[msteffen]

My understanding is that scientific research has a dual problem, where the number of students needed to carry out existing professors' research is much larger than the number of junior faculty positions generally available. The result being that most trained PhDs must leave (US) academia because there are no jobs for them. In fact, I've heard scientists complain that universities owe it to students to provide more help finding a job in industry after they graduate.

Given all that, where are professors supposed to find and hire students who don't want to stay in academia themselves? I think a lot of these students wind up being aspiring immigrants, and I'm not surprised that a lot of them would also have a hard time finding a place for themselves after graduating and that many of them would leave. Also, the abstract seems to argue that that US still benefits greatly from this arrangement: "though the US share of global patent citations to graduates' science drops from 70% to 50% after migrating, it remains five times larger than the destination country share."

[MostlyStable]

If the culture normalized such that a much larger proportion of research was conducted by permanent, non-faculty, research employees, this would both reduce the need for so many students and increase the jobs available for students, and create a new employment niche with a different balance of teaching/administration/research. It would basically be turning "post doc" into an actual career rather than a stop over.

This would be better for everyone involved, at the admitted cost of being quite a bit more expensive. My guess is that the market would naturally converge on this equilibrium if the information of job placement rates on a per-program (or even per lab/advisor) were more readily available.

[Fomite]

This isn't really a culture problem, IMO, as much as a funding one.

My group currently employs two people of the description you have, and it does reduce the need for students (and honestly, increase productivity).

It's also by far the most stressful part of my job. Funding them involves writing multiple grants per year (because the expectation of any particular grant is low, even with a decent hit rate) and I am constantly worried that I won't be able to keep them employed.

If one of them leaves this year, I'm not likely to replace them, simply because in the current funding environment, I can't look someone in the eye and promise them a long term position. There are so many more ways to fund a student, and they're inherently time limited, so even if things collapse, there's ways to white knuckle through it in a way there aren't for staff scientists.

[mnky9800n]

Also, consider what the postdoc is.

A person arrives on a 18 month funded postdoc (believe me, plenty exist). They have just completed a PhD which means they probably have a couple papers published and maybe another one or two in the pipeline. So as they spin up their time with you, they are also finishing these papers from their previous job. By six months in they are done with that and fully onboarded to the project. So they spend six months working. But now, they only have six months left of contract. You don't have money to keep them or perhaps your country will require you to offer a permanent contract if it is being renewed so you cannot offer them to extend their position with you. So they spend the final six months of their postdoc looking for a job. So, for 18 months of salary, you get six to eight months of work. It's unreasonable. Things need to change.

Or lets say you have a mission critical project that must be done by a postdoc. You offer them a 3 year contract that is grant funded. It is three years because most grant agencies work on three year cycles. The project requires a year commitment to building an apparatus (maybe its a lab experiment, maybe it's training some foundation model, whatever). After that year, the apparatus can be used for science. Your postdoc comes to you in year 2 month 3 and says, well I have been offered a faculty position at university X so I am leaving in the fall. So you get 18 months of work out of them and now cannot hire anyone else because you only have 18 months of funding left, but your country requires you to offer a minimum of 24 months contract. Things need to change.

It's important to note that academics often keep projects from their former positions going at their new ones. But as soon as someone leaves to industry, this falls apart. Because industrial positions expect the person to work on the project they specify, they rarely hire someone to work as an academic, pursuing their own research directions.

I think the solution here is as others have suggested, spend more money on hiring people for longer term and with higher salaries. But we shall see if anyone listens to that advice.

[godelski]

Notably even the role of the professor has drastically changed in the last few decades. The "publish or perish" paradigm has really taken over and changed the type of research being done. Higgs famously said he wouldn't make it as a non-tenured faculty in today's academic culture.

Not to mention that the type of research being done has drastically changed too. There's many more projects that require wide collaboration. You're not going to do something like CERN, DESI, LIGO, or many other scientific mega projects from a single lab, or even single field of study.

The academic deal has changed. It used to be that by becoming a professor you were granted facilities and time to carry out your research. In return you had to help educate and foster the next generation. It is mutually beneficial. There were definitely abusers of the system, but it is generally not too difficult to tell who in your own department is trying to take advantage of the system, but incredibly difficult to identify these people when looking from the perspective of a university administration. There's been more centralization in the university administration and I'm afraid Goodhart's Law is in full force now.

What I'd like to see is more a return to the Laissez-faire approach. It shouldn't be completely relaxed, but to summarize Mervin Kelly (who ran Bell Labs): "You don't manage a bunch of geniuses, they already know what needs to be worked on. That's what makes them experts in the first place." At the end of the day we can't run academia like a business and it really shouldn't be. The profits generated from academia are less direct and more distributed through society. Evaluating universities by focusing on their expenditures and direct profits alone is incredibly naive. We're better able to make less naive evaluations today, but we still typically don't (it is still fairly complex)

[j7ake]

Your suggestion would have fewer fresh eyes to look at the problem. If the scientific enterprise were just about churning out widgets, then yes it’s better to have permanent staff.

But having a strong training pipeline for the globe is a huge plus for US prestige, and the top people are still offered jobs as faculty or industry within the country, so it still a net gain for USA. But it’s brutally competitive for the individual scientists

[turtletontine]

What you’re describing sounds a lot like the Department of Energy national labs. They have (or had) many permanent-track research roles without teaching obligations, where scientists can have long stable research careers.

The problem, as always, is funding. In the US, the federal govt is essentially the only “customer” of basic research. There’s some private funding, often from kooky millionaires who want someone to invent a time machine, but it’s the exception that proves the rule. Universities sometimes have pure research roles, but they’re generally dependent on the employee paying themselves with a constant stream of grants. It’s a stressful and precarious position.

[cafebeen]

To a large extent, I think this could be solved by labs having more long-term permanent research staff (technicians, data analysts, scientists) and reducing the number of PhD students. Many students would gladly stay on in that position instead of leaving, so it increases job opportunities. It would also improve the quality of the science because the permanent staff would have more historical knowledge, in contrast to the current situation where students constantly rotate in and out with somewhat messy hand-offs. The students could also then focus more on scholarly work, planning and overseeing research execution with the team. The problem is that the incentives are aligned to allocate students to doing all lab tasks, not long term staff. I think we could change this through changes to the requirements and structure of science funding mechanisms however, since ultimately that's the source of the incentives.

[materials4028]

> much larger than the number of junior faculty positions generally available

Expanding on this a bit, insight credited to bonoboTP: in a steady state the number of junior faculty positions will only open up at the same rate as current faculty retires. But each faculty member is expected to train dozens of students that are all in principle qualified for such jobs. Therefore, the vast majority, let's say 95%, of PhD graduates have to take industry jobs, there is no way around it. But this does not seem to be the goal of the 95%, hence the incredibly tight job market. Returning to their home country for a faculty job acts as another release valve, but sooner or later those will be filled as well, except in countries in the rapidly expanding phase in terms of university education.

The tenure system is incredibly broken as a result. Ideally, I think there needs to be more non-faculty careers available for PhD graduates either outside or inside academia. After all, there is clearly some value in the work a PhD student does, otherwise they would not be paid. Perhaps we can have public or semi-public research institutions that hire these scientists for actual development. Most likely this will require an upstream incentive change so that grants are awarded to these newly minted organizations.

Universities charge a large overhead in part to cover the "tuition" for the PhD students, which is really a meaningless number since it's taken out of the same check they give you the remainder of. If we just strip out this part and give most of it to the scientist, economically it should be a viable salary.

[analog31]

When I was a physics grad student ~35 years go, this was called "the birth control problem. I had every intention of going into industry. I described it to my dad who got his PhD in the 1950s and he said it was the same back then. But there's a perennial "this time it will be different."

[Fomite]

"Ideally, I think there needs to be more non-faculty careers available for PhD graduates either outside or inside academia."

For awhile, I loved that my field had lots of opportunities outside academia for PhD students, and that they were held in pretty equal regard, prestige wise, with academic positions.

Then the current administration gutted the entire field.

[epolanski]

Academia is a pyramid, like most organizations, eventually most PhDs cannot get a full time position.

The fact that many PhDs leave is..normal..if you get few high impact publications you can find full time positions outside US, even as an associate professor and not just a researcher.

And the reason why many go to universities around the world for PhDs is not because they want to stay in that place necessarily but because you're more likely to fund your PhD research and get a high impact publication.

[Onavo]

There's that and the fact that a lot of people who attain graduate degrees are immigrants who do so for the sake of immigration.

The whole system essentially self selects for cheap labor and exploitation.

If the feds put a high salary requirement on it like the E or O series visas, perhaps the system might change.

The scientific minds of India, China, and Russia don't come to the US and slave away in the lab purely out of passion for advancing science, they do so because it's a path towards the green card. The PIs and laboratory heads all know damn well how the system works, they are no better than those bosses of H1B sweatshops, except perhaps they do their exploitation from ivy filled ivory towers rather than in Patagonia vests.

[aleph_minus_one]

> The PIs and laboratory heads all know damn well how the system works, they are no better than those bosses of H1B sweatshops, except perhaps they do their exploitation from ivy filled ivory towers rather than in Patagonia vests.

In my observation there do exist quite some people among the PIs and laboratory heads who are quite highly idealistic for research, but have no other option than playing this rigged game of academia.

[khannn]

I was accepted into a PhD CS program despite applying for a masters. The advisor had something on his door about the limited number of slots open for people who graduate from grad school. Tried to discourage me from the program.

Quit after two semesters.

[godelski]

  > where the number of students needed to carry out existing professors' research is much larger than the number of junior faculty positions generally available.

This is definitely true, there are more physics PhDs graduating from the top 2 schools than there are total faculty positions listed each year.

BUT you are missing that there is still demand for positions out in industry as well as government labs. But there's also a decline in that right now as we're going through a time of encouraging more engineering and less research.

In reality there's a pipeline of research. If you haven't been introduced to it, I like to point to NASA's TRL (Technology Readiness Level) chart[0]. The pipeline is from very basic research to proven systems. Traditionally academia and government labs do the majority of work in the low TRL while industry research handles mid level (stuff that isn't quite ready for production). The reason for this is due to the higher rate of failure of low level research and so shifts risks away from industry. Not to mention that industry has different incentives and is going to be more narrowly focused. Academia and gov labs can research more long term projects that will have large revenue growths but may take decades to get those returns. I mean how much do we get from the invention of calculus? Or the creation of WWW? We'd also get far less growth and profits were these not more distributed.

So while yes, getting a professorship is a challenge and highly competitive, it is far from the only path for these graduates. We can also do a lot to increase (or decrease) their options by increasing (or decreasing) funding for science. There's a lot of science that happens outside academic labs and they still depend on PhD graduates to be able to do most of that work. If you want these people to have jobs, fund more low level research[1]

  > I've heard scientists complain that universities owe it to students to provide more help finding a job in industry after they graduate.

A big reason for this is that networking is still a big issue. I can tell you as someone who does not have a good relationship with my former advisor that this has made job hunting a much harder experience compared to other peers. While my credentials are better than some of those people they come in through a side door (often skipping things like LeetCode challenges) and instead I have to go through the standard applicant pool. I don't think they don't deserve those jobs (most of them do), but just pointing out that networking is still a critical part of hiring. I mean even one simple part is that when applying you might not even know what a group is doing and if that's what you want to do. Solicitations are often vague. Even if there were no advantage to the hiring process networking still provides a huge advantage to the filtering process.

I mean even putting the personal experience to the side, don't we want to make the most use of the resources we have? Don't we want to get graduates connected to labs/work places where they will be most effective? This is still a surprisingly complex problem to resolve and even limiting the hiring problem to PhDs (where there's far less noise than general hiring) it is still a complicated problem.

[0] https://www.nasa.gov/directorates/somd/space-communications-...

[1] But I'd also say that we might be encouraging too many people to do PhDs. Doing a PhD "for a job" is a bit odd. A masters is better intended for that. But a PhD is more directed towards doing research work. That said, in the worst case a PhD says "this person can work on ill-defined tasks and has the diligence to see them through." Regardless of the industry, that is a pretty useful skill.

[aleph_minus_one]

> That said, in the worst case a PhD says "this person can work on ill-defined tasks and has the diligence to see them through." Regardless of the industry, that is a pretty useful skill.

Very few companies and industries want employees who

- are very conscientious ("has the diligence to see [the tasks] through"), and

- are much more effective working on their own, i.e. are no "team players" because they don't really need a team ("this person can work on ill-defined tasks").