesn’t happen, as you can merely dismiss the non-ideological fits at other schools without needing to explain yourself, and only select the ones that do fit ideologically. Anonymity neuters outrage.
For what it’s worth, my personal mental impression is that this IS happening in other jobs. At least, the ones that also confer a lot of white collar status.
Anecdotally, I’ve heard people say a similar thing about lawyers (many more law firms hiring people as paralegals who they would have hired as lawyers; far fewer people being offered partner than in the past).
I don’t know how it is in medicine, but it certainly seems like, whenever I have a medical complaint, I can take my pick of a 100 RNs or PAs, but man, finding an actual MD with an opening anytime soon seems REALLY tough.
Here’s an interesting aside - writers and bit part actors are currently complaining as much about the job market as anyone. Supposedly, mega-star salary stays the same, but the people on the lower rungs - who in past times would have had a stable mid-paying tv writer job, or consistent living as a recurring guest star on a smaller sitcom - claim it’s becoming financially untenable to exist in these smaller jobs in huge industries.
Two immediate thoughts here: white collar jobs have prestige that people might be more willing to compete for.
In my experience, applicants to this jobs also are more likely to have wealthier parents who can help pay their way if they end up as adjuncts. Part of me wonders if plumbing would be able to get away with milking the “apprenticeship” period way more if they could get away with it. If the apprentices were more likely to be able to pay their own way for the first couple years, would salaries show the same pattern?
As an academic, I believe it is mostly the case of the superstar theory. There appears to be an extreme Pareto principle / power law at play, where the best researchers are monumentally more impactful than their only slightly less talented peers. Partly this is because of funding and publication dynamics, where a successful researcher gets exponentially more money/opportunities to continue being successful (see also Lotka's law on publication count). Given this, it makes sense to take a shot at getting the next big thing, more than having a wide pool of moderately successful mid-career researchers.
Colleges (at least R-1 in STEM fields, see diddly's comment) do not hire all low-commitment adjuncts as you suggest because the startup packages are substantial and they invest a lot in making sure new hires can be successful. But there is a similar process in the usually 7-year tenure review: professors are given 7 years to "prove" their research output before either being given permanent tenure or being fired. It tends to be a very stressful 7 years in a young faculty's career. So in some way there is a bit of a "free trial" for even tenure-track faculty that serves the same purpose as an adjunct pool would.
Maybe every academic is faking their data, you want to hire someone who is happy within the first year to start faking data because someone who is reluctant is dangerous.
And you want people to move around because otherwise fellow adjuncts will notice their not very brilliant has suddenly made great discoveries and the simplest explanation is fraud.
Probably overly cynical but I think it ties into a world where only 17% of authors provide data when it is requested.
The peak oil folks would say that this is just a normal symptom of increasing hierarchical behavior under conditions of declining real resources.
The top research universities do compete against one another and do bid up salaries, bonuses (aka summer salary), and other perks for the top people. At the same time, there are norms against the ratio of top to bottom salaries within rank -- typically top tenured faculty will receive 3-4x the bottom in the same discipline and rank. Top academic salaries now run in excess of 400k for 9 months.
Thank you for stating the obvious.
One of the things I have long espoused, in the sense of thinking it but literally never telling anyone, is that the United States Congress should establish a system of National Teaching Universities that do exclusively undergrad and non-degree training. Recruit early- and mid-career academics with modest conventional prospects by using "Moneyball"-style strategies on the assumption potentially great teachers are being underutilized. Recruit outright non-academics, such as senior enlisted military personnel, who have an extremely deep reservoir of meta-knowledge about how to train people that is underutilized in the white-collar world for the crasset possible class reasons. Completely exclude research from the main career path and consider imposing outright guild-like rules preventing your hires from even publishing research while they're employed with you, and perhaps for six months afterwards. Require everyone to go through extensive, demanding paid training from which many drop out, and make sure all the training materials and training apparatus is built from private-sector and military people uncontaminated by the schoolmarm-industrial complex and its fraudulent "education research."
Basically, the bargain here is just that if surplus academics are willing to compromise SOME of their self-regarding and comically left-wing "ideals," move to some underserved working-class community, and actually make themselves useful, while abstaining from the corrupt academic system somewhat in potentially controversial ways, right-wing taxpayers should and reasonably will open their wallets and hand out genuinely life-changing compensation. It would require legal cover because the academic bums have written their special privileges into the legal system itself in deep ways, but that's why you have a legislature. You can literally just change the law to fix things. Conservatives control the Supreme Court and it would be pretty perverse and surprising if they actually sabotaged a program like this by enforcing special rights for academic dweebs who want nonsense guild rules enforced as a matter of first-rank constitutional principle.
Nobody is going to do this because conservatives fundamentally do not believe it is possible to fix anything and do not care. Their idea is just simple-minded buffoonery involving a rotating cast of villains, always tied to modernity, diversity, and social change in some vague shifting yet obvious way, who are thought to be corrupting the otherwise pristine and workable system through secret collusion. So even though solutions like this are actually kind of obvious and plausible (defuse social tensions by paying surplus academics enormous amounts of money to educate people who could actually use an education to do something, as opposed to just to sustain the education system) you will never see a Republican legislator ever propose something like this. Because their social base is the rural grandees and local bigwigs who claim to speak on behalf of dispossessed working class folks while privately colluding to cheat them so they can send their kids to Harvard as they always have.
This all seems right on. Particularly as regards obscuring who is doing the actual teaching.
I think the first graph's caption is wrong?
"Acceptance rate for tenure-track positions by how long it’s been since the candidate has gotten their PhD."
The chart doesn't show the rate, it shows the number of hires. That's very different: It depends on the number of applicants, and five years from PhD is a very weird time to be applying for tenure track jobs (given that you're close to tenure, but probably not close enough to be hired with tenure.) And it's from the American Historical Society, which is not likely to be representative of academia as a whole.
“Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?”
This does happen. First as post docs in many scientific disciplines. And then you are hired w/o tenure on a 2-4 yr contract, hopefully to get tenure at year 6-9 (varying on field and institution).
Experience isn't always good. Fresh faces bring enthusiasm, attention and a joy for life. If the numbers really matter, bring in the old hands. If it's more of a wishy-washy social thing, out with the old and in with the new.
Plus, in the 2-tiered system, do professors really want a similarly-experienced, worse version of themselves challenging their authority, angling for their job, or to feel sorry for?
Plus plus, and this is the big one, in the current system, students only see academic success stories and aspiring successes. Hiring old adjuncts is putting the losers of the process on display, (potentially) discouraging students from entering the meat grinder themselves.
“The value of the #1 brightest new PhD is that she has a 5% chance of becoming a future superstar ... Once you’ve been around for five years, colleges can see your track record, satisfy themselves you’re not the next Einstein, and lose interest.”
That’s a large part of it. So much hiring into the entry-level tenure track is based on buzz, and any university with pretensions is hoping they will lock in (even for a few years) the next superstar. They would rather take a chance on that than hire a solid performer 5 years out who would do a fine job but probably not become a household name. Some of this becomes a self-fulfilling prophecy; someone who starts their career teaching four courses a semester at a second-rate place with little or no support is trapped. Publishing your way out of that (no grant writers helping you, poor lab facilities, few good students or research-active colleagues, paltry travel funds) is like climbing Everest.
At the senior level there is poaching, but again, only for the creme de la creme. Otherwise I think there is (1) age discrimination, and (2) the fact that new PhDs are (relatively) cheap and senior faculty might not expect big bucks, but more than entry-level pay.
> Here’s another question that confuses me even more: Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?
That's what grad school is for.
There's another, more mechanistic explanation for why adjuncts who transition to tenure track faculty do so outside of their own university.
If you were to rank the university positions by prestige you might get a ranking which looks like
1) First Tier Tenure Track
2) Second Tier Tenure Track
3) First Tier Adjunct
4) Second Tier Adjunct
Adjuncts who get promoted one level up either get more prestigious adjunct positions or get tenure track at a lower ranked university. Getting tenure from the place you are adjuncting at is harder because it requires getting promoted two rungs up the ladder.
"Supply + demand" and "weird path dependence due to historical circumstances."
Maybe the most interesting book on the topic is Louis Menand’s *The Marketplace of Ideas: Reform and Resistance in the American University* (my comments are at https://jakeseliger.com/2010/01/21/problems-in-the-academy-louis-menands-the-marketplace-of-ideas-reform-and-resistance-in-the-american-university-2/). Basically, the academic job market worked reasonably well for would-be professors up until around 1975, when the number of undergrads fell (baby boomers worked their way through the system) but the number of PhD-granting institutions and programs continued to hum along. Since then, in most but not all fields, the number of PhDs has way outstripped the number of slots. When supply exceeds demand, strange things happen.
Tenure is the "weird path dependence due to historical circumstances" thing. Probably no one would set up the system the way it is today if one were starting from scratch. Tenure worked somewhat okay when lifespans were shorter and mandatory retirement hadn't been struck down the Supreme Court (one old article: https://www.nytimes.com/1994/06/15/us/new-law-against-age-bias-on-campus-clogs-academic-pipeline-critics-say.html).
Schools, like many employers, will change when they're forced to change (https://jakeseliger.com/2016/02/25/universities-treat-adjuncts-like-they-do-because-they-can/), but they tend to have a surfeit of applicants relative to jobs in most fields that aren't things like computer science and electrical engineering.
The most interesting question is why people keep starting PhD programs that are designed to produce all-but-dissertation "students" who can teach classes for very little money, and adjuncts. Granted, I was one of them, at one point.
Maybe academia is the brainiac's version of gambling in Vegas: likely to lead to losses, but fun in the moment.
Assoc. Professor at major R1 "brand name" university here (~12 years experience). This is mostly right. A few things:
1. STEM is very different from non-STEM. Title is the same but there is little actual overlap in what the job is like day-to-day. I am in STEM and run a lab with 20 graduate students and post-docs. This just doesn't happen on the "other side of campus".
2. Superstar effect is a big explainer. People want Steph Curry, and are willing to fight over potential candidates that have that potential. Experience does not turn [insert middling player] into Steph Curry.
3. You shouldn't think of it like you are locking someone in to a career path at the first blush. They have gone through (often) 10-12 years of college/grad school/post-doc by the time they even try for TT. Like it or not, potential is seen as at least somewhat baked in by the time they finish their PhD or post-doc. Remember, they are often nearing 30 years old at that point. In other industries this would be like a promotion of your leading 30 year olds into a management or leadership track.
4. Teaching is a small part of my job. Not going to give an effort/time percentage, but in practice it is low. This is why there are effectively "two markets" for TT and adjunct.
5. Adjuncts are treated horribly. Thankfully we have very few here (remember, "name brand" university). It just kind of sucks, and is very similar to "winner take all" dynamics happening more broadly in the economy.
Here in Germany, 1) you have to change universities - "to avoid intellectual incest". You can not simply do your Master at Humboldt-Uni, then your PhD AND then get tenure at the same place. Some faculties solve this by close cooperation with one other department at another college/Universität - swapping their presumably best and brightest. 2) There are now some laws to stop colleges to give their cheaper teaching-staff just one 2-year-contract after another - now it's either end the co-operation or give them a long-term contract (not tenure, but quite secure).
I was never going for that rat-race (way too dumb for it, too). Still, in my early career I got some very fine&well paid teaching jobs at the DAAD (German academic exchange service), though still kinda inexperienced - but 5% of us turned out fine PhDs or even professors (id est: valuable alumnae). When at 40 it was obvious to them I just wanted to continue being a good teacher, they closed the door. Which is fine; grateful for 11+ years on German tax-euros.
Here is another question, not addressed here. Universities rarely fire people. I know they cannot fire tenured people. But many of the people hired to just teach are really bad at the teaching part, but it is very uncommon (compared to the private sector) to let someone go. Whereas on the tenure track there is real accountability for grants and publications, off it there is almost no accountability for teaching quality.
This seems very simple to me. You've worked for me for several years without getting tenure. You are likely to continue to do so. If I am going to offer someone tenure, it should be an individual I believe that I would not be able to entice without that offer.
For one thing, football and basketball teams are way more visible to way more people than all but the very best researchers, and frequently not even them. And you can't really 'watch' researchers, even excellent ones, do their research in competition with others.
And in my discipline (Political Science) about 98% of the research is just stone stupid. (I can say that. I'm retired.) Irreproducible, statistically incoherent, and generally of little practical consequence.
Having said that, I've had friends do good, solid work because they enjoy the topic, but the topic is not 'sexy' so it just disappears. I have worked on some very good stuff but even that has pitfalls if it doesn't adhere to whatever the current narrative is.
Anyway. I was a "late bloomer" and got a full-time adjunct position that paid very well and had great benefits. But I negotiated that first contract before being forced to join a union. So, on average I got paid about 3 times what my fellow union members got. (Thanks, AFT.) Also, I was qualified to teach courses that were in high demand. Very high compared to supply.
After about 5 years (I was then getting 3-year contracts) I found that I had too little time to do both good teaching and good research (I contributed causal modeling on the work of others) plus I got asked to do something I considered unethical, so I dropped the research and spent the rest of my career just teaching. When an appropriate tenure track positioned opened (3 times) in my department I was asked each time to apply, with some assurance that I would get the position. By then I liked what I was doing, so I refused. Why do relatively worthless research just to gain tenure? I had the equivalent by being very, very good at my job, and a lack of easy replacements.
Look. We really need a two-track tenure process. One for research, and one for teaching. I know more than a few good researchers who aren't that good in the classroom. Why not reward both more appropriately? I also know that that currently tenured loathe that idea.
Also, remember that I was at a second-tier state school. Maybe around #150. We were good at what we did when we concentrated on it. But we were NEVER going to get a #1-100 new PhD. Sometimes they would apply, but they just used the interviews to bounce to the next one at a school they liked.
So. Why did my school try to hire the same way? Because the top admins are just as socially conscious as the top at Top 25 schools. Plus, to the extent that they can hire adjuncts cheap, it looks like they save money. (I'm not convinced of that, but that's a whole nother story.
I do feel bad for those who take adjunct positions in the hopes of landing a tenure track job. But several of our ABD adjuncts have made that happen, and one full-time adjunct colleague landed a tenure track job at a decent private college. But it's kind or rare.
We also pump out new PhDs for lots of bad reasons, and at least in my discipline we've been told since at least the '70s that a "wave of retirements" was coming and there'd be lots of jobs. But we study politics, so you should expect lots of lies.
Also, many "intellectuals" seem to enjoy having lots of indentured servants around them. That's a psychological problem that I'm not competent to help with.
The analysis in part I is spot on. As for why they only hire inexperienced people, I would add that adjuncts have experience in the wrong job. They have experience teaching, but excellence in teaching is not what is being hired for. They don't have experience in research (because adjuncting does not provide the time or resources to seriously pursue research, in general cases). Plus, there is an inverse signalling problem. You see that someone is an adjunct and you conclude that they were not good enough at research to get a tenure track job. So experience as an adjunct has zero actual value, and negative signalling value. And the more time you have spent as an adjunct the more negative signalling value this has. (`Jones was so hopeless he couldn't get a tenure track job even after ten years of trying!). Although practically speaking, the negative signalling value of any time as an adjunct at all is so strong that spending accepting an adjunct appointment is basically a kiss of death to chances of landing on tenure track.
In my fairly limited experience, it's less the case that teaching-track faculty never get hired for tenure-track jobs at their own institutions than that they rarely get hired for tenure-track jobs at any comparable institutions.
Academia is very hierarchical, and pay, teaching workload, research resources, and prestige are all very different if you're at a top-20 institution, a top-100 institution, or a non-R-1 institution. In my field (economics), pay and teaching load for permanent teaching-track positions at top universities are not too different from pay and teaching load for tenure-track positions at lower-tier institutions.
When I was on the job market, I was considered for tenure-track jobs at some top-100 research universities and some liberal arts colleges. I was also considered for teaching positions at elite universities. If I'd taken one of the teaching positions, done really good work, and gone back on the job market, it would be reasonable to expect that the jobs I'd be competitive for would be in the same basic ballpark as the ones I was competitive for when I went on the market the first time. It would require a pretty dramatic change in my productivity and quality to go from someone that the 100th best institution in the country would consider to someone that the 10th best institution in the country would consider.
As for why academia is so hierarchical, I'd guess that it's a combination of the fact that the most skillful/productive academics like to work with other very skillful and productive academics, plus the fact that outsiders (i.e. the people funding the work through their tuition, donations, or grants) don't know as much as insiders about what constitutes good work, so rely on academics giving each other seals of approval to know who to trust and who to give money to.
As for why superstars are not so mobile - there is a key difference to sports stars, in that sports stars are typically single and in their twenties, whereas tenured professors are middle aged, typically with working spouses and kids, and this makes them significantly less open to geographic moves (and very seldom are there two really good universities in the same geographic area). Also, for experimentalists, moving means recreating a lab from scratch which is a multi year undertaking (in which time you are not doing any research and thus losing prestige), whereas even for theorists you have developed a system that works for you, which presumably involves some web of local collaborations and so forth, and moving means tearing that up. People do move certainly, just the friction is greater.
There's also a second-order effect here: once grad students know that the odds of getting tenure-track positions are highest in the 12 months after they go on the job market, their incentive is to optimize all their work in progress to maximize the amount of publications, etc. that they have done as of when they go on the market. So they often have relatively little incremental publications that will occur after that 1st year, because they've focused on getting as big of a python lump of work out at once, rather than establishing a steady stream of incremental publications after that date.
As for your closing question: universities can and do engage in bidding wars, trying to outbid each other for superstars. However, there are a couple of factors constraining this (1) morale issues (plus pay transparency, which is the norm at universities) means you are somewhat constrained in how much variation in pay you can have (you can and do have factors of a few, just typically not orders of magnitude). (2) entering a bidding war and losing has reputational costs (university X tried to get Jones and failed, guess X sucks), which makes them reluctant to enter a bidding war unless they have reasonable costs of success. (This then gets compounded with principal-agent considerations. Whoever pushed to try and hire Jones personally loses a ton of political capital if Jones declines).
Why can't you get hired at the institution you got your PhD in ?
That's mostly to ensure that institutions at least try to hire good candidates rather than their own students. It also help with intellectual cross fertilisation - people from outside can bring in knowledge and methods that are not available in the department.
Why do they only hire inexperienced people ?
It seems like something particular to either the US or social sciences. In my field (maths) and my country (France) you typically only get a permanent position after a couple of postdoc positions (hiring committees get suspicious if you have not been hired after 5-7 years so being to experienced is not good either).
A lot of this agrees with stuff already said by others, but:
1. The pay difference between adjuncts and research/tenure-track faculty is largely about "research success is what drives prestige", as you say.
2. The "negative premia" for experience is basically just observing that people (mostly) go on the academic job market when they are completing their PhD, some succeed and get TT jobs (or in fields where postdocs are ubiquitous, good post docs that sometime after hopefully lead to TT jobs), others fail to and get adjunct roles / leave academia. For an adjunct after that point, there is a massive negative selection problem, plus as you say it's hard to do good research while teaching 6 classes a year.
3. "But then why do they only hire inexperienced people?" In fields where post-docs are ubiquitous, obviously people have more experience when applying for TT jobs (but if a post-doc counts as experience, why doesn't a PhD? Isn't the last few years of a PhD relevant job experience which people are judged on for TT hiring?) But ignoring post-docs, there's an initial clearing market of new PhD grads, some get TT job offers, others don't. (See 2). Beyond that, obviously they don't "only hire inexperienced people". If a TT professor is successful at lower-ranked university X, higher ranked university Y will poach them, but with a promotion to Associate or Full Professor. That people hired at Assistant Professor level tend to be inexperienced is just synonymous with the job title.
4. "Why only people from outside their own institution?" This is sold in part as being about avoiding nepotism. And in part its a "go into the wild word and prove yourself, and then if you are successful we will hire you back". In equilibrium it often works, especially seeing 99% of the time the school someone did their PhD at is higher ranked than their AP placement.
5. "Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?" You could say this is what post-docs are. You could sort of say this is what PhD students are. You could also say "yeah, but why do the top universities bother to have Assistant Professors at all, rather than waiting to see who becomes a star and then hiring them?" The latter is an interesting question (especially given that a large share of TT faculty at top schools don't get tenure). My guess is part of it is historical tradition, and part of it is that faculty will tend to show a decent amount of loyalty to their current employer, so you get an "unravelling equilibrium" in that the most promising candidates are poached early, because it's much harder to poach later on (who actually wants to upend their entire family life to move city for a modest pay bump at a slightly more prestigious university, especially if they are already successful. Just get current university to match the pay bump!). Part of it might also be that senior faculty like having productive junior faculty around. And another part might be that a lot of the most influential research is produced by faculty while they are fairly young, and the departments might care about this (even if it's too inside baseball for it to influence "prestige" in the public's mind).
I think in modern academia the incentives are so misaligned that it's hard to really model it in normal economic terms.
Students are mainly there for signaling purposes, not to actually learn stuff. And the tuition is being paid for out of subsidized loans. But also, the institution is nominally a nonprofit, in many cases literally run by the state. But also, the departments are usually in charge of hiring. The main department stakeholders are the tenured faculty, not the people nominally in charge of the university as a whole.
IMO this makes academia really susceptible to fads and trends. There's not really any strong competitive pressure to impose discipline. I don't necessarily think there's any "rational" reason for it to work how it does. There are definitely *some* structural elements in play, like budget cuts. But the structural economic elements dont on their own determine all the specifics that develop in a relatively small and close-knit community.
this is a huge topic but there's a few things that make easy analysis different.
1) as many have pointed out already, STEM is very different from non-STEM. I'd further subdivide things - fields have really different hiring patterns, adjunct use, etc. and that has a big impact on the way the market works. Comparing the market for a law professor to those for an English professor to an engineering professor to a chemistry professor is at least several kinds of apples, and probably a mixture of fruits and vegetables. Then there are multidisciplinary things (like public policy schools that hire economists, sociologists, political scientists, etc.) where things are even weirder b/c multiple strands come together in a weird mix.
2) schools differ a lot - someone has already noted the difference between R1 and others, but it is even more differentiated - the very top private schools behave very differently in hiring than do large state flagship schools even if both are R1s. There are whole separate "teaching institution" markets, the 2nd tier (which no one would admit to being in) research schools, community colleges, non-flagship state schools, liberal arts schools, and so on. Each submarket operates differently.
3) put those two things together and you have a large number of submarkets that operate very differently in hiring, use of adjuncts, etc.
Then we get in to allocating slots for TT faculty at a university. Some disciplines are going up in enrollment (Data Science); some teach huge "service" courses (English) rather than / in addition to large numbers of majors; some are going down (Classics). When a university hires a TT faculty member, that's a long term commitment to funding a slot. Some slots are cheap (English), some are really expensive (STEM). Administrators want flexibility in adjusting staff - which leads to more reliance on adjuncts and/or "APT" or "Academic Professional Track" professors, who generally don't write much and teach more, serve on committees. These folks get contracts from a year to longer. But they allow adjusting head counts. And most university central administrators know the demographic cliff is coming in terms of students for many places and want to keep flexibility to adapt if their institution's enrollment drops (as it will for many schools).
And then there's the need to bring in adjuncts to provide "real world" perspectives. Not much of a need in, say, English for that, but there is in law, business, politics (sometimes), etc.
There's more too - the academic job market is really a lot of submarkets which behave differently. Many of them are quirky at best, dysfunctional at worst. But you have to look at the institutional details of each to tease out why they behave the way they do (at least to some extent).
If I'm reading that graph correctly then it absolutely does *not* reflect my field (astronomy). Almost no one gets a tenure-track job within one year of earning their PhD. Nearly everyone does a postdoc, which in the US usually lasts 3 years. I don't have data in hand but I would guess that the median tenure-track professor in astronomy is somewhere around 3-5 years after getting their PhD. I think physics is very similar. Is this really that unusual?
One advantage may be that Universities are both consumers and producers of new PHDs. If the only way to get into Tenure Track was to suffer through years of adjunct teaching, would the PHD > Tenure Track career route be as popular?
As an outsider it seems that people going for PHDs in the hopes of staying in Academia are playing the odds that they are the lucky few that get a first year Tenure track posting, and if not either leave Academia or are stuck as an Adjunct.
From the University's perspective, the value of not shifting senior adjuncts to tenure track is that the tenure carrot can stay open to as many hopefully PHD students as possible, many of whom will end up spending years on a PHD with no tenure track to show for it. At that point some have no better option than to teach under whatever terms they can get.
I'm a former theoretical physics postdoc in the UK. My wife is currently a postdoc in biosciences at a top UK University. Here is my two pence:
1) Einstein didn't work in a lab. I get the point, but I had to say it! Sorry.
2) In my experience universities don't usually differentiate their workforce like that. Rather, most experienced people get forced to teach, early career people not so much, and only the very few (maybe <10%) get to avoid it (usually not by negotiating with universities but rather by getting grants which bring money and prestige and happen to force them to spend their time not teaching). And usually the more experienced and well-paid you are the less time you get to spend doing research rather than writing grants, supervising, and teaching.
3) Having talked to people on the inside during hiring processes, I don't think it is accurate to describe hiring people out of their PhDs as trying to find the people with the highest chances of being the next superstar. People usually try to screen for some measure of quality, but typically if you meet some minimum requirements (like having published in decent journals and/or having graduated from decent institutions) then what they try to optimise for is complementarity of research interests. Basically, they care more about how your work will fit with the work of other people in their groups than about whether you are a bit more likely to be the next Einstein. If you would be a good collaborator, who has a track record in lines of investigation which don't overlap too much with the ones they have going but which have potential for interacting with them, then they'll want you. (Bonus points if your niche area of expertise happens to be hot at the moment, where "hotness" often lasts less than the average PhD.) Which makes sense. Why hope for Einstein when you can more reliably aim for a multiplicative impact now?
4) Even the people in the dream superstar jobs are usually earning not that much compared to industry - or even non-academic techy jobs in the public sector.
I don't have data on this, but there's a similar phenomenon in the publishing world. A debut author with a book that sounds exciting and topical will have a decent shot at a six-figure book deal, and those books frequently go to auction. But once you're published, if you're among the majority of authors whose books don't turn into juggernauts, your advances drop dramatically, with some previously well-paid authors grateful to get $20k or $30k for a new project.
This is a "shiny new thing" phenomenon - an exciting new author has enormous potential... until their book hits shelves and they don't hit bestseller lists. My initial guess is that hiring committees are susceptible to the appeal of the "shiny new thing" and get so excited by the boundless potential of all the newly-minted PhDs that the adjuncts who've been slogging for years now just don't stand a chance.
After a couple of disappointing books, an author will often change genre or age category, choose a pen name to publish under, and go to a new publisher to leave behind the stink of their sales record. This can be a way to reboot their career, re-invigorate the marketing department (who will have given up on them at their previous publisher), and try to capture some of that "shiny new thing" energy again. It's interesting you mentioned that successful teaching- or adjunct-track professors will go somewhere else for a tenure-track position when they make a breakthrough, and feels related to this.
A little concerning that an industry like publishing that acquires talent almost entirely on *vibes and sparkle* appears to have similar patterns to academia!
As someone who's spent ages 15-41 in academia: you're *very* far from wrong, but it's also not as drastic as someone reading this might fear. Academia isn't sharply divided between "Einsteins" who teach only occasional upper-level seminars, and poorly-paid adjuncts who teach huge numbers of intro classes. There's also a large "middle class," e.g. all the non-Einstein tenure-track faculty, which both does research AND interacts regularly with undergrads, sometimes even at the same time (at least in CS, lots of undergrads get involved in research). Maybe that just helps confuse prospective students all the more! :-)
(Also, agree with the other commenters that postdocs are now a huge part of the story.)
Sports team analogy is probably even more apt here than Scott realizes, especially salary capped and contractually strict ones like the NBA. Just like in academia, in the NBA a rookie who has the chance to develop into the superstar is far more valuable than an established veteran who has proven themselves to be a reliable above-average player. This is true even if, on average, the rookies do not become above-average players. This is reflected on the trade market, where the #5 pick in the draft (which becomes a first-year player in the league once drafted) is valued more highly in a vacuum than solid rotation veterans, even if "solid rotation veteran" is actually probably the EV of what you get from the #5 pick.
There are additional wrinkles and dimensions to this, as rookies are also on a more affordable rookie scale contract, NBA players are locked up into defined contract lengths than is difficult for the player to exit early for other teams, etc. The limited free agency explains superstar inertia in professional sports - there are typically only a few defined windows in a player's career where they truly are pure free agents. But "superstar inertia" I don't think is very difficult to explain for most tenure-track academics. Unless you are in the true 1% of academia there's a lot of disruption to uprooting your life and your intellectual networks to go somewhere use and the upside needs to be very high to entice you to do so. Given the natural geographic distribution of academic employers it's not even like other working professionals job-hopping to go to the competitor firm down the street. And it seems to me that you do see the true top 1% of tenure track academia, celebrity public intellectuals etc., move around, though perhaps my perception is shaded by the fact that these moves are naturally going to be more newsworthy than random tenure track faculty moves.
Some of what's going on is historical, going back to about 1950, when the Feds started throwing money at universities to do research. The NSF was founded c. 1952 b/c everyone realized that science had won WW II thus needed serious funding. The question was HOW, and we got a bastard version of the German system and our old tradition of colleges being for educating elite youth. Remember that many of the important scientists of WW II were German. There, all universities were research universities--there was no UG education as such--and their goal was to produce research and prepare scholars/scientists. They were supported by the German state, and departments were given budgets they could spend as they chose (and there was, basically, just 1 fully tenured "Ordinary Professor" per department, and he ran it, often like a fief. That was what US scientists envisioned after the war--fund the best scientists directly to produce the best science. But this met howls of protest from most US colleges & universities who feared all the money would go to the usual suspects, Harvard, Yale, Illinois (go Illini!), etc. And it was unAmerican. States urged the money go to states for allocation to their schools. The scientists howled -- nobodies at places like the Southern University of North Dakota at Hoople would get, and waste, scientific research dollars. Bad Science would be done!! The compromise is what we have today is the result, anyone, even a nobody at SND@H, could submit a grant, and if it was, in fact Best Science, it would be funded. So, virtually every school in the US tries to be a research oriented school, not just for the prestige, but because of the sweet overhead money that comes in, helping fund student-attracting facilities like salt water pools and paying big money to sports coaches.
The adjunct vs. research professor divide reflects the old teaching-college vs. research institution divide in the US. If you read what the poor adjuncts write, they generally thought of themselves as teachers expanding and informing the minds of their students and did not want to be researchers. They suffer for pursuing their dream. Between that and high course loads, they don't do research and so can't transition to the main tenure track. They'd like to be paid well and have job security to do what they love, teach.
Top stars get hired (note rise in numbers for older faculty) when a university wants to set up a new field of study or some "center." Then they go for an established name who will attract bright and ambitious folks quickly.
I'm an odd duck in this case. I wanted to be an UG teacher back in 1974 (my mentor said, "I suspected that of you."), before the Federal money started streaming to my field, psychology. But there were no good jobs available then, so I took one at a comprehensive university that then got ambitious and shot up to R1 status remarkably quickly. So I had to get with the program and do research, and did get tenure, though I could not today, as I never got grants.
"colleges have decided they only need a certain amount of prestige before they stop caring and fill the other teaching positions with warm bodies" - No, it's more that they buy as much prestige as they can afford, but (in part because its hard to hire good research faculty while requiring them to do lots of teaching) that isn't enough in order to meet all their teaching obligations.
"Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?" - This is what the best universities do - they hire "tenure track" professors, but very rarely grant them tenure - instead, their tenured faculty are maximally prestigious people poached from elsewhere. This works for the best universities, because being tenure track at Harvard (with the expectation that you probably won't actually get tenure) is a very good deal for the very best new PhDs. It doesn't work at other universities because being tenure track at Podunk College, with the expectation that you won't even get tenure out of it, is not an attractive offer, and so good people don't accept offers at lesser schools if they expect it won't lead to tenure (assuming they have a better option elsewhere).
Note that Devereaux is writing about the *history* job market, not all of this applies to academia in general. Like, I've never heard of the whole "experience actually hurts" phenomenon in mathematics.
Also worth noting is that in math and various sciences, it's typical to do a few postdocs (which doesn't fit into any of Devereaux's categories; these are neither tenure track nor teaching track nor adjunct) before getting a tenure track position. Devereaux mentions this but *mostly* ignores it because he's mostly focusing on history.
So I don't know if the market is as broken as in other fields.
As for history, though -- I do think you may be making a mistake in assuming that people are necessarily doing things for good reasons rather than due to biases. Devereaux explains the "experience hurts" phenomenon via a bias for hiring from prestigious colleges; I think that answer is definitely worth considering.
I think academia and performing arts have a lot more in common than you'd think. A record label and an academic department both want to sign someone that will HIT. But you have no actual idea what prospects will do it. For recording artists, you can give a thousand advances and if even one hits it will pay for the rest. But you can't do that for academic departments. So you make choices based on what you can actually observe (a degree from a top 10 program) and hope that will give you a better chance at a hit.
Why do departments only hire people with no experience? The same reason that there are no 35-year-old breakout pop stars. Labels and departments buy based on POTENTIAL. Someone with experience has no upside.
> What process carves off 30% of professors to get good pay and benefits, but passes over the rest?
You're neglecting the influence of unions and worker activism. The institution, like most employers, is coordinated and can apply power effectively with ease. It has most of the power. Workers can only exert pressure by leaving (and going to another institution that is likely similar) or striking (difficult to coordinate, pisses people off, risky, etc.).
Often, the low-paid teaching staff are contractors, have little clout in the union if they're in it at all, and may well be gone in a year or two. The union is incentivized to do the most for its long-term, prestigious members, and so it uses its power mostly for them.
You end up with a scenario where the uni would like to pay everyone crap, the workers would like everyone paid well, and the way the power imbalance/application of political capital shakes out is that some small amount of workers are paid well and the rest are paid crap.
I don't think your explanation about trying to trick students about who is teaching them works, or at least it's not complete. Disclaimer that this is based on my anecdotal experience in astronomy.
Teaching professors are a relatively new phenomenon but colleges have existed and have been teaching students with something like the current model in the US for decades. Why would they only now start trying to offload teaching responsibilities onto a different class of professor behind students' backs? Also, the teaching professors that I'm familiar actually have the title "teaching professor" right there on the department webpage, so if they are trying to hide this from students they are doing a poor job.
Also, in my field, adjuncts are people who have a primary appointment elsewhere but who then get a secondary affiliation with some college or university, usually so they can directly advise grad students, and sometimes as a way of funneling research funding through the university. They aren't paid positions and don't come with any teaching or department service responsibilities. I think one needs to be careful about differentiating between very different positions with the same or similar titles when doing this type of analysis.
One element you might be missing is the nepotism angle. My partner was a an academic who recently left academia, and one thing they claim is an open secret in academia is that most new tenure track hires have parents who were tenure track hires before them. There could be reasons other than direct nepotism that account for this (the children of tenure track people are more likely to go into the field and be suited for it, and also they know the most common pitfalls to navigate), but however you parse it, it makes sense to me that in a system that has a nepotism component, you see a lot of people being snatched for high-prestige jobs the moment they become eligible for those jobs, not the moment they have a proven track record (because nepotism generally works counter to meritocracy).
It would be interesting to do a deep comparison of the academic and entertainment industries, because entertainment is a famously nepotistic industry where lots of people toil in competition for low-paying, shitty gigs and a few people rise to the top, and those people are often related to other people in the industry.
"Why only people from outside their own institution?"
There's a strong social tradition of avoiding 'academic inbreeding.' Hiring your own graduates is seen as both an obvious form of nepotism and a recipe for intellectual stagnation. Not that it doesn't happen, when a favored student can't find a job.
I'm tenured at an R-1 equivalent uni in Europe.
Adjuncts are de facto on another planet, and tt assistant professors are by design junior, i.e. you typically hire them immediately after their PhD or a few years of postdoc. In many countries, only associate and full professors are public servants like judges and diplomats, i.e. not on a contract.
Teaching is incidental, and can easily be done by graduate students or postdocs. What matters is research, grant funding, and prestige. In my experience, poaching exists but up to a point, especially in Europe where there is less room for negotiation.
I don't know about teaching-focused institutions.
Hypothetically, a person doing a job doesn't look like they can do more challenging work, at least if management is sufficiently unimaginative. So they prefer shiny new people.
This seems to import the assumption that "Einstein/superstar" researchers are a thing outside of STEM. I don't know if this is true -- I'm STEM myself -- but it strikes me as counterintuitive in a field like ancient history (Bret's field) where research is driven more by laborious study than clever ideas. What is an Einstein-equivalent in history or English literature *doing* to be so valuable?
How do you discover someone is a good researcher while 1) soaking up their time teaching and 2) not giving them any significant field-appropriate resources for researching?
I think the explanation mostly is institutional capture of the hiring methods and standards by the tenure-track profs. They were hired into the prestigious TT in this manner, and they do the hiring/control the hiring of TT positions, and if the methods or standards are questioned that would question their entitlement to the prestigious TT jobs they have.
Just think of 'adjunct professor' as 'substitute teacher' and all shall be clear.
Scott: Your question "Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?" has an easy answer.
Once you're hired as an adjunct you really have no hope (except in the most unusual non-STEM circumstances) of getting a good TT job. You have to start producing amazing research/scholarship almost immediately, and the adjunct grind just makes this impossible.
I am a non-tenure track research professor (meaning I don't teach classes) at an R1 institution and have been an adjunct professor at a different college. I can share some insights into hiring decisions.
There is a bias toward hiring young people for tenure track positions for several reasons:
1. They are trying to hire people who will be making a name for themselves at that institution. Hiring committees do look at a persons career trajectory in addition to accomplishments.
2. Faculty want to make sure their department is in on whatever the current trend is in order to appeal to grad students and postdocs. Naturally, the best way to buy in to such trends is to hire young people.
3. There is an extreme level of concern about diversity. Since diversity is increasing, the most diverse subset of candidates is the youngest group. This leads to a strong bias toward hiring younger people to faculty positions. (This is probably one of the most important factors and I have known cases where it was explicitly stated that no white men would be hired. That clearly ruled out a lot of the more established candidates.)
Some of this post also assumes some popular stereotypes that are not true. One is that research and teaching are incompatible. Almost all the time a person who is good at research is also an excellent teacher. Besides which, students like having a professor who is active in their field. Researchers generally like teaching, in spite of the stereotype that researchers wish they could avoid teaching so they could focus on their research.
Universities hire adjuncts because they're cheap labor, not because they're expert teachers. Don't read too much into it. Teaching at the college level is really not difficult for someone who understands the subject. I think Scott has a post somewhere showing that most innovations in education are just showing selection effects. Faculty know this, even if they do not usually admit it. So, there is not much value to be gained from having people who are hired for teaching abilities.
A mildly snarky take on not offering TT to adjuncts in-house:
As a successful academic on the hiring committee you've successfully jumped through a large number of hoops, starting in high school. If you've hired someone as an adjunct already, your department's process has confirmed this person as a 2nd tier researcher. Giving them a TT position would imply a flaw in your hiring process (if they're good enough to be TT, why would they take an adjunct position in the first place?), the same hiring process which has confirmed you as a successful academic & we can't just go around calling that into question, now can we?
I'm a department chair at a state university (with unionized faculty), and have experience with both tenure-line and adjunct hiring.
When we hire a new adjunct, the pool is generally limited to people who are already in the area and open to more employment. Usually, in an adjunct's first semester, we can only offer them part time work. This is partly because the collective bargaining agreement requires that we offer classes to any qualified current adjuncts before hiring someone new. For example, if someone retires, gets sick, or goes on sabbatical, I have to offer their courses to any qualified current part-timers who would like to increase their teaching load. There might be a course or two left over for a new adjunct to start part-time. It's essentially impossible to offer a full-time position for a new adjunct.
In this situation, we can't cast a wide net for candidates. We aren't offering a level of compensation or security worth relocating for, and in many cases the need arises last-minute. So, we hire the most qualified person we can find locally, among people with the right credentials who neither got a tenure-track nor took a full-time job in industry. It's often difficult to find anyone.
For a tenure-track hire, by contrast, we get a large number of applications from across the country and internationally. If one of our current adjuncts is in the pool, they will get the same consideration as everyone else. But the odds that they will happen to be the best fit out of that large pool are very low.
An important point to consider is that in many fields (e.g. economics) adjuncts typically do not have PhDs, only masters degrees. As such they cannot move into tenure track or teaching track jobs. This is often raised as a point of contention, that adjuncts are not considered for such jobs, but they are also explicitly unqualified for the job. Adjuncts in these fields are in effect below even PhD candidates, as they are not even working on their dissertations, and more like regular grad students.
That isn't to say anything bad about adjuncts, just that simply going by job titles isn't putting like with like.
I should also point out that when I was in the job market it was common advice when applying for a "normal" tenure track economics position to not even admit that you like teaching. I use admit intentionally there, as it was seen as an admission that you were not serious about research. This is important, as most schools wanted researchers and offered extremely low teaching loads (something like 1 course a semester, maybe 2), one even promising TA support such that "you don't even have to interact with students, ever." Accreditation is easier if you have more courses taught by PhDs, however, so depending on how it is counted (number of courses taught by PhDs vs percent of faculty with a PhD) teaching track roles are becoming more valuable to counter balance the number of research only faculty.
I have a couple of other thoughts here that have been mentioned before, so I won't belabor them, but one that seems to be left out: networking.
My professors have explicitely told me that at some level, whether I can land a post-doc is a luck of the draw for how many of their friends have recently taken post-docs. From undergrad I moved to a grad school in large part through my mentor's contacts, and the assumption is that from grad school I would move to a post-doc largely through my advisor's contacts. This is a pretty common story in math, which seems to be a particularly small and close-knit community as opposed to many other STEM fields. While I am in graduate school, and for a year or two out, my advisors will be concerned with my success, as it reflects on them. After a few years, they will have other students to go to bat for, and I imagine that they would reasonably focus their efforts on landing their current students' jobs. The longer I'm out, the less contact I have with my advisors, and thus the harder I expect it will be for me to land a job.
DISCLAIMER: I think a lot of merit goes into getting grad school/post docs/tenure track jobs. I just also think that there's a lot of networking in there too; you sort of need both, or overwhelming amounts of merit.
I think Scott's basic explanation is correct. I'll just add tweak that I thinks clarifies the matter further.
As Scott notes, there are two distinct functions: research and teaching. In a model of perfect competition, salary (Y) would form a power law when plotted against the ratio of research-to-teaching (X), with the long tail extending toward 100% teaching. However, there is not perfect competition, because bureaucracy. For budgetary reasons, hiring and salary must be legible to administrators, so they segment the curve into distinct roles: tenure-track, teaching, and adjunct, resulting in the present system.
The power law in competition for researchers is also congruent with the preference for superstar researchers and the disfavoring of older adjuncts. But it's the bureaucratic dimension that prevents this from being rationalized into a system of free agents.
My experience is in the humanities, and it's also about 30 years out of date now. I had a tenure-track job right after I finished my dissertation. My husband was a lecturer for two years before getting a tenure-track gig at a different school. (There's another distinction for you -- lecturers make a real salary and get bennies, but they're one-year renewable hires. So better than adjuncts, but worse than TT faculty).
My observation is that all the universities assume that the best candidates will get snatched up right after they graduate. If they didn't get snatched up, well, then I guess they weren't the best candidates, right? Almost anyone you ask in academia in a one-on-one conversation will agree this is stupid and incorrect, but in committee meetings it's kind of the default assumption.
I'm female, and I also felt like there was a similarity between fresh, virginal candidates/fresh, virginal young women and slightly used, been-around-the-block candidates/been-around-the-block less-young women. Do I have any social science to back this up? Nope. It was just a feeling I had.
I eventually left academia so that I could live in the same state as my husband (who also left academia), and we both got higher-paid, lower-status jobs (lawyer and IT manager). In the same town! Mirabile dictu.
I guess I should also mention that humanities folks don't usually do post-docs, though there are exceptions.
I think you should question your approach of asking why colleges don't do the thing that would be more fair, rather than asking why any organization ever does the fair thing. No explanation is needed for colleges behaving unfairly. An explanation /is/ needed when one sees an organization behaving fairly.
As Adam Smith said, What needs explaining is not why so many nations are poor, but why some are rich.
Colleges aren't for-profit corporations, and my experience with organizations of all types has been that organizations that aren't run for profit, don't incentivize their people to be fair, or to do their jobs well in any respect. The only exception I've encountered are religious organizations, volunteer organizations, and Dept. of Defense front-line organizations, which all attract only people who are already incentivized to do the mission they're signing up for.
(You might think that the non-random behavior of colleges shows they're incentivized in some way. That's correct; but when white-collar workers aren't incentivized to do their job correctly, their personal preferences take over. These are highly correlated between different people in the same job, even in different organizations; so the behavior across organizations has a similar distribution.)
Substack tells me this comment is too long, so I will be splitting it up into two; the second half is in the reply. (Edited to remove some potentially identifying info and to soften some things I think I was too harsh on at first.
I was recently on the CS academic job market and successfully got a job. A few thoughts -- this is mostly focused on the top 30 CS departments in the US. Once you get out of there I don't know how much this holds anymore.
1. First of all, your caption on the graph doesn't actually match the chart. Assuming the chart label is right, that's not acceptance rate of people being hired, that's the *number* of people being hired at each rank. (Yes, these numbers are low, which matches my expectations for a specific field. See the annual Taubee survey for CS hiring stats in the US and Canada -- it's typical to hire about 200-400 professors in the US per year: https://cra.org/resources/taulbee-survey/ (last available for year 2021). I would be shocked if the acceptance rate for recent PhDs is 90 and the acceptance rate for people 1 year out (including postdocs, people who did a year of research prof, etc) was so much lower -- but the graph is just raw numbers as far as I can tell. The fact that the "Within Previous Year" bar is so high does not indicate that the acceptance rate for recent PhDs is higher. It just indicates that more of them were hired, which is not surprising considering it's by far the most common time to go on the job market.
2. Speaking about top R1 schools (let's say the top 30 computer science departments -- and everything I say here counts double for the top 10): The reason it becomes much harder to get a tenure-track job after becoming a teaching-focused professor (either teaching-track or adjunct) is that the application for tenure-track jobs is almost exclusively focused on your research. Aside from the allure of the "full package," I'm not sure why the charade is maintained that research is just one of three equally-weighted considerations in a tenure-track job (research, teaching, service). In reality it seems to be about 95% research, and as long as you don't stab a student or something you'll probably keep your job. (I will say: some places, including the place I ended up going, did actually have one teaching faculty interview me as a tenure-track faculty. This is rare and a good sign that this institution/department actually treats its teaching faculty well. But I think aside from that, I got asked exactly zero questions about teaching at R1 interviews. Yes, I had to provide a teaching statement that described my teaching philosophy. Did anyone actually read it? I have no idea, but I know they read my research statement, because I got questions about that.)
And if you are a PhD student or a postdoc, your full-time job is research. Most PhD students and postdocs don't work a day job, have only minor teaching responsibilities, and have collaborators and peers who are also doing research. Plus they (theoretically) have guidance on which well-scoped problems to work on from their advisors. So, PhD students and postdocs are basically in the best state to spend the 5+ years of run-up to a tenure track application. Doing research as a PhD is way easier than doing it on your own time as a hobby, if only because your university will pay to get you access to all the electronic library resources you could want, lab space, equipment, administrators, conference travel, seminars to watch, administrators, etc, plus you don't have to hold down another job while doing that. This is the biggest pool of people who want to become tenure-track professors, so this is everyone else is competing against.
Teaching as a teaching-track professor or adjunct is a full-time job. My understanding is that a typical teaching-track load is 2-3 classes per semester. The advice I've been given is that it takes 4-8 out-of-lecture hours to prepare materials (lecture, homework, "fun" activities, etc) per every hour of lecture -- not even including grading, office hours, helping students who want to chat, dealing with students who decided they hate you, dealing with cheating, paperwork, interfacing with the rest of the department, etc. As far as I can tell, it is one of the hardest jobs ever, and it doesn't pay nearly as well or have the prestige of being a research professor. But, if you want to be tenure-track faculty, you will be evaluated on very little of this, and realistically if you're teaching 2-3 classes per semester you will not have time to conduct the kind of research agenda that will help you compete with tenure-track candidates. The competition is the PhD students and postdocs (and maybe research professors) who *have* been spending full-time jobs on research, so even if you do a great job in the <25% of your day remaining to you after teaching, that's a *really* uphill battle. I've spoken with several teaching faculty and asked why they did that, and I got two reasons. Almost all of them: they love teaching! Which is great! One person said he liked academia and was thinking about the tenure-track life, but didn't want to be busy with the 20 jobs of being a tenure-track professor, so he went with a teaching job so he could be overloaded with only 10 jobs instead. I didn't talk to any adjuncts because I don't know any, which probably says something about how much they are overlooked in academia in general.
Maybe a more charitable way to look at this is: Research, most commonly done during a PhD/postdoc, *is* the experience they're looking for. Everything else, including teaching, is a "distraction." (I don't honestly know if it would be better if everyone gave up on the charade and had tenure-track faculty just stop teaching. More on that in another post later, perhaps.) And people who do research-focused jobs, especially in research-oriented academic roles (see next), or occasionally from particularly paper-publishing-focused industry labs, do occasionally switch into tenure track.
On the point of not hiring from your own institution I have seen one senior professor in economics claim they would do this specifically for purposes of diversity in intellectual thought. They argued that on principle their doctoral students should get experience somewhere else and that they should try and get people from elsewhere so that the intellectual climate didn't get stale. No idea if this if this value system is strong enough to override more base incentives, seems unlikely to me, but could be part of an explanation.
As a STEM PhD, in a very qualitative way I feel like I can tell who will be hired and who won’t pretty easily. The truth is you have to be exceptional to be hired in my field (not bio) and it’s easy to tell which phd grads will be chosen based on how many papers they published and who they worked with in graduate school. They are the real go getters and they are zealous.
Related to this, much of the hiring is based on network effects and prestige of the PhD institution. See here: https://www.science.org/doi/10.1126/sciadv.1400005
Hi Scott and thanks for this post. I'm an associate professor at a European R1-like university and I think there are relatively straightforward economic explanations as follows (with tounge partly in cheek).
As an academic in a research position, my main job is not to do good research, nor to teach classes. It is to bring home research grants. Overheads from these grants keep the lights on for everyone else. Therefore, the university needs to try to give tenured research positions only to people that can be expected with high confidence to earn more than they cost.
Next part of the explanation: a successful grantmaking career is a trajectory that needs to start almost immediately after getting a PhD. I am not sure whether it is rational for funders to allocate funds in this manner, but it is broadly perceived that this is how it works: more money comes to those that already have it, and if you don't get on the escalator from the start, it quickly gets much more difficult to climb up.
You ask why universities do not hire lots of people in a low-commitment position. But in a sense this is what the tenure track is. The salary is not high and you get kicked out after a few years if you haven't established a track record of external funding. Low risk with high upside for the university.
You also ask why colleges aren't more like sports teams, trying to buy each other's stars. This does happen to an extent. But a closer analogy might be to senior partners in law firms or management consultancies. If conditions are better elsewhere they may leave, but transaction costs are high, and most of the money anyway comes from what they bring in on their own.
Thus, I think the economics of a dependency on external grants goes a long way towards explaining the phenomena you describe.
I'd point out that the incentives at top institutions to care about superstar researchers run deeper than prestige. Top-tier profs pull in grant money, in many cases millions of dollars for departments (indirect costs on grants, i.e., the percentage of a total research grant that a uni takes in for its own operations rather than a prof's research is around 50%). The difference between A tier and B/C tier profs in this ability to pull in grant money is likely substantial in many fields that are heavily dependent on research funding given the many biases present in professor brand name recognition, exponential growth dynamics in citation counts, and knowing how to work the grant system, let alone differences in actual talent (which from my observations are substantial).
"But why aren’t colleges more like un-salary-capped sports teams, trying to outbid one another for their rivals’ superstars?"
Institutions do this (we always heard that many top-tier Stanford profs were paid in the low to mid six-figures). However, departmental research budgets are finite with high potential upside but definitely not startup go-to-the-moon upside (how much grant money can one prof pull in? How can you predict who will win a Nobel just from a track record early in someone's career? patent potential is rarer) Switching universities is a huge move and as always, you have reputational risks if you e.g. move around too frequently and lose research productivity. What motivates academics is mostly not money, but freedom to do research. This means freedom from teaching/departmental responsibilities as well as the ability to collaborate closely with other leaders in your field (different schools have different specialties and you may not be able to find similar collaboration opportunities at peer institutions no matter what they pay you). Academics (in the hard sciences/engineering fields at least) could generally get paid way more out in private industry.
There's only so much grant money to win. It could be an interesting question whether unis are optimized for winning grant money based on faculty counts (whether they're over or under hiring)
Some hires appear to simply be political appointments, like Noam Chomsky at University of Arizona. I know they never replaced Ed Abbey. But he was the eco terrorists' darling. Really? Noam Chomsky for Ed Abbey? I don't get it.
Chomsky merely mumbles his way through his monologues; Abbey had a goal in socializing, you might say.
Yitang Zhang is a fascinating case: he failed to get a tenure-track job because his early work had no real successes, he bounced around as a lecturer (and even fell out of academia for a while), and then he produced a monumental advance on the twin primes conjecture and was instantly given tenure at his university (and naturally, he was poached the next year). https://en.wikipedia.org/wiki/Yitang_Zhang
I’m a tenure track faculty member at an Ivy. From my experience, I think this analysis gets almost everything right, except:
1) take Scott’s observation that every hour spent in the classroom is an hour not spent doing research together with the observation that the greatest research contributions tend to come earlier in people’s careers, and it’s no big mystery why universities don’t hire everyone as adjuncts and wait for the superstar researchers to rise to the top — if everyone had to teach a moderate amount, those future superstars wouldn’t have enough time to do the research that would make them superstars and everyone would be less productive as a whole.
2) I don’t think students are really clamoring for famous thought leaders to teach them. With maybe a few exceptions at the very top (eg. Steve Pinker), few undergrads have any idea who the field-famous research stars at their university are and almost none of them know when they are being taught by a research star and when they’re being taught by an adjunct who isn’t doing much research at all. That might leave us without an explanation for why the titles are so confusingly similar, but I’m not sure that needs much special explanation. All titles are pretty confusing to outsiders. How many people know the difference between the Dean of the College and the Dean of Student Affairs? I’m not sure they’re less confusing outside of academia, either. Is the difference between a Software Architect and a Software Enginneer really less opaque?
Because, a brand new Ph.D. can be all things to all people. Someone with experience can only be what they are.
This looks like a multi-armed bandit solution. If the variance in outcomes among inexperienced hires is greater than the variance in outcomes among experienced hires, and you are trying to maximize your likelihood of making a great hire, you're going to make a substantial number of offers to untested but possibly great new hires. If they don't turn out to be great, you can decline tenure later.
I don't think it's a mystery why the job market (in humanities) penalises experience. It's a bit like the dating market, desirable candidates get snapped up quickly, and if you're single for five years straight there's probably something wrong with you.
It's only slightly different in the sciences, where there's an expectation that you need to be a postdoc for a few years. This really just means the clock starts a few years later.
Universities hire based mostly on prestige rather than ability. Whether that's because prestige is an easy-to-measure proxy for ability or whether they genuinely want prestige more than ability, that depends on how cynical you are. But you don't become more prestigious just by sitting around on the shelf. There's a chance you might, in that extra year, do something that makes you extra-prestigious, but it will need to be something pretty special.
> But why aren’t colleges more like un-salary-capped sports teams, trying to outbid one another for their rivals’ superstars?
Because there's not enough benefit to doing so which would offset the huge cost of a bidding war. The fifth best basketball team can become the best basketball team by hiring the best player in the world, but no single hire or handful of hires is going to loft U Chicago above Harvard. Bidding wars cost a lot of money, and they also piss off your existing staff who are annoyed that Professor Big Shot down the hall is suddenly earning four times more than them without doing anything meaningfully better.
This is what I’ve been told: universities like to hire young PhDs for tenure-track because they want to make sure you can be productive for as many years as possible into the future. If you’ve had a PhD for 6 years already, the university fears that’s 6 less years of productive output and once you get tenured, the university is basically stuck with you. That’s what I’ve been told by profs who are / have been on hiring committees.
acoup is generally on point about the structure of the academic profession. Basically, much of teaching is done by specialized teaching faculty who get little money and are looked down upon by "research" faculty.
With that said, we have major disagreements.
First, I don’t think universities can easily save a lot of money. Universities spend huge sums on administration not due to incompetence but because all stakeholders are demanding more IT support, more support for student mental health, more support for finding jobs, etc. There is spending I see as clearly wasteful, but is generally spending that pays for itself – fancy buildings requested by donors, luxury dorms that attract out of state rich kids who pay the bills, conditional government grants, etc.
Second, it isn’t clear to me that spending more on adjuncts is the most compelling priority for any money that can be freed up. Adjuncts are highly educated adults repeatedly choosing to take crappy wages in exchange for a nominally cool job and a chance of a better job. Compare that to, for example, undergraduate students who are often literally children when they commit to spending in the low six-figures on a college education that is now a barrier to entry for a middle-classhood. I'm not sure I'd spend extra money on the adjuncts over just reducing tuition.
Full disclosure – highly remunerated tenure track faculty at R1.
Some additional variables:
1) Grants. An Einstein doesn't just bring in prestige, they bring in millions of dollars of funding, which can then be used to hire more people. As a result, there's pressure to find people who have markers of future grant success. But grant funders are cautious, they want to make sure that their millions of dollars are going to be put to good use (which is exactly the kind of strategy that en masse makes research worse, as moonshots almost never get funded), so another vicious cycle develops where they preferentially fund people who have already had grant success, or certainly look like they will. This then interacts with....
2) Hyperspecialisation and illegibility. Grant decisions are often made by non experts. Job searches often feature someone from management or outside the department. Even within departments it can be hard to assess the quality of someone's work if it's not your exact area. As a result other markers (E.g. outputs in ranked journals) become the lowest common denominator. So if youre a few years post PhD and don't have these markers, it makes you look much less likely to have future success compared with the newly minted PhD who already has a few but hasn't had much opportunity to fail yet either
3) Relatedly, the pressures for more presige, publications, and funds point towards 'collaboration' and interdisciplinarity - it's easier to tell a future grant success story if you can say you've teamed up with Einstein at another university, and universities that work together can feed off each other's prestige, and if you add your name to their paper while they add theirs to yours you can get more outputs. So new hires who have connections with other instutitions, or whose supervisor was Einstein, can be a genuine asset, and have an advantage over someone local who only knows everyone you know.
4) It's also important to remember that the Adjunct market isn't normal. Departments preferentially give jobs to their current grad students so they get the experience. As you finish your PhD, there's a new batch of grad students needing work, so it gets harder and harder to justify giving the job to you even if you are more qualified. This means that as you get further out adjuncts often have to find non-academic work which makes it difficult to research and get those other markers.
Regarding why potential research superstars aren't made to teach, and why teaching track is incompatible with research track:
https://slatestarcodex.com/2017/11/09/ars-longa-vita-brevis/
Regarding why there's some poaching but not a huge amount: the people are involved are all quite intelligent, and can recognize Moloch traps, even if they're not entirely immune.
https://slatestarcodex.com/2014/07/30/meditations-on-moloch/
Overall, the academic job market is like all the worst parts of asymmetric dating markets. I think literally you can just pick the worst dynamics from dating markets, and there'll be an academic equivalent, and vice versa.
On that note, maybe one of the reasons it hasn't gotten fixed is that you just don't talk about it. The impression I get is that it's airing dirty laundry in public, and erodes trust in the Academy at large. I've been in a discussion with some academic friends about Bret Devereaux's blog, and they were distinctly unenthusiastic, and I think one part of it was the dirty laundry aspect.
I don't get why my nice tenure-track job exists. I teach political science in a largely undergraduate program (no PhD students, sort-of-related Masters program) in an institution that excels in STEM. Nobody is expecting truly great research out of *anybody* in my liberal arts college. Anybody in it who does do great research moves on to a more prestigious university (especially for social science) in no time. So, why even have 30% of the faculty in my department (and college) on this much more expensive tenure track. The university's overall reputation is made by scientists and engineers. I think there's a lot of path dependence and a lot of prestige to publishing research even if almost all of it published in my college is... fine.
My wife is tenure track research faculty. Based on her experience, I would say the elephant in the room is grant funding.
Essentially, research faculty are as valuable as the grant money they can bring into the university. Universities get an additional 30-50% of the value of the grant paid directly to them for overhead costs (and there is little accountability for how the university chooses to spend that money). Also, research faculty are expected to fund a portion of their own salary from their grants. The percentage varies, but my wife funds 50% of her salary from her own grants, so as long as she has grants then the university only pays half. If you are research faculty with funds you are an asset. If you are research faculty without funds you are a burden. Anyone the university considers making tenured faculty needs to prove they will not be a long term burden.
The best NIH grant you can get is an R01. If a researcher has that, then your lab is funded securely for the next five years. Once you show you can write an R01 grant and get it, universities are willing to make you tenure track. If you write and get an R01 as part of your post-doc work, then you can apply to tenure track positions immediately and are likely to get one. Then all you have to do for tenure is get a second R01. However, if you finish your post-doc without an R01 then you need to find a non-tenured position and scrape by until you get a big grant.
In short, there are no superstars. There are professors with a track record of bringing money into their university and professors without that track record. From the university point of view, if your people keep getting grants it means they are probably rising to the top of their field. Or maybe a university could poach a superstar, but the actual value added is the money they bring with them. A 'superstar' that has no active grants is no longer one for the purpose if hireability.
Academia is basically one of the "dream careers", in that there are tons of people who dream of doing research and/or teaching at the college level, and so you have a huge field of folks fighting over limited space. It doesn't help that by the time they're looking for tenure track jobs, your professor folk have spent virtually all of their adult life (and a big fraction of their overall life) inside academia with its peculiar ways.
"Maybe colleges really do want “superstars”, not just moderately good researchers. The value of the #1 brightest new PhD is that she has a 5% chance of becoming a future superstar; the value of the #100 brightest new PhD is that she has a 1% chance of becoming a future superstar. Once you’ve been around for five years, colleges can see your track record, satisfy themselves you’re not the next Einstein, and lose interest."
Reminds me of this by DeBoer on NFL hiring:
https://freddiedeboer.substack.com/p/the-nfl-rookie-pay-scale-is-somehow
Others have said this for specific other fields above, but my main reaction to this was just: that graph of decaying chances to get into good jobs looks *right*. It looks like the way the world should be if it's functioning properly. These are all PhDs, which means that the academic world has already kicked the tyres of these people and determined which ones are roadworthy. On the first pass, most of the good ones should get swept up into the good jobs. On the second and third passes, there are fewer good ones who escaped notice first time round, so there are few hires into the good jobs.
There are presumably lots of ways in which academia is messed up, but I don't see how that graph can be one of them.
My experience is that academia is not a closed system. It's competing with private industry. And industry will always have more openings and bigger budgets.
I did my master's at a research university. Not tier 1, but well regarded in a few specific fields. My lab was swarming with tech and defense companies looking for specialists. Tenured professors would consult with or at least exchange interns with these companies. For a fresh PhD, then, the BATNA wasn't settling for a different track or a lesser university. It's getting a six-figure job offer in Huntsville and leaving academia entirely.
If industry and academia have similar demand for fresh PhDs, but industry's demand grows faster with experience, I would expect industry to win. That could generate the weird curve that Bret showed.
A few thoughts on this:
a) a lot of the most groundbreaking research in an academic's career happens when they are relatively young. Since you mention Einstein, it's worth going to take a look at how long he lived vs. his age when his most famous theories were elaborated. So this means that, to paraphrase Pulp Fiction, academic's minds don't age like fine wine like they think, but instead slowly turn to vinegar. Hence older academics who haven't become famous are not very valuable.
b) Tenure committees are probably incapable in most cases of distinguishing between candidates based on objective quality criteria (the few true megastars everyone knows about and get hired quickly, leaving a bunch of grinders who are more than capable of doing job but don't quite have the "it" factor). In that context as you say hiring the young guy who less proven either way provides more upside, but conversely if people have been on job market for years without getting tenure anywhere that also becomes an easy way for the committee to ding them. Sort of like a house that for whatever reason sits too long on the market, you start wondering what's wrong with it even if you can't see any obvious issues.
c) am skeptical it's politically acceptable to start paying professors in a differentiated and market-reflective way. If professors with similar job titles make vastly different amounts of money that's going to create tons of jealousy/infighting, and if they get truly paid free agent type money the marxist types all too common among student body are going to be picketing the president's office to fire them (like they regularly do to try to cap pay for the finance people who manage the university's giant endowment programs). Happy to be enlightened otherwise, but would be very surprised that universities compete on individual-specific pay packages in a meaningful enough way that professors would be tempted to regularly jump ship.
I’m a professor… some additional clarifications. 1) Adjuncts are sometimes Ph.D.s (particularly in the humanities), but in other fields are frequently simply people who know a lot about the field (practitioners). Practitioners don’t do research, therefore they are not candidates for tenure track positions. Adjuncts with Ph.D.s are almost always out of the research game almost entirely. 2) Teaching faculty don’t have much time to do research. A full course load in many schools is 4 courses a semester, at others it’s 3. Grading, teaching, dealing with student issues, etc. all take up a great deal of time. Research is not a side-activity — it takes an incredible amount of time as well. Anyone on the teaching track is likely to not have much time for research. 3) In fields supported by grants / labs / access to technical equipment, universities frequently must fund the up-front costs in the hope that the professor will obtain grants to cover future costs. In Engineering, a startup package for a new tenure track faculty member might be in the hundreds of thousands of dollars, for example.
Overall, however, the analysis that universities have a “prestige” track and a teaching track; that the teaching track has a surfeit of people willing to go into it, and that it is deliberately obfuscatory so that the “advertising” can be about prestige is spot on. To speculate a bit, I think that one of the reasons that teaching is easily undervalued (not just at the university level), is that it is so difficult to measure teaching quality. Clearly some teachers are better than others at getting students to learn the material. However, our instruments for measuring teaching quality are really terrible. By contrast, everyone can count papers, citations, and grant dollars and can assess the quality of journal that the paper is in (but assessing actual impact is very hard as well, which is why that also isn’t a big evaluative component).
Once you are on the teaching track or an adjunct you spend way more time prepping new teaching materials and applying for jobs the next year so you don’t get research time. Changing universities or being given new teaching to do every year has an enormous marginal cost compared to teaching your same courses year on year when you are TT. In some professions the journals are corrupt and they will publish the work of people who already have good prestige signals so when you get a little prestige it compounds very quickly
Working on the research side... I think it's even more dysfunctional.
There's research and academic staff. Research staff typically have to chase their sallary from grants.
Academic staff typically get a steady paycheck and don't need to chase their salary and have more teaching duties.
Many research staff enjoy research and want a quiet life but almost universaly hate the constant grind and uncertainty of chasing grants.
Many research staff at prestigious universities will at some point move into an academic post in a less prestigious university for more money and more stability.
Almost any particular field is a small world where everyone knows everyone and networks, collaborations and connections have their own value.
I was one of those PhD students who graduated with a much better teaching CV than research CV, so I was considering applying for university teaching positions when I graduated. (I ended up at an tech start-up instead.) I think I had very different impressions of the market.
(Context: American R1 institution, graduated a couple of years ago, in engineering.)
My impression was that the academic job market was primarily for research, and if you mostly wanted to teach, you were basically looking in a different job market. Traditionally, some of those people would be job-titled lecturers, and there was some grumbling among teaching-focused faculty about being treated as second-class citizens. But there was some optimism about change, as some institutions came round to treating teaching more seriously. Some departments in some research universities had started hiring for tenure-track teaching faculty positions—"with voting rights", one of my professors explained to me—with it explicit in the job title, say, "Assistant Professor (Teaching)". It wasn't super common, but those departments would've wanted to build very strong undergraduate foundational programs. Or, you could go for teaching-focused colleges, like liberal arts colleges, or engineering colleges like Harvey Mudd or Olin. But this was still functionally a different, and smaller, market to the "real" academic job market.
In the real job market, people seemed to think that hiring committees claimed to care about teaching, but didn't _really_ care. And as a student, every faculty member I knew was research-focused (and possibly loved teaching, but it wasn't their primary role), or if they were teaching-focused you knew it from their job title. So I never really felt that there was any attempt to obscure the distinction between teaching and research roles, more just that teaching roles… weren't that big a thing.
That said, I admit I can't really provide any insight into how the adjunct thing works, other than to say I didn't think it was about teaching.
The question Scott asks is an excellent one (Why Is The Academic Job Market So Weird?). I've often wondered this, and it's weird in so many different ways that answering it touches on almost every aspect of academia and higher education. It's perhaps pointless to write when there are already 300 comments (!), many of which are excellent (esp. noting the existence of postdocs in STEM), but here are a few notes from the perspective of a tenured faculty member in a natural science (Physics) at a large U.S. public university. These comments here are fairly minor; perhaps at some point I'll attempt an actual answer to the question, which I think involves (i) the existence of multiple competing, incompatible, and often cryptic goals and (ii) the mismatch between these goals and where money comes from.
1 Poaching. As many have noted, in the sciences it is essentially hopeless to go from an adjunct to a tenure track position, mainly because the former is (usually) based on teaching, and the hiring the latter is based largely on research. A more puzzling question is why there isn't much more poaching of tenured professors, since these people have *demonstrated* that they can succeed, and postdocs newly on the job market have not. I think the answers are (1) a somewhat irrational desire for young blood and (2) the higher cost of more senior faculty, especially in terms of lab startup funds (in most STEM). Also, the lack of immediate incentive for tenured people to move leads to attempted senior hire processes dragging on forever (> 1 year) and often being unsuccessful.
2 Research income. "Colleges want two things from their professors." Actually three, the third being grant funding, from which the university takes a cut (at least 30%). This also relates to the "superstar" part of the essay -- grant success isn't uniformly distributed, and superstars can bring in a lot of money.
3 "colleges will enter a bidding war to get you" -- there is remarkably little variation in salaries, and often a lot of institutional constraints on this. There are, however, bidding wars in terms of STEM startup costs. In Physics, Chemistry, and Biology, these are often ~$1M.
4 Postdocs. As others have noted, postdocs exist, which is in some sense a trial period pre tenure-track. However, schools generally don't hire their own postdocs, and this makes sense: that postdoc's research area is probably already represented by the lab he/she is in, and hiring the postdoc as faculty would create questions of nepotism, independence, etc.
Devereaux's post is also excellent, though I do think he underestimates the role of the willingness of people to take terrible jobs in encouraging the existence of terrible jobs.
It's honestly very difficult not to read Devereaux's post as mostly being an argumentum ex butthurtum over not having the tenure-track job he admits he wants. I don't think there's any real mystery about why there are many adjuncts and few tenured professors: adjuncts are cheaper and they don't have tenure so you can fire them. And as you suggest, the fact that they're teaching (a necessary evil) instead of doing research (what academics actually want) means they can be pretty crap – and since they don't have tenure, you can fire them if they're *too* crap.
And Devereaux not having a tenure-track job also seems like just a good decision by his admins, since he's rampantly ideological to the point it hurts his academics – for example, he freely admitted that he wrote his terrible "exposé" of alleged Spartan incompetence pretty much because he hates laconophilia and the ideology of those moderns who admire them, but flat out admitting it doesn't make it any better. Academia may be grossly ideologically skewed, but "being wrong on purpose to own the cons" probably still isn't what even very liberal faculty want in a researcher.
I think there are two equilibrium tenure cultures: a strong presumption of tenure, and a strong presumption against tenure. Anything in between leads to too much politicization and infighting. In legal academia (where I work) there is a strong presumption that a tenure-track professor will get tenure. Tenure denials are rare and traumatizing events for law faculties. One consequence of this is that there is much more pressure on the entry-level market hiring decision, because you are basically hiring this person for life unless they are good enough to lateral somewhere better.
These days hiring in law is mainly focused on how exciting/interesting/productive you are as a scholar (and secondarily on diversity and course offering considerations). In prior generations there was less of a focus on scholarship, and the model was just to hire the smartest recent law graduates (many of them supreme court clerks). But now there is basically an expectation that you will have a PhD and two or more full published articles. Adjuncts in law faculties are on an entirely different track - they are either explicitly time-limited fellowships of one or two years (with the idea that you will go on the market for a tenure track job), or they are longer-term teaching position with 0 expectation that you will ever move to the tenure track (e.g. legal writing, clinical professors, etc.).
The emphasis on scholarship for tenure-track hiring (as opposed to teaching) is, I think, a product of institutional culture rather than incentives. Faculties are self-governing and (at least at most law schools) concerned with producing interesting/important ideas. I'm not sure how much the law students themselves care about how good their professors are as scholars (I suspect very little), and the new US News rankings basically don't weight faculty scholarship at all.
Within my immediate circle of current and former colleagues there has been pro-athlete style poaching of researchers by universities and non-university academia. At t=1, t=3, through t=25 years from completion. Within Engineering, Physics, and CS.
My advisor was poached from R2 to R1, months before I joined his newly forming group. 50-ish guy then. My sister’s advisor had the same - her first advisor had been recently poached from a DOE lab to R1 university. Both cases were university corporate donors who committed to funding specific research expansion, then deans etc headhunted - not existing grantors coming along with professors/researchers deciding to go looks for greener pastures. Full disclosure: I quit mine, my little sis quit hers.
Within CS there are 10+ professors or PhD’d researchers I know personally who have been approached by approached by at least one university dean/board member/prestigious alum with an offer of “come join us, we want you, and have funding lined up + new space for you + (perks)”.
But since it’s CS, and decently talented folks with recognized achievements start at $1m-$2m base salary + $xM I’m stock at any one of hundreds of tech companies, universities have to compete to play for that talent.
As an academic drop-out I do want to mention that in my experience the tenure track profs are sometimes Einsteins, and often just people who know how to play the funding/publication game.
My advisor was a "meeting post" type. He published a paper in the 80s that sparked a bunch of futurist speculation in the media. The research itself did not end up being seminal in the field, but it happened to catch the public's imagination at the time.
He leveraged that fame/prestige into setting up a self-reinforcing cycle. He has access to a lot of funding, so he attracts a lot of smart people into his lab. A small subset of those students end up doing interesting work. He takes those students and parades them through conferences and such, which secures him more funding.
So cynically, universities mostly compete for prestige. And prestige has little to do with meaningful contributions to human knowledge, at least in my experience.
In terms of colleges being good at predicting research success, and virtuous/vicious cycles, this is what things look like from my perspective as a math postdoc.
There are 50-100 top universities that hire full research professors. When I talk about tenured positions I will be talking about these research positions, despite the fact that many other positions are also tenured with some level of teaching. You can determine if a professor is a research professor or not by looking at their teaching load. It varies by field, in math the lowest teaching load at top universities is typically 1,2 which means you teach 1 class one semester, and 2 classes the other. A position with 1,2 teaching load is a research position. A 2,2 teaching load is mainly a research position as well. 3,3 is a teaching position with maybe a little research on the side, and anything more like 4,4 is entirely teaching. This is just a fact of life because there are only so many hours in the day and time spent teaching is time spent not doing research. (these numbers are for math. For chemistry/physics and other sciences the teaching loads are lower, say 0,1 or 1,1 for research positions. This seems to be because everyone takes math classes in college, but many students never take a chem or physics class so less teachers are needed, also likely because other sciences bring in big grants for labs and universities like this. Math is a bit weird in being a STEM field which requires next to no funding or grants to do good research.)
Research professors are hired entirely based on their research prospects and teaching is essentially irrelevant. Tenured Research positions tend to be very comfortable, and have many more good applicants than there are positions. You can tell how impacted/competitive a field is by looking at how many years people with recent tenure track offers spent in postdocs (short term research positions that are not tenure track) before they got that offer. In math people typically do ~3 years of postdocs. In theoretical physics I hear it is more like ~6 years. In econ I hear it is often ~0 years so many people are hired right out of grad school. There are people who get tenure track offers in math right out of PhD but these people are exceptional even among people who make tenure track and were successful enough in their PhD research that they were at basically no risk of not getting tenure somewhere if they did a few years of postdoc.
I see 3 reasons for the empirical fact that if someone does not get a tenure track research position after the standard number of years in their field plus one or two years, they are likely to never get one. (So if you have been a math postdoc and not gotten a tenure track research offer after 4-5 years you likely won’t get one.)
First: People have different levels of aptitude in research. Some people are strong on some combination of intelligence/hard-work/choosing good research problems/having a personality well suited to research. People who are sufficiently good at this are usually apparent by the end of PhD or a few years after. People get tenure track offers when it becomes obvious they have this, so if you haven’t gotten an offer after doing a PhD and a few years of postdoc, then you have spent a lot of years doing research without displaying this. Universities want someone who has a chance to do crazy good research. After you have spent 6 to 8 years doing research in PhD and after, the chance of you suddenly getting a bunch of way better results is not that high. Skill caps are very real in math research and the top end is extremely high. I was pretty successful in PhD and have a Postdoc at a top 20 school, but I knew several grad students who made me look like a lazy idiot. Think 5 times as many papers as me and their best papers are orders of magnitude better than mine.
Second: The more promising you appear as a researcher, the more likely you are to find postdocs with no teaching load or very little teaching load. This means you have more time to focus on research than the people who looked a bit less promising, and the gap widens.
Third: If you have done obviously good research in the past then people are more likely to reach out to you about exciting questions in the field because they know you might be able to help solve them, and people are more receptive to working with you on projects you are excited about. This means if you have better research than someone else at the end of your PhD, then you will likely collaborate with better researchers in the following years and have more/better research problems to choose from.
Another fourth thing that is not very relevant in math but is relevant in experimental sciences is lab funding. Someone with more promising research gets a postdoc at a top 5 lab with a bunch of cutting edge equipment, and their friend with less promising research goes to a top 50 lab with decent equipment. There are more opportunities to do exciting research with better lab equipment. I work in math so I don’t know how big an effect this has.
Getting a tenure track research position many years later than usual does occasionally happen but it is quite rare. Examples of this look something like Yitang Zhang, who spent 22 years after finishing his PhD in several positions including lecturer and accountant, then surprised everyone by proving an extremely impressive result related to the twin prime conjecture. He won a bunch of prestigious prizes for this and got a tenured position at UCSB 2 years later. This is an extreme case, but the general idea is colleges want research professors who get results that are so good they attract people to the college. For people early in their career, universities speculate a bit based on the quality of your previous research. If you want to get a tenure track position more than a few years later than usual, you probably need to get a result that big before you get that position. That late in your career colleges will not speculate on you.
Maybe graduate students are the missing link in all of this? Prestige doesn't just come from top-level professors, it also comes from students which find immediate success upon leaving the school. Job guarantees are a big recruitment tactic, after all. So, tenure track positions are given to recent graduates, in order to pad the school's success rate. But, hiring too many of your own students is suspicious, so colleges dance around this restriction by hiring externally, weaving a sort of web of academic interdependence. This probably wouldn't be feasible in a different industry, but with all of the collaborative research and critique going on, I'd wager that tenured professors from a broad range of colleges are fairly well connected.
When asking the question "What process carves off 30% of professors to get good pay and benefits, but passes over the rest?" it's very important to note that the process is not in equilibrium. Tenure track roles used to be a much higher proportion, but that proportion fell in the past few decades and is seemingly continuing to fall.
You have to think of the "superstars" thing a bit like venture capitalists with Unicorns. They're gambling on "potential", but the potential they're hoping for is even more rarely realized than you're thinking: literal Einstein (or at minimum Weinberg, who according to legend demanded to be paid more than the football coach). Everybody else, they're trying to find a really big story really early. So what you get is kind of the same dynamic where people invested in Theranos rather than a comparable startup that wasn't run by a charismatic Stanford dropout. (It's also like Theranos because of the reputation cascades: the top people like this person, so they must have potential, so we bid them up, etc.)
Now, does that mean they're unhappy with superstars they hire who are not literal Einstein? No, because everyone else (grant agencies, philanthropists, heck on a "what is worth covering" level science journalists) is trying to make the same gamble. So as long as a sub-Einstein/Weinberg superstar can keep seeming like they could end up as Einstein/Weinberg later, people will keep throwing money and status at them, and the university won't need to spend much in the way of resources to keep them around.
(Disclaimer: I'm saying this from the perspective of a (sub-)field that in general does not do the adjunct thing. I know zero people in my field who became adjuncts, and a lot more who left to do data science or machine learning or finance.)
I'm a professor at a "teaching college" in Europe. I don't mean that research doesn't happen here, but (1) it's not supposed to be a priority and (2) what research does happen is applied research, i.e., industry projects.
Even so, research sounds so sexy to the school administration. I am a teaching professor, and always have been. A few years ago they put in the requirement for me to...publish? At first, I cheated by publishing pop-sci articles. Now, with only a couple of years to retirement, forget it. By the time they notice my paper deficit, I will be gone.
What I find hard to understand is this: Educating our young people is a hugely important task. Why does the school administration not understand that? Why are they suddenly trying to compete with research university that have different staff, nationally funded laboratories, etc, etc? We cannot compete in that space, and wasting effort to do so distracts from what we are supposed to be doing.
Finally, allow me to offend a lot of academics: There are too many low quality PhD students, and low quality universities awarding them degrees. Almost every university has a bunch of PhD programs, which they want to fill, because PhD students make for cheap labor as TAs and even teachers. Most of those PhD students don't do useful research, but they serve their purpose for a few years and get their doctorate as a reward. Then they join the ranks of adjuncts, driving the pay even lower.
> But why aren’t colleges more like un-salary-capped sports teams, trying to outbid one another for their rivals’ superstars?
There are three big pieces missing in this analysis:
a) The "superstars" have their own ambitions about what kind of places they want to work in. Adjuncts who become "superstars" move to more prestigious places, but they can be offered better positions where they currently are too. See e.g. Yitang Zhang, who solved a big maths problem and then, quoting wikipedia: "He served as lecturer at UNH from 1999 until around January 2014, when UNH appointed him to a full professorship as a result of his breakthrough on prime numbers. Zhang stayed for a semester at The Institute For Advanced Study in Princeton, NJ, in 2014, and he joined the University of California, Santa Barbara in fall 2015."
b) It's not so much about money - working in a "top tier" place with "top tier" people and better research funding (often from external grants) seems to be more motivating than getting more money yourself. The people who want to make money are already pre-selected against, or they make money from side-hustles.
c) Such poaching *does* happen, but is usually more subtle. See e.g. KAUST, a Saudi university which pays enormous amounts of money for researchers to be affiliated with them (and sometimes also work there). It's working moderately well.
Wrote up my take:
https://rgonstuff.substack.com/p/whats-the-deal-with-adjuncts-suffering
The idea is that deficient guidance leads to sticky unrealistic expectations causing adjuncts (and to a lesser extent other "trying to break into academia" players) to engage in what to outsiders seems like self-harmful and irrational behavior
I cannot claim I read all 351 comments carefully, but one thing I have barely seen mentioned, which is the big thing, is that most of these tenure track jobs are held by people who have been in their job for years and likely decades. As they retire, many of these tenure track lines are simply abolished instead of being filled by someone else. The 30 percent mentioned by Devereaux is likely to shrink a lot further, and many subfields are basically dead as tenure-track already. Tenure as it was once known is likely to vanish in our lifetimes. (And this is completely separate from any political attacks on the concept. DeSantis et al. are simply trying to hasten the inevitable.)
Going big picture for a moment, this is clearly the same process that Turchin describes for lawyers.
https://peterturchin.com/cliodynamica/bimodal-lawyers-how-extreme-competition-breeds-extreme-inequality/
Lawyers are traditionally elite jobs, but there's just too many credentialed professionals to admit them all into the elite, and on the other hand too much demand for their work to limit the number of jobs available to them, and on the third hand there's little incentive for the current elite to forego their own professions' elite status. So the elite status remains available, but only for a select few - some actual superstars, but mostly the current elites' progeny and proteges. The rest remain commoners with elite job titles.
I think this is right. It’s all about money but also about who controls the money. University’s make money from three ways: student income, research grants and philanthropy. They are all proportional to prestige which is proportional to research metrics of top researchers more or less so hiring them brings in more student dollars. Once students are committed the teaching just needs to be at a certain level and you’re good so no need to compete on teaching just hire bare basic there and cut it to the bone and use the rest of the money to subsidize research so you can max out support for top researchers. And I agree make the titles confusing so everyone thinks they’re being taught by the top researchers. Prestigious researchers also bring in more research money as they can put in more competitive grant apps because they have the metrics and the money for a big team who are all providing them with data and help writing grant apps. Those people can then make demands about how they want things to be as otherwise they’ll go and take the money with them. Also it’s the prestigious researchers and their friends who all decide who gets the funding as they are the ones who review grants and who funding agencies and journal editors seek advice from. So they structure the system to benefit themselves. This is natural and unsurprising. You might think it’s ok if the system were actually good at picking excellent researchers but really it’s not, it’s good at picking people who can play the game well. Which means making friends with the right people to get good reviews getting invites to the right conference etc. These tend to be very outgoing people who are socially very skilled but often scientifically pretty mediocre. Also they tend to be a lot older as relationships tend to accrue. Same reason the leadership got real old
In the Soviet Union. Like it takes forever to get high up in academia to a position where you can actually influence things as you have to be a senior researchers which means old. Like it’s almost structurally impossible to lead anything in academia unless it’s in a field where you can bring in large amounts of money from a rapidly growing new industry. Like in comparison to the age of ceos of breakthrough companies it is kind of silly how old academic leadership tends to be.
I haven't managed to look at all the comments, so sorry if this repeats someone. I've spent time as a part-time teaching lecturer (in the UK, may be a bit different), and one thing is that you spend your time preparing and teaching – there isn't much time to do the research/publishing stuff you need in order to appeal to a TT job recruiter. Perhaps a fresh PHD is [a] current in all the required new stuff a university needs (both academic and cultural – i.e. progressive, whatever, doctrine), [b] fresh and ready to do research as that's what they have just been doing. [c], possibly, young and malleable to be moulded for the politics of the institution. (No, I'm being frivolous...)
The usual path for research STEM faculty is to get a PhD and then do a postdoc, after which the best researchers have had time to develop and often have stronger records than people with much more experience. Some people are late bloomers and do a second or third postdoc and succeed after that, but being hired as a teaching adjunct usually means failure to get a good additional postdoc.
I was hired as tenure track at MIT after being a postdoc at MIT and doing a PhD that involved working with MIT professors. If a college really wants you, they'll be happy to higher you for tenure track even if you were already there. If you perform above expectations along the tenure track, other colleges will try to higher you away and offer large salaries. If you perform below expectations you might be denied tenure.
I haven’t read all the comments, so apologies if this point has been made.
I think a hypothesis has been skipped. Departments (who do the vetting of tenure-track candidates) have a pretty good idea of who can be a thought leader based upon their graduate level research and pedigree. A candidate who attends a top tier institution and has a thought leader as an advisor has a better chance of becoming one of the prestigious research professors of the future. So, the promising candidates get hired out of grad school and are retained by the institution that hired them. Of course there are surprise outliers, adjuncts who publish important work while maintaining a heavy teaching load, but this is uncommon.
There are also errors in the other direction, hires who get to tenure then rest on their laurels, but this is also uncommon.
I’m not defending the system, but noting that the data can indicate a type of market efficiency.
> But then why do they only hire inexperienced people? Why only people from outside their own institution? Here I’m even more confused, but a few guesses:
Surely some degree of this is the woke purity tests?
A young person is likely better at being for all the required beliefs, like supporting newly trans rapists being legally considered women, have a shorter history on social media and head just has exploded at the rapid swapping of beliefs where they confuse the politics situation of 6 months ago as the current beliefs, maybe they don't remember bird flu at all so when corona happens they are a blank slate for whatever medical claims are acceptable this month(masks dont work, we lied about masks for front line workers, masks outside 50 feet away from anyone are saving lives) or that obama bombed civilians and so can believe that if a black man was in the white house Americas forien policy would be magic.
A far away institution probably has less knowledge of what someone has said in private.
All this is making me *so glad* I left academia.
I'll echo what others have said about this post, and the relevant source, not generalizing well to STEM, confusion as to why postdocs were not mentioned at all.
To Scott's final confusion: Why not hire everyone in some low level capacity?
I can only speak for STEM fields, but my impression was that this is what Assistant profs used to be. There were many openings and it wasn't too competitive, but getting tenure was quite hard. This is essentially placing low-stakes bets on many people, and seeing who becomes a superstar. Now it is flipped in that it is very hard to get a faculty job at anywhere doing high quality research: my department recently went through ~400 applications for a position and made no offer... but tenure (although not guaranteed) is much easier. I've heard that MIT and other places are still "old-fashioned" in this regard.
The confusion is then stated as: why did things change? I don't have a great answer. Perhaps it is related to funding structures. I'm not convinced that Scott's idea would be any better than what we have now, because most people don't end up becoming superstars.
Another subtlety: faculty in some of the sciences can be divided into two categories:
1) Research faculty: these people win grants (from which the Uni takes a cut), advise many students, and teach ~a course per semester. Typically these would be more applied fields like quantum information, biological physics, condensed matter be they theorist or experimentalists
2) "Liberal arts faculty": Profs who can't win grants because they work on fields no one funds (String theory, quantum fundamentals), but the uni takes them on because of the "prestige" of having them. They aren't making the uni any money, and many of their students (if they have any) are supported by external fellowships. Dominantly theorists, and they all teach; most of their courses are hard! (A friend of mine took a course from a string theorist who handed out the final, said "I'm sorry" and left the room)
There are vanishingly few of 2), if any, in the biomedical sciences, a few in Evolutionary bio, and a bit more (but still few) in Physics/Math. Many don't properly separate these and get further confused when trying to understand why the responsibilities and hiring for some faculty can be so varied. The vast majority of real liberal arts faculty follow a model like this where support comes almost entirely from the university.
Does anyone understand the first "plot"? I've only known one STEM grad to get a faculty job directly after their PhD (and they were told it would be better to do a postdoc first!); most postdoc gigs are 2--4 years. Among Social sciences / Humanities I've known a few, -- but my impression is that there are far fewer openings per year in these fields. I'm very confused....perhaps only the ones who know they can get a job apply their first year out, and the applicant pool is just much larger in the 2-4 year range?
Article brings some interesting points, despite it has been deeply discussed in media and with professionals and organizations in the field.
As I am involved, I’d like to put here my perspective here more impartial possible.
So, at first, and been absolutely partial only in that subject, I don’t think everybody who was taking to this job, actually was looking for fame and money. Primarily, I dislike generalizations once they never will represent the whole, soon even if it is the minority, this shouldn’t be desmoralize or not listen because of the main majority.
FACTS:
Actually, particularly in this situation, the individual who was recruited and he/she didn’t look for or apply for the job.
He/she capability of production has to be measure by the amount of tasks, challenges and daily work and homework, which, maybe you doesn’t have this known, but the program is established in a progressive school, where students are launched. While the tests requires a great deal of memorization, (it’s all about “mind games”), it can be told it is not purely educational. It is about stressing in the maximum level students to make them learn focus and learning in the worst scenarios. In that way, they can learn how to unblock and access contain they even know it was there.
The main factor that some scientists disagree is about not to teach and develop the necessary skills before start the training. The program consist in make the student self-learning. Detail: it is require lots of technical skills.
Considering less than six months, I consider the improvement and increase of the individual very impressive.
But, to be honest and impartial, I will let the own scientists give you the results:
https://youtu.be/xoVJKj8lcNQ
The numbers before and after it is really impressive.
Thank you for the space and anyone who wants to discuss or ask something I am at disposal.
🥰🙌💃
I think the major component of "why is it so weird" is that status draw.
There are literally something like 20 people working on PhDs and post-doctorates, hoping for a hire of some sort, for every possible position to fill.
The closest analogous market I can think of is acting in movies, and we know just how messed up Hollywood hiring is.
All these theories assume that departments are acting in their own rational best interest, so I'd like to propose a totally different explanation: the failure of decision by committee.
An adjunct who has worked in the department for 6 months to a couple years is known to the people on the hiring committee; their work, their attitude, their skill set. Some of the people on the hiring committee will like them and want to bring them on, but others will be looking for someone else (more research focused, less research focused, more interested in teaching a particular class, better at teaching certain classes, particular personality differences, etc etc etc). A new applicant is unknown, and everyone can convince themselves that "maybe this person is what I'm hoping for", and so gets more general support (or at least, less severe no votes).
I think I can answer Scott's questions but warning: this comment is long. I did my Ph.D in the natural and physical sciences at a top 10 institution and now have a tenure track job at a small liberal arts college where I teach several classes per semester and do some research. Some background first.
About teaching: Teaching is particularly time and attention consuming. When I'm teaching, especially large undergraduate classes, the top thought in my brain is usually about that. Research is also time and attention consuming, so even for talented people it's hard to do both. If I take on research that's too ambitious I'm likely to fall behind the rest of the field or get scooped entirely because I work more slowly since my attention is split with teaching. Research also moves fast, with new techniques being adopted over a period of a few years. Being out of a research area for several years, or just being slow to keep up because of a heavy teaching load, means you have a lot of learning to do if you want to try to ramp up your research productivity, on top of the time and difficulty of the research itself. Most people who try to be good at both solve this by working on problems that aren't very interesting to anyone else where it doesn't matter if they go slow. Or, they work at prestigious liberal arts colleges with lower teaching loads and more resources for research. Ruthless time management is also important, but not enough on its own.
About grants: Productive researchers don't just add to a college's prestige, they also contribute financially. They are expected to bring in millions of dollars in grants of which 40-50% goes to "overhead", where the college takes the money and uses it for whatever they want, and the grantee doesn't see any of that. Grant funds are also used to pay graduate student "tuition." This is another transfer from grant funds to college since on paper graduate tuition is as much as or more than undergraduate tuition even though for most of their Ph.D the student isnt taking any real classes, just doing research full time. The student never pays a bill, it just comes out of their advisor's grant funds.
About networking: Getting grants is partly about actually having good ideas and results, and partly about convincing reviewers that you have good ideas and good results. For research that is good but not outstanding it's a subjective decision who to fund and who not to. Since reviewers are other researchers in your subfield, there are high returns to networking and selling your work. People preoccupied with teaching don't have time to do this at all, while prestigious researchers spend a huge amount of time on it in the form of giving talks and going to conferences - my former advisor was often away for weeks or months at a time. Top 10 graduate schools are also full of networking opportunities, eg. the college pays for students to take visiting professors in their subfield out to lunch.
Ok, now to answer the questions:
1. In the natural and physical sciences colleges can't hire everyone into a low commitment job because just to get started requires a huge commitment, usually a startup package in the hundreds of thousands or millions. Colleges put this up because they expect to make it back later in graduate tuition and overhead from future grants. They can't do this for everyone because if the new hire flames out, they wasted a bunch of money on stuff that will be hard for others to re use like salaries for graduate students and postdocs, consumables, and specialized equipment or instruments that may not be useful to anyone else.
2. Colleges mostly hire new Ph.Ds or recent postdocs because they are current in their field and haven't fallen behind by spending all their time on teaching. If they are politically savvy they have also been networking with others in their subfield during their graduate and postdoc years which makes them more competitive both for grants and for getting the job in the first place. Colleges do sometimes hire experienced superstars by poaching them from other colleges, but usually you need to offer an established superstar a very sweet deal to get them to move, like their own building. Superstars bring in more grant funds but not that much more, so the decision usually only makes sense if you think trading money for prestige is a good deal, which it only is at top 10 institutions.
3. I'm not certain why departments have a cultural norm of preferring to hire from other institutions. My former advisor told me it's because they dislike the appearance of neopotism. Maybe, but I suspect it's also because cross hiring is good for networking. If your entire department was trained at your institution and mostly knows each other you'll be at a disadvantage when applying for grants or looking for collaborators.
These weird confusing things happen less in fields where research is cheap, like math, or where cushy highly paid industry jobs are readily available, like CS. In math Yitang Zhang went from adjunct to superstar by solving a huge open problem. This worked because he didn't need specialized equipment, worked on it slowly for literally decades, and the problem was so hard he wasn't in much danger of being scooped. In CS it's more common for senior people to transition back and forth between academia and industry.
Not a current academic but family members are.
1.) You really, really can't do good research and have a heavy teaching load at the same time. Your proposed idea of "hire adjuncts and see who turns out to be superstars" doesn't make sense. The adjuncts will not have the time (or, in lab science, the grant funding) to demonstrate that they are superstar researchers.
2.) As many other people have said, poaching absolutely happens.
3.) Maybe this has changed recently and my experience is pretty math-specific, but traditionally my impression is that academics fundamentally trust Academia's ability to recognize a researcher's quality, and trust that department rankings accurately reflect department quality.
The model is, "good people get jobs at top institutions right out of grad school; those who don't are probably not good."
Sure, there's general acknowledgment that there are more qualified people than there are tenure-track positions, so good people get rejected all the time, and there's an element of luck in career advancement. Everyone knows this. Everyone also knows that there are "who you know not what you know" elements and you're more likely to get hired/promoted if you have personal relationships with people influential in the department.
Somehow this doesn't seem to translate to an equally firm belief that there are underrated people who "got unlucky" on the job market but would be as valuable to the department as their luckier peers, or a desire to "scoop up" underrated applicants.
Losing, it seems, brands you as a loser...even though *everyone knows* that many excellent people must lose every year through bad luck alone, by virtue of too many people chasing too few jobs!
While a lot of the incentives here make sense, having spent most of my life being raised by/married to college professors who have been part of or run the hiring process for their departments, this doesn't match the process on the ground as I've heard about it.
I think it mostly fails by imagining 'the school' to be a unified entity with coherent goals and plans, rather than a conglomerate institution with a a bunch of heavily divided and often openly antagonistic factions.
To whit: my impression is mostly that colleges and universities have pretty strong unions for professors, which want to have 100% high-paid tenure-track positions. Who are in a constant state of cold war with the very powerful and goal-oriented administrators, who want 100% cheap adjunct positions.
(Or, if they're trying to be a top-tier prestigious institution, maybe want 10% tenure track all-stars and 90% cheap adjuncts. But if you're looking at national statistics rather than anecdotes from the ivy leagues, those few prestigious schools will get swamped by the thousands of community colleges and 'average' local colleges that make up the bulk of the numbers)
What I've seen is basically a slow shift in the power balance, with administrators ratcheting in few more adjunct positions every time there's a lean year of budget-tightening, and giving handouts to the old tenured professors who are about to leave anyway to molify the unions, while never actually adding more tenure positions back in when things are good again.
As for why there's a preferences for newer professors in tenure-track positions, I think there are three main factors I have seen in action:
1. Being an adjunct for 8 years and applying for a tenured position is sort of like being unemployed or working as a Walmart greeter for 5 years and then trying to get hired as a program lead on a large software project. Everyone knows that being an adjunct sucks, and there's an assumption that if you haven't been able to get a better job than that for many years on end, there's probably a reason why. It's taken as strong evidence against your quality as candidate.
2. Professors always want their department to tech exciting new ideas and use exciting new techniques that the field has developed sine the last time they were is school, but they almost never want to take the time (or have the resources) to learn enough about them to teach them themselves. Hiring the freshest graduate means getting the newest techniques and ideas and feeling a little bit less behind the times and a little more useful to your students.
3. Because the administration keeps shrinking the number of tenured jobs, the remaining tenured faculty form sort of a 'core' that serves as the heart and soul of the department, and ensures its values and character get preserved into the future. This is probably tacitly the #1 thing that existing tenured professors care about when they interview for tenure track positions, they are basically hiring the new pillars that will have to fight with them in the trenches and continue fighting after they're gone to keep the department healthy and independent, and they'd much rather get a new pillar that is young and will be there for 40 years than one that is experienced and will be there for 20 years. (And of course, the less experience they have, the more you can impose your own values and priorities to teh new hires who will have to carry them after you retire)
“Colleges want to pretend to students that the same people are doing both these jobs, because students like the idea of being taught by prestigious thought leaders.”
I get the 2nd half of this in the most basic sense, but outside of the most elite schools, I'm wondering what anyone's thoughts are on the percent of college students (or their parents) who even know that some professors do both those jobs, what the tenure system is, or what a research university is and whether or not they attend one, especially the further you get from the hard sciences. I was a philosophy major in the early oughts and don’t think I had a clue. I'm currently doing a PhD at a public R1 and would put the percentage very low in the classes I've taught in human development/ed psych- both undergrad and master’s level.
The worst part, is that industry professionals aren't in the mix.
Let's face it, for STEM degrees, the student is really earning a ticket to apply for a vocation. Let's give them a step from university to industry.
Who are the most impressive vocational instructors? Those who have spent ten years in the industry, not the bloke who couldn't really cut it in industry. What do graduating students really need? Internships, industry contacts ... the very things that vocational instructors who rotated out of industry can provide.
Most of even our high school teachers should be drafted out of industry for a year of teaching, then return. In my senior year, I saw the least motivated students say 'I'll just take the BA and become a high school teacher.' Oh right, that person is going to be a great motivator. [last sentence was sarcastic tone]
Suprisingly few mentions of the fact that this isn't a steady-state situation. It's adjunct*ification* that he's talking about.
The process is optimised for hiding problems more than anything else. It's easier to quietly shift more people to the 2nd class track than to make the experience of members of either track (especially the 1st class one) worse.
Honestly this would be an interesting situation to throw a marxist lens at. The preference for choosing the tenure-track ἐκλεκτοί as soon as possible in their career separates those who will make the decisions from those who will bear the brunt of them.
I'm reading B. Devereaux and thinking; this is just like the moral mazes (by Zvi), (middle management capture of the 'ropes of power' of the institution.) It's a piece of Moloch. What to do? Maybe teaching colleges and research universities. With only a smaller percentage (of students) going onto research. There would still be research at the teaching college, but the focus would be the students and not some new result... though 'new' stuff happens every where, all the time.
Oh and cut out the middle management, Profs have to run the place.
I'm in computer science, and think maybe what I say below is somewhat true of close disciplines like math or engineering, not sure about other fields though.
> Or it could be a vicious cycle - if hiring committees dismiss you, then grantmakers and journal editors will also dismiss you, and you won’t have the resources you need to do great work.
> Here’s another question that confuses me even more: Why don’t colleges hire everyone in some low-commitment capacity, maybe as adjuncts, wait to see who becomes superstars, then poach them?
The vicious cycle is more like this (and explains why the second strategy won't work): It is almost impossible to do good research without a job that grants you significant time to do research. Yes, Einstein was a patent clerk, but guess what... he's Einstein. The rest of us need much more time to think and devote to research if we want to get anything done.
Adjuncts have high teaching loads and little time for research, so it's rare for them to improve their research record while working that job.
To counter your claim about preferring newly-minted PhDs... actually more experience DOES help you get a tenure-track job. I was a postdoc for 6 years and got a WAY better job than I could have straight out of my PhD. With a few exceptions, most people we've hired in my department since I got there have several years of postdoc experience (or were tenure-track faculty at another institution).
Postdocs are somewhat similar to adjuncts in being lower status and having no job security. (In fact, their job is guaranteed to end after 1-5 years, whereas an adjunct can in principle stay indefinitely.)
But the key difference: since a postdoc's job is almost exclusively research (or say for math postdocs, have a low teaching load, so still lots of time for research) they can improve their research record in a way that's much more difficult for an adjunct teaching 2-3 courses per term. No one is as *single-handedly* productive in research as a postdoc: they have the research skills of a professor coupled with the free time of a PhD student. (Professors with lots of papers every year typically have lots of PhD students and postdocs working for them, but if they were on their own they'd publish less than a typical postdoc due to their extra job duties.)
I've never seen anyone on a hiring committee ding someone just because they are an adjunct. I've just never seen an adjunct with a strong research record apply in our department. Obviously there will be a counter-example someday, and likely there are counter-examples now, and I'm not making a claim about any individual adjunct. But the law of large numbers kicks in at some point to make the *average* postdoc much more research-productive than the *average* adjunct.
> But are superstars really that loyal and inert?
No, they aren't. Lots of poaching happens, as discussed by the other commenters.
But you really have to be a SUPERstar. If you're merely doing well, then that's typically not enough to get hired by a place even at the same rank as your current university, and certainly not enough to move UP. In general, it's much more difficult to get hired as an Associate Professor (the rank just after you get tenure) at a university than it is to be granted tenure (i.e., to be promoted from tenure track Assistant Professor to tenured Associate Professor) at that same university.
Am I right in saying that this article focusses on the US academic market? How does it generalize to e.g. Europe or Asia?
Here's my guess: Once someone has been an adjunct patiently for five years, you figure they'll probably still stay on even if you don't give them tenure.
Or roughly the same thing from another perspective, maybe people who care about money will leave academia within five years unless offered a tenure or tenure-track job. The rest can be taken advantage of.
The article mentions this:
"Now you may ask why anyone would take a job like that with poor pay (for a job that requires a PhD!), no job security and no benefits. And of course the answer is ‘because they have no other choice;’ leaving academia, even temporarily for a non-academic job is generally a career death sentence"
This is my question, and the answer is provided, though I consider it unsatisfactory. Someone with a PhD should be able to be hired into some useful, PhD-related position, unless the degree is in something purely academic, like history. One should, therefore, if one aspires to a career involving a university and cannot get on the tenure track, then one should hope to come to the university as a "professor of practice".
Anyone who can earn a PhD ought, by definition, to be intelligent. I still don't understand how the majority of adjunct professors would maintain that position and fruitless hope for a tenured position under those conditions for more than two years.
I have recently started as an Assistant Professor at an R1 institution, in biology/life sciences. I think the reason places try to avoid hiring internally is twofold: the first as Scott identifies is that there would otherwise be pressure to keep everyone, including more marginal hires, since you know them. But there is a second stronger force - R1 institutions are trying to hire faculty that will become the next superstars. And it is a LOT harder to become a superstar, with your own independent brand, if you stay at the institution where your previous mentors are located. Initially most people in the field at large will assume the ideas came from the mentors and you just executed them. There are exceptions of course - post-docs and trainees who are so good that they already have a lot of their own recognition in the field, and often these are the very few people who are recruited and retained internally.
Academic in mid-tier university in the United Kingdom here. I want to explain concretely how the graph makes sense (which others have done fairly well), but also to provide some context about the financial incentives to the university that may be UK-specific.
For academic teaching & research jobs (which may have a greater or lesser research focus depending on the university and department), one of the main factors considered by the hiring panel will be the applicant's publication record. What we are specifically looking for is high-quality publications in the last few years, rather than 20 spam papers. (High-quality usually means prestigious journal/conference, as it's not like we have time to read the papers ourselves.)
If you have just completed your PhD or postdoc, you will almost certainly have some good publications co-authored with your PhD supervisor or PI. We're prepared to take the punt that you'll be able to produce more, or kick you out after 2-6 years (depends on which university) if you don't.
If you've been working as some kind of adjunct (which for us, usually means fixed-term contract full-time teaching-only staff, although can be part-time in other universities) or as a permanent teaching-only lecturer, you probably won't have been publishing recently. So your number of publications will be the same as someone finishing a PhD or doing a post-doc, your publications per year will be lower, and you've probably already been rejected for teaching & research jobs when you applied for them with the same research record when you were younger.
I think that's sufficient to explain the graph in the post. But why do we care so much about publication record?
Scott is partly right when he says it's because of prestige. There's a specific mechanism by which that operates in the United Kingdom. About every 5 years, the government runs an assessment called the REF (Research Excellence Framework). This is unpopular, but is meant to hold universities to account for the quality of their research. Details vary from cycle to cycle, but each department has to submit around 3 papers per research academic (teaching-only staff don't count) to a panel of external academics in the same subject, who score the papers on a scale from 1-4. The score of the department is published and feeds into highly publicised university league tables, as well as affecting the amount of government funding the universities receive.
So it is about prestige, but also about money. Hiring low-quality researchers lowers your department's average, sending it down in the league tables and losing you government funding.
Of course, there's a more direct mechanism that makes hiring active researchers financially important. If you have a good publication record, you are more likely to be taken seriously when writing grant proposals to the government-funded research councils. Some of this goes directly on hiring new post-doc researchers or buying equipment, but maybe half goes towards the general running costs of the university. As ever, it varies by university, but in the top research universities, most of the income comes from research grants, not student tuition fees. (In Oxford/Cambridge, the university loses money on home undergraduate students.)
Once again you admit to basically being clueless about how an institution functions, only to then speculate brazenly. You don't have to endorse the structures, but if you really want to add something of value to the conversation you should ask a lot more questions before attempting to spread a bunch of guesswork.
Many very smart people have put a lot of thought into centuries-old norms such as tenure and non-incestual hiring. Such practices are fundamental to the mission of higher education, but are easily ridiculed with the help of facile libertarian assumptions emerging from a basic lack of interest in that mission. (Preferring to weigh the efficiencies of markets, status implications, etc, may be more in your wheelhouse, but without investigating how/why the structure was built it amounts to just idle disruption.)
Of course there's a ton to be fixed in academia, but it's like democracy: if your heart's not in it, you're contributing to its failure. Yet you settle for a cartoonish version of colleges writ large, personified as a businessman who "prefers" a game of pretense with his customers, while most colleges have in fact been wrestling tirelessly to retain many core, long-held values during this period of late-capitalism, where popular opinion treats them as if they were just glorified trade schools.
There are other institutions with this same two track system, for example officers and enlisted in the army or partner-track and non-partner-track jobs in a law firm or consulting company. So whatever is causing this structure probably isn't driven explicitly by academia, but rather some property of academia that it shares with both the army and legal partnerships.
"Second, they need them to do good research, raise the college’s reputation, and look prestigious."
The obvious problem being that there are entire fields where "good research" is literally impossible, because those fields draw no distinction between fact, taste, and opinion. Most of the original academic disciplines fall into this category (I'm looking at you Queen of the Sciences.) Nikole Hannah-Jones was insta-tenured, yes in order to give Howard prestige (among the right sort of people) but you're going to need a particular sort of metaphysics and a lot of hope to think she's going to increase the amount of knowledge in the world.
What about the trivially simple explanation:
When a college wants to hire for a research position it's in their interest to consider all possible interested applicants and have no incentive to favor internal canidates.
Ok, so in that system how often would an internal canidate get promoted to a research position? Well, only when the college had an opening for a research position at the same time as an adjunct at that university had put together a publication record good enough to warrant hiring there but not so good as to let them secure a better job and hadn't yet been hired elsewhere...and the fates smiled on their application (lots of randomness in who gets hired).
Given that most academics on the market apply for multiple dozens of jobs the chances of this happening are just going to be really low.
--
Yes it does suggest that there isn't that much of a premium on staying put (academics prefer a TT job at Harvard than a TT job at no name uni even if it means moving) but that seems mostly correct.
Ohh and the difference with sports teams just mostly comes down to the fact that there is a greater variability in what different universities want. Universities often want to complement existing research teams or have views about what kind of research is valuable.
This makes the system much more like a matching market (eg medical residencies...but more so) than bidding on a commodity.
I think this post is misleading for a few reasons. Someone else probably already mentioned the points I'll make, but in case they haven't:
1. The chart in the post is based on data about historians, not all academics. Your post doesn't make that clear.
2. The bar chart depicts the *number* of historians hired N years after completing their PhD. Even if there was no preference for less-experienced applicants at all, the chart could look the same. Clearly, people are more likely to look for a job right after they get their degree than several years afterwards. There will be fewer history professor job applicants with 3 years of experience after getting their PhDs than with 0, because a lot of people with 3 years of experience are already in a TT position and will not apply for jobs. The chart doesn't correct for that.
I'm not denying that there's a preference for less-experienced applicants; I think there likely is, but that chart in particular doesn't really show this.
3. In a lot of STEM fields, it is vanishingly rare to be hired for a TT position straight out of a PhD; you need a postdoc first. Bret Devereaux mentioned this in his post, but your post makes it seem like getting a TT job right out of grad school is the expected career path in all academic fields.
Perhaps it's an instance of efficiency wages? According to efficiency wage theory, overpaying some employees can be profitable because it raises loyalty, morale, and productivity. https://www.investopedia.com/efficiency-wages-5206757
Tenured professor at a mediocre public university in the US here. I think a lot of what you say here about the separation between research and teaching, and why people fresh out of grad school are likely to get jobs, is probably right. But I think you also need to factor into the equation just how *bad* universities (and everyone else) are at predicting who's going to be the next Einstein. Including the case of Einstein himself, who in my limited understanding was not respected in his field at all until after he revolutionized it. So I don't think it's true that there is competition over the top few PhDs, because I don't think universities will have enough of a coherent set of criteria to even converge on the top few; once you get into the class of 'wrote a competent dissertation at a respected PhD program', within that class of people you're pretty much throwing darts at a dartboard. And it's definitely not the case, at least in my field, that there is a shortage of such people: there are far too many promising recent PhDs to fill the small number of tenure-track positions that come up each year. That said, I think 'able to secure large external grants' might be more predictable than 'will do amazing research in the future'. But I have no data to back that up; it's just a hunch.
A few mechanisms I've seen in play that I haven't seen mentioned in the comments I read:
- For impressive profs you don't want large research accumulation, you want evidence of ability to identify and lead important research. You can estimate pretty well from PhD (plus maybe post-doc years) whether people are doing important work per year.
- There's a huge pool of applicants you need to filter. Having been on the job market for many years is strong evidence that nobody else wanted to hire you. A cheap heuristic for filtering is therefore that lots of time on the job market indicates a candidate who is unlikely to succeed in your hiring/isn't valued by other hiring committees etc.
- Perhaps an uncomfortable point, but teaching university students and leading research are very different activities, and experience in one is not necessarily even helpful for doing the other (past a point). "Experience manifestly hurts applicants" seems very misleading, because they are experience in different jobs. Insofar as there is a problem here, it is with unrealistic expectations about a path from teaching-focused posts into research-focused posts (which might be encouraged unfairly by research institutions who want high supply of teaching).
Few thoughts. First of all, this is not just true of colleges. It’s also true of funded startups. I’d think of it like this: there are probably two Nash equilibria when it comes to getting talent at “low risk”. (A Nash equilibrium is a situation where no one can improve their results by deviating to another strategy.) One is let another institution shoulder the initial risk of fielding a farm team and seeing who shines and who washes out, then poach the winners. This probably seems less risky than fielding your own farm team because if true 1% stars are one in a thousand but your local farm team can only be 100-deep, odds that the true superstar will come from your own back bench are low. My guess is that this is priced in when poaching superstars, so you’re probably overpaying substantially for someone else to incubate them.
The second Nash equilibrium is probably “moneyball” where you take the pain up front and invest in becoming better predictors of talent so you raise the odds of somebody in your own incubator being a break out star. Because this requires patience and up-front sacrifice, we’d expect it to be less prevalent than the first Nash equilibrium, especially in a niche known more for playing the game than questioning the game, until the price of free agents is literally outrageous.
If my theory about the second Nash equilibrium is correct, though, it’s what somebody at one of Scott’s Bay Area house parties would call “ripe for disruption”.
If superstars are generally underpaid (to prevent e.g. "upsetting the apple cart" of other employees' expectations of no one making waaaay more than them for doing "the same job"), then they are from a hiring perspective much more lucrative and their allure to institutions totally makes sense. If they were paid what they are worth, there'd be no incentive to hire them over others. The perverse incentives created by underpaying stars hurts others rather than just the stars.