So Many Research Scientists, So Few Openings as Professors
nytimes.com
nytimes.com
I honestly think that the lack of large industrial research labs and the drying up of good academic research positions is very much an issue of what would be R&D funding going to other things that provide better near term ROI: VC, stock market, etc.
R&D is very much an investment like anything else and for whatever reason (patents, existing IP portfolios, etc.), it's just not a place that gets much attention at the moment as the money handlers have found that they have more ROI opportunity elsewhere.
I think the modern substitute for old fashioned R&D is the modern tech startup. The siren call of these ridiculously inflated "valuations" is simply too much for investment managers and it locks up more and more money that could go towards other investments. There's more parallels there as well, R&D and Startup investment is a high-failure game. But the outcome of successful startups is likely to be much larger than the outcome of a successful R&D venture (which still has very long tails of IP capture, product development, marketing, etc.)
I think it's a shame in the sense that, while we end up with dozens of photo sharing sites and social network startups, there's very few actual world changing ones out there. Most of what we get are not "hard" technical areas and their value derives from their faddish popularity - they're kind of the Karashians of technology.
But I guess that's the way of the world.
For the organizations and countries involved, this helps provide for an advanced industrial base that continues to build and drive an ecosystem that supports powerful economic engines.
Another major reason governments fund research is to produce people with PhDs. This is why the apprenticeship structure of academia is so sticky, and why in order to get tenure professors have to graduate students.
Unless they don't mind working for peanuts, in which case there's no problem.
1. Society uses price signals to drive activity to areas "we" want to develop
2. In research, the tournament model prevails so a few winners get most of the gains
However, the question arises:
The tournament model prevails in many other domains including startups, the movie industry and so forth. However, in those areas most people don't feel that the enterprise is severely underfunded. (I am assuming this so if people have contrary data, I will re-assess).
The question stands: in spite of the tournament model, why does society fail to deliver enough reward for basic research when we as a society believe that there should be more of it.
(We could ask the same of teachers, and so on.)
It could also be the case that "society", whatever that is, doesn't really feel that science is underfunded.
Repeat for grad school->postdoc, postdoc->tenure track, and so on.
As someone finishing up grad school, this doesn't seem quite accurate. To me, it seems like a lot of the interesting research is being done in industry. I remember scrolling through jobs online toward the end of my undergraduate and seeing "PhD required" for all of the positions I was interested in. Most of my friends working on their PhD are aiming either for industry or a research scientist position at a government lab (like ORNL). So I'm not sure who exactly constitutes this group that believes "not professor" = failure.
1) Emphasis on funding translational research using government grants, which industry would otherwise be doing on its own. Hence, industry jobs have evaporated while basic science has languished.
2) Institutional salary caps on Grad students/Post docs, so training funds stretch too far resulting in an oversupply of entry level jobs. Better to pay the best people at competitive wages.
3) Growing divide between industry and academia, making job hopping even harder. Everybody is left worse off.
Most research is actually performed by trainees: undergrads, grad students, and some postdocs. This is silly because
1) the "training" is often pretty minimal (my PhD involved less coursework than an undergrad masters) and
2) most of the people doing most of the work have no idea what they are doing.
Nevertheless, this still happens because there are TONS of established funding mechanisms for trainees: REU (and similar per-university programs for undergrads), and individual (NRSA) and institutional training grants (T31) for postdocs and grad students.
In contrast, there are very few ways to fund more experienced individuals who are not running their own lab. This is bad for those people, and also bad for the institutions.
That said, actual training on doing research was pretty damn thin on the group.
Why do you think they don't now?
> Merck’s move follows a major trend in biopharma R&D, as the biggest companies concentrate more and more of their work in the big hubs. And virtually all of the major players have downsized at one time or another.
> Close to three years ago, Merck triggered a major reorganization in its R&D ranks, as the then new R&D chief Roger Perlmutter set in motion a plan that involved 8,500 layoffs, all of which were piled on a restructuring effort that was announced earlier.
> Those layoffs followed a years-long gap in significant new drug approvals and a string of clinical setbacks. Since then, though, Merck landed a landmark approval of Keytruda, now the number two checkpoint inhibitor on the blockbuster cancer market, along with an OK early this year for its hep C combo, Zepatier, which is being sold in a rival-infested field.
The private sector cannot innovate until monetization is in sight and the public sector is increasingly squeezed to publish papers with limited resources.
The overall effect is less innovation and bad science.
That feels like a sweeping condemnation of R&D in the private sector in general. I know very little on the subject, is there a strong base for this position?
I used to work at a large company that in the past performed a large amount of fundamental research into physics, since it affected their products. I took a tour of the labs recently and they are a shadow of their former selves. It was depressing.
I think reduction in private sector STEM comes from several sources. One is the private sector not willing to take the risk of something not working, no long-term thinking, etc.
Another might be the relative difficulty in measuring the impact of your R&D department. In an era where every cost has to be justified and everything has to be measured by some metric, R&D begins to look like an expense rather than a profit generator, at least in the short-medium term.
When big conglomerates ran big horizontal and vertical businesses, they did general R&D work that benefited their broader corporate mission. Bell Labs wasn't there because AT&T was some sort of benevolent charity.
Now companies are mostly brokers between outsourcers and have a narrow focus. We haven't "caught up" with the capability that we already have, so startups have replaced corporate R&D, mostly using off the shelf stuff to make minor incremental changes.
What would be news is any sign that things are changing, with applications to grad school dropping, or maybe the expansion of graduate school degrees that are explicitly not aimed at professorships.
Then, a few years ago, I heard it was the case in the sciences.
Now it is in the technical disciplines, and it is egregious. (4x oversupply). The magnitude and the discipline is the news here (at least for me).
Lots of good info from the IEEE Spectrum (2013): "The STEM Crisis is a Myth"
http://spectrum.ieee.org/static/the-stem-crisis-is-a-myth-an...
When I first started I thought these people sucked, and I was going to be great. What I learned is that the graduate development opportunities are rare, and many of the reason people suck are beyond their control. For example, working on tedious projects or bullshit projects, do to your PI's conflict of interest. Or simply put that your research rests on a foundation of lies.
So, on one hand we have many people on the other hand we don't have many qualified people.
I've also heard the opposite of your first example used as a criticism of candidates too. After completing my undergrad degree in math and then grad degrees in statistics, I was astounded how in industry, either describing yourself more as an engineer or scientist who does not work on the abstract math stuff or describing yourself as someone who is very interested in abstract math will both cause you to get rejected.
The only way to win is to happen to have done a lot of difficult abstract math in the past and remember it all well enough to pass tricky interviews, but then to be overwhelmingly happy and satisfied with a job that will not ever ask you to use it and will instead burn you out on dumb shit stuff like fitting a regression to KPI data and using t-stats to directly do (fallacious) model comparison.
The biggest career risk I've seen from having done very computer science-heavy statistics is underemployment. Your math chops will be pure credential, sometimes used by your manager to try to win arguments from authority about e.g. that dumb shit KPI regression. But you will 100% never be given open-ended modeling work that could actually have a positive impact on your business's bottom line.
Essentially, you are hired to be some more senior person's sycophant political darling. You function internally much the same way that a big consulting company functions externally -- people already know what they want to hear, they just want you to tell them what they want to hear, slap together some plausible-sounding rationalizations for it, dress it up with buzzwords about big data and "insights", and be a show pony for talking about it.
They emphatically do not want you for doing anything that would be called "real" work.
There are occasional exceptions, especially in certain teams at certain established tech companies. But then, landing a position inside one of those teams is nothing but the same lottery as winning a professorship all over again.
I'm not talking about academic overqualification. I'm talking about someone hiring you, talking at length about how you are being hired specifically to do X because it matters to the company's bottom line, and then after you're hired they switch it and say actually you're going to do Y but you're going to be a political mouthpiece for X.
One common set of values is
Y = statistically invalid model fitting that actively causes the business to lose money but which is easier to reduce to pliable metrics for political jockeying
X = (deep) machine learning and/or Bayesian stats
It's not at all about academic overqualification. The actual business need, for reals, can benefit from the pragmatic and cost-effective use of the tools, and the person is actually skilled in using to do exactly that.
Yet, they are prevented politically.
As for starting a company, I think I would guess that's one of the least plausible ways of doing important or useful work. You'll only get funding if it is a trite variation on consumer bullshit -- even though consumers themselves don't want that and would rather that your labor is allocated to solving more fundamentally important social problems that there just isn't money for solving. Plus, you'll be so burnt out over all the auxiliary stuff like HR, marketing, sales, that you won't actually do any of the underlying quant work that was the whole reason for starting the company in the first place.
It kills me how people here seem to think that "start your own company" is some kind of "put up or shut up" gauntlet to throw down to challenge someone who is lamenting the shitty state that things are in.
"Start your own company" is not some venerated challenge-call for those brave few who want to change the world. Starting your own company is just a different format of the same bullshit phenomenon.
Fixing what's broken inside of companies and organizations that already have huge leverage and capital to positively impact fundamentally important problems -- that is perhaps worthwhile, if you can manage to deal with the political fighting without getting too burned out.
What I'm trying to say is that you're told you'll be hired for X, and, crucially, that it's easily verifiable that X actually would help solve the business problem better than Y. Failing to do X actively hurts the business.
Yet you're still forced to do Y. It's not because in the real world you only needed simple, trustworthy Y to get the job done. No, you're failing to get the job done, need X to get it done, are told you're the one to bring X to the table, and then you're made to do Y for destructive political reasons.
The point I'm trying to make is that there's no defensible "real world" pragmatism to support the focus on Y nor the bait-and-switch to hire someone who knows X. Whatever the reasons for that, they are not about improving the firm nor making money for the firm. They are about optimizing a bonus or promotion or whatever for a single individual or some small faction, even at the expense of the organization's overall progress.
Sure, it's closed source, expensive, and can be clunky, but it is not a horrible tool for many jobs that involve a mix of signal processing and other data analysis.
Scipy is often a viable altenrative, but if the company already has a bunch of Matlab code, rewriting things in Python probably wouldn't be worth it...
And slow. Don't forget slow. Closed-source, expensive, clunky, difficult to read, and slow.
(The first time I took machine-learning in grad-school, our assignments were in Matlab, and testing them required me to stay up until 04:00 in the morning at least once a month. A validation run just would not take anything less than five to six hours. The second time I took the class, I took it in the computer-science faculty, and the professor gave the assignments in scipy/numpy. Validating and debugging that was easy.)
(Just this year, I actually went and rewrote a probabilistic program in Haskell from a probprog package based on Python. Interpreted languages are fucking slow for numerical jobs.)
Seriously, Matlab = MATrix LABoratory. It's a DSL for matrix math, and it's really pleasant to use for that.
Also, my Matlab code is a joy to read. It's not Matlab, it's that most people who write Matlab code aren't professional programmers / don't care about making it pretty.
> It's not Matlab, it's that most people who write Matlab code aren't professional programmers
This, 1000x, this. There's also the time-honored tradition of a quick one-off script "to see if it works" that somehow becomes permanent.
If a place finds themselves having a lot of MATLAB code, it means that earlier they didn't refactor and retool when the choice was cost effective to do so. That's strong evidence that for whatever they are working on right now they also will not refactor it to at least try to avoid the future costs of current poor designs. That's a huge red flag.
I guess if they paid an insanely high salary or gave you some other type of assurance that you personally valued, you could trade it off against the red flag evidence. But generally these places also tend to just hire maintenance engineers, since they know they can't offer interesting work. It's best just to avoid and work for places that bend over backwards to refactor and evolve their tooling over time specifically to prevent this problem.
Serious question: do you feel that way about other older languages too? Would you balk at a C++ shop (which isn't doing high performance or low level stuff?)
I have some additional skepticism about (a) shops that very quickly adopted Go and then display a cult-like dogma about how super perfect and awesome at all things networking that Go is and everything else isn't; and (b) shops that use Scala or Clojure purely as "better Java" and often have thin, entirely unfunctional layers wrapping bad legacy Java code -- in such places, the Scala and Clojure often exists specifically because they had a hard time recruiting people to be legacy Java maintainers, and having them do it through one layer of indirection, such as Scala, let them tap into a more vibrant labor market with a lot of people who just naively believed that e.g. usage of Scala == adherence to good quality standards.
The age of the language has nothing to do with it, and I actually quite enjoy C programming. I'm not a big fan of C++ code bases in which some of the more esoteric language features are greatly abused, but generally feel like C++ is a great tool and would in fact enjoy the chance to dig more into in a future job.
I also quite like code maintenance and legacy code extension, because when the company takes it seriously and values that kind of work, it involves a lot of cool abstraction, working with interfaces, design tests a la the Feathers book on testing in legacy code. I think that stuff is quite fun, and a healthy culture of refactoring makes it interesting to work on legacy systems.
While it's not an iron-clad rule, certain kinds of legacy choices, however, are not really compatible with a spirit of quality, legitimately valuing maintenance or refactoring, etc. MATLAB and Excel VBA are by far the strongest indicators that it's not a good situation.
The items above are just language choices, too. There are also many other things to worry about. For example, when a place uses MATLAB, it often means some of the work they do is scientific / quantitative prototyping. Do they enforce good code standards even at the prototype stage so as to minimize the distance between research prototypes and production? Failing to do this is a seriously gigantic red flag. It can mean that the management chain values the domain science more than the quality of the implementation, and often this means that domain scientists are allowed some of the following
- to manage their own working environments (leading to lots of awful "but it works on my machine" errors)
- to write every thing as giant, messy, linear scripts that start off with 200 lines of boilerplate data loading and model setup code that should be factored into a library but is instead copy/pasted and finish off with 200 more lines of custom plotting code that also should be factored into its own internal library for standardizing reports and charts
- to turn things into short-term fire drills for production programmers solely by virtue of them being "someone who uses MATLAB, not <some real language> that we use in production", and get management support for this.
- never bother to learn object oriented or functional programming principles -- google a design pattern and then just paste some ill-conceived bastardization of it wherever they feel, and then argue with production engineers that you can't refactor it "because it's a design pattern"
- ... I could go on.
When I see places that heavily use MATLAB or R for research systems, I break out in a cold sweat, since those languages are entirely unsuitable for professional software design. They are good for ad hoc linear algebra, optimization, model fitting, plotting, and statistics. But the software design underlying how those things are carried out in MATLAB and R does not translate at all to a system where the scientific code is maybe 1% and the business reporting code, customer-facing services code, etc., are the 99%. And putting the scientific code at 1% is generous even for a company that is solely about scientific computing or quantitative services.
Anyway, there is just a large difference between organizations that work from a systems engineering perspective first, and build that way, and mercilessly require non-programmer PhDs to get up to speed on actual, principled software development even for doing ad hoc domain specific work.
That is the crux of the issue. Many Matlab and R users don't see themselves as producing software per se. They are producing "recommendations" or models or papers; the code is just a means to that end, and so who cares if it is terrible? No one will do this exact thing again anyway... Obviously, that is pretty myopic. If you have good "infrastructure" code, it makes writing the one-off parts easier, faster, and less buggy--which it makes it easier to turn the one-off code into infrastructure.
Out of curiosity, how could I convince you that joining our group (which does have a ton of matlab code) wouldn't be awful? Or what would convince you to hire someone whose last job was predominately matlab?
Good luck and I hope the work is published.
So at least in my area there is no research career. It's a dead end.
1.) I am sorry that you are going through this.
2.) Your experience perfectly matches that of a good friend of mine. She has been doing excellent work, she has had some great pubs and yet she considers getting tenure to be about as likely as winning a lottery.
I wish I had something tangible to offer, but the best I can offer is my good wishes. Seriously, best of luck!
For biomed stuff, it's frustrating that the NIH seems like it could address this problem (they control a huge proportion of the available funding), but have no interest in actually doing so.
If I were king for a day, I'd seriously consider shrinking the training grant pool and redirecting that money to developing viable career paths for "staff scientists" or other experienced-but-non-PI employees.
Giving money to staff scientists would work against both of these benefits. It would cost more and could reduce the number of trained scientists entering industry. (Although that isn't clear. Creating more NTT positions might just give hope to more of those eternal postdocs.)
I agree, but it certainly does not feel like the NIH has thought about this carefully either. My impression is that the NIH does what does because this is how it has always been done (and that worked well enough for the PIs they consult). I would love to see data suggesting that the current arrangement is optimal--or even close to it. The only thing that comes to mind is a recent paper arguing that peer review scores for grants were mildly predictive of the number of number of resulting publications.
As for staff scientists, one good staff scientist could certainly be more productive than a gaggle of newbie masters' students on their own. Right now, everyone is incentivized to do just enough to get the next paper. Permanent employees could build up infrastructure (protocols, code, etc.) that is actually reusable and ultimately more efficient.
Look at UCSF, they don't even have an undergrad program.
Were it not for the historical accident that we do things this way, I strongly doubt anyone would suggest we adopt it for whatever tiny educational benefit it may impart to undergraduates -- most of whom won't even go into the academic fields they are studying.
This makes me think we (as a country, a people), having trained all these scientists, have a market opportunity for inexpensive research, and we aren't taking advantage of it. Sure, the best and the brightest might fight their way to the top, but we can get perfectly good science out of the middle of the bell curve too.
There are decreasing marginal returns from throwing more people at a problem. Especially if they are publishing, because now you need more people to curate their publications. And if you throw enough researchers at something, they are going to find a bunch of spurious results simply because of the luck of large numbers.
I don't know if we are at that point. But adding a bunch of average people to a project can easily slow it down.
If you want to help the existing researchers, mow their lawns and buy their groceries and take care of their house maintenance problems. Let them concentrate.
We're not out of work to be done
Throw enough tenure against the wall and eventually something will stick
Right,
What is the difference between Civil/Environmental and Environmental ?
It seems the best bet is to study something like mining since the R_0 is so low there.
That's roughly how it was divided at my college, anyway.
Also Business schools need professors.
Where is the money going?
Buildings and administration. The "undergraduate experience" that involves 24-hour gyms and subsidized restaurant-quality cafeteria food.
The truth is that professorial salaries are low and tuitions are high (for full-price payers) for the same fundamental reason: universities set the levels where they are, because they can.
http://www.huffingtonpost.com/2014/02/06/higher-ed-administr...
The interesting thing about this is that nobody complains about it.
Research is expensive and funding is hard to come by. Therefore, if you want to have a job next year, you end up doing 'safe' experiments that only incrementally advance your field.
In order to get funding for more radical ideas, you need to have a history of good results (and past funding). Which usually means you have to be a professor already.
(Also, many people who aren't professors, such as postdocs, cannot directly get funding from many funding agencies. They usually require you to be a professor already...)
You just denied tenure to Albert Einstein.
(Luckily, in real life, he was appointed a lecturer in 1908, three years after finishing his degree, and became an assistant professor in 1909 and a full professor in 1911.)
I'm a few years out from my PhD. The classmates who stayed in academia are neither much smarter nor harder working than those who have left. It does select for certain personality traits, like risk tolerance and willingness to self-promote, but I'm not sure that's really what we want.
Usually they are so good at their subject that they tend to skip steps in their explanations and forget how difficult a certain concept is and expect everyone to breeze through it.
Unfortunately, universities optimize for research rather than teaching.
When you optimize for, say, the number of publications, you encourage aspiring professors to vomit out more minimal publishable units and self-promote more loudly rather than thinking interesting thoughts and writing them down after some rumination.
The property you actually care about, the interestingness and quality of the research, is a latent property. You cannot optimize for it by optimizing any of its indicators; they are only noisily related, and intervening on the indicators will only break the connection between the indicator and the latent interestingness.
Getting money today (GMT) is all that matters, so until the govt finds a way to take everyone's not hard-earned money and redirect it toward more noble pursuits, I'm confident the trend will worsen.