How would we know when science trainees bail on the career?

We all know about the oversupply problem in academic science wherein we are minting new PhDs faster than we create / open faculty jobs to house them.

Opinions vary on what is the proper ratio. All the way from "every PHD that wants a hard-money traditional Professorial job should get one" to "who cares if it is 1% or even 0.001%, we're getting the best through extreme competition!"

(I fall in between there. Somewhere.)

How would we detect it, however, if we've made things so dismal that too many trainees are exiting before even competing for a job?

One way might be if a job opportunity received very few qualified applicants.

Another way might be if a flurry of postdoctoral solicitations in your subfield appeared. I think that harder flogging of the advert space suggests a decline in filling slots via the usual non-public recruiting mechanisms.

I am hearing / seeing both of these things.

68 responses so far

  • Established PI says:

    I'm afraid that what I am also seeing is a decline in the quality of graduate applicants. Perhaps I am just looking at the past through rose-colored glasses, but our graduate applicant pool doesn't seem as deep as it used to compared to the quality of medical students (and I have no idea how they all compare to students going into entirely different careers such as comp. sci., engineering, finance...). My overarching concern is that the well-publicized dismal prospects of Ph.D.s deters very talented students who never even enter the pipeline.

  • dr24hours says:

    This suffers from the standard assumption that a faculty job is the only thing worth considering when considering the pipeline. Maybe large numbers of your PhDs are going outside academics because it's obviously, tangibly, objectively better for them there, at the moment.

  • drugmonkey says:

    How is that in any way in conflict with the points I am raising?

  • Ola says:

    Regarding your first hypothesis (trainees exiting before looking for a job), I've certainly heard examples of post-doc's being told "you're not going to be competitive" based on publication records and the like. Such trainees usually get turned town when they ask to be promoted to RAP, or even ask for permission to submit a K award. They then drift around for a year or two doing "filler" post-doc's and occasionally they're lucky enough to find a sugar-momma to promote them to RAP and start submitting grants, but by and large they just say "screw-it" and join the local adjunct pool which is legions thick. Biology tenure track positions at the numerous SLACs within a 50 mile radius are routinely getting 150+ applicants, with the bulk of the pool from the local failed post-doc' population.

  • Busy says:

    My overarching concern is that the well-publicized dismal prospects of Ph.D.s deters very talented students who never even enter the pipeline.

    We see this all the time. A bright undergrad can easily land a $100K starting job in industry, or they can earn $20K a year for the next five years while they pursue a PhD, only to have about 1-in-4 chances of landing a permanent academic job.

  • Morgan Price says:

    In the big research group that I belong to, post-docs mostly bail to industry *before* applying for tenure-track jobs, even though when they arrive, they almost all say that they're hoping to become professors. I think the competition for funding is a huge deterrent. The drop-outs include some very bright and successful scientists.

  • dr24hours says:

    @DM: It isn't. Except that you suggest that the situation is "dismal" if qualified PhDs aren't applying for your postdoc jobs. Maybe the overall job situation is fine, and academia just can't offer an enticing career ladder anymore.

  • dr24hours says:

    When academics act like a faculty position is the end-all be-all of science and people should be willing to work long hours for low pay and low job satisfaction because it's a "calling", and anything else is failure, they should be unsurprised at the number of PhDs who call bullshit, and work better jobs with more scientific freedom and better pay with decent work-life balance.

  • Busy says:

    Over the last three years we had four very bright students finish their PhD and go to industry, not so much because they wanted to but rather because they didn't want to play the PostDoc for eternity game.

    I don't mind people choosing their own non-academic career path because they wanted to. I mind missing the best brains because of ridiculous hurdles that reward the patient not the talented.

  • Amboceptor says:

    Giveaway is someone who doesn't care much about where or when the work gets published. More and more, publications are only a big deal for the career of the PI, and non-star "trainees" might as well treat the job as a technician does.

  • Amboceptor says:

    One sign is someone who doesn't care much about where or when the work gets published. More and more, publications are only a big deal for the career of the PI, and non-star "trainees" might as well treat the job as a technician does.

  • jim says:

    I would argue, as sincere as an Internet troll can, that the best Canary in this coalmine would be to look at fields like comp bio, where there has existed a steady-state 'exit' pipeline (tech) for the past 20 years. If things are dim, those fields will suffer first and hardest as the experienced trainees bail en masse. I'd predict that other fields where trainees have a harder time transitioning will show much slower drawdown times, as the exit pipeline will consist of younger trainees refusing to enter the field, rather than experienced people leaving en masse.

  • zb says:

    "The exit pipeline will consist of younger trainees refusing to enter the field, rather than experienced people leaving en masse,"

    yes, as EstablishedPI is reporting. I'd also suggest that an increasing proportion of foreign applicants/post-docs is a sign, because they are a subgroup in which the value of an American visa can be added to the direct compensation of the job.

    (that's not to say that foreign applicants/post-docs can't be great, but that the compensation rewards are different, if the visa has any value at all, which it does).

  • jim says:

    @zb Sure, but upstream indicators like 'decline in the quality of graduate applicants' could also be due to other universities increasing their trainee #. I'm guessing lots of Established PI-types were saying similar things about the 2002-era pools, right?

    Foreign post-docs is a good place to start, though. Not only is the visa of real value, many times they (Euro mostly) have reduced or no tax obligations, as well.

  • FU says:

    As someone who said fuck you to a five year post doc just to start applying for faculty positions I can attest that the vast majority don't even try. Sure, you jerk off the grey beard with the "I want to be just like you" just to get into a big name group at big name U. When I started grad school, a two year post doc for my field was the norm. Now you have to spend a year writing the F32, spend year 2 and 3 spending those dollars while writing the K99, which can't kick in till the F32 dollars are all spent. 5 or 6 years in pd hell, of course the vast majority don't even try any more. The cull is happening and needs to happen. I just got fucked by the timing. My boss gets daily emails from pd who are losing funding and want to stay in the Boston/Cambridge area. The system is fucked. The best and the brightest can see that.

  • becca says:

    You know how you feel like you want to cry when you realize NIH isn't actually collecting data on grants? That's how I felt when I realized *nobody* is actually collecting data on grad student/postdoc career outcomes, like the kind of data that could actually empirically answer this question properly.

  • Lady Scientist says:

    My hypothesis: all of the smart potential trainees are reading up about the funding situation in science and moving on to other careers. Those who aren't reading up on the funding situation are the ones who decide to pursue a career in science (and, by default, they aren't the smart ones).

  • Ola says:

    @FU
    FFS what is the problem with a 5 year postdoc? Why is every PDF who's 3 years deep shitting their pants about the godawful prospect of a second post-doc?

    Jim and Lady Scientist have it right - the smart ones are already out of it and don't bother applying. This actually makes interviewing for grad school real easy: "Why do you want to do this?" Any answer other than "because I am totally fucking in love with this job and can't envision myself doing anything else for the rest of my life and I don't care if I go broke/mad/insane/divorced trying", is a fail. That's what it takes.

  • Grumble says:

    @FU: 5 or 6 years in pd hell, of course the vast majority don't even try any more."

    Ho, hum. How come I rarely hear surgeons complaining about their 5, 6 year residencies? Maybe cause they're committed and you aren't.

  • DJMH says:

    Grumble: there's not a lot of surgeons who go through residency and then can't find a job though, are there? OK then.

  • DJMH says:

    @Ola, that's a recipe for monasteries or starving artists, not jobs.

  • FU says:

    @chOLA & dUMBLE

    How fucking stupid are you? "I'm so cool and being a professor is so awesome i will suck whoever s dick it takes...." DJMH has a brain. You get a fucking job after a residency. For surgeons 350,000 fucking dollars/year starting. After a 6 year pd you get to enter a fucking lottery where the prize is ~80,000 with a few years of summer salary till you get funding ( if you get a real job and not just jerking off at some soft money). You people are delusional if you still think people with half a fucking brain are in the PI game anymore. Unless you need a green card, the big boys/girls have woken up to the reality.
    The fucking pretentious idiocracy amazes me. " Don't wanna sound like a dick or nothin', but it says on your chart that you're fucked up. [edited -DM] What I'd do, is just like... ha ha... like... aha... you know, like, you know what I mean, like... haha..."

  • jmz4 says:

    "FFS what is the problem with a 5 year postdoc? Why is every PDF who's 3 years deep shitting their pants about the godawful prospect of a second post-doc?"
    -It really depends on your personal situation. If you're living in a big city with a spouse and or child, then the prospect of waiting till you're 35-38 to start saving any kind of money is pretty unappealing to most sane people.

    "That's what it takes."
    -That's what it takes to breed a mentally unbalanced, intellectually deficient workforce, yes. You don't want to just have the crazies entering the pipeline.

    That being said, I disagree with the trainee issue. Those of you that are professors could probably get some quick stats from your schools, but when I was on the admin review committee (3 years a go, as a grad student) I hadn't noticed any drop in the past 5-6 years in any quantifiable measure of applicant quality (GRE scores, GPA, papers published, etc). I think that's because there's always the indefatigable arrogance of youth to account for. Back then, I figured I would shatter all the averages and be a professor by the time I was 28. Most of my cohort had similar (highly unrealistic) aspirations. Science is always going to attract people that think they're the outlier, and some of these people will be right (or at least as good as historical applicants).

    To DM's question, my guess is that applicant quality to faculty jobs will remain stable even as things get worse because so many of our measures are relative. If paper quality goes down the tubes, then more mediocre papers will start finding their way in top journals. If the postdoc pool gets worse, then 1-2 papers will become the norm, etc. Is there anything that hiring committees use that is truly objective?

    To be honest, (and speaking as a postdoc) I don't really see the hypercompetition for PI jobs as a problem that will damage the quality of applicants. I know plenty of top-notch people willing to stick it out as long as it takes. But the holding tank for those jobs has few outlets and is fairly cramped and miserable, and is leaking talent that could be put to better use.

  • CD0 says:

    I would not change my academic freedom and the creativity of my job for any better paid position dealing with golfers of the type that I see in the nearby business school.
    I agree with OLA: What I do has no other significant reward than satisfying your curiosity and your passion for discovery. Also gaining the respect of your peers; the only ones who understand what we do. Is it enough for most people? clearly not and I have lost very talented individuals who could have made it in academia and decided not to do it. But I certainly do not envy any of them.

  • drugmonkey says:

    jmz4-
    And are you not concerned that hypercompetition may select for unwanted traits such as data faking?

  • Busy says:

    Unless you need a green card, the big boys/girls have woken up to the reality.

    I've came to that realization in the mid 90s, when things turn sour in our field. For other disciplines this has been very much the case since the late 70s.

    And are you not concerned that hypercompetition may select for unwanted traits such as data faking?

    Or schmoozing or trend chasing. Back when there were enough jobs to go around you could still land an ok position if you were a decent researcher but perhaps not the most likeable fellow. Not so anymore. Now you need to go to finishing school in non-academic traits to have a shot at a position, with the number of candidates being so much larger than the number of positions.

  • Grumble says:

    @FU: " You people are delusional if you still think people with half a fucking brain are in the PI game anymore. Unless you need a green card, the big boys/girls have woken up to the reality. The fucking pretentious idiocracy amazes me. "

    Relax. All I said is that you aren't committed whereas medical residents are. That is, in fact, the truth, by your own admission: you are not committed to a 5-6 year post-doc. So why the pissy-ness?

    So, as to your comment above, implying that only stupid people are left in science because... well, I don't know why because. It's hard? It's risky? It doesn't pay as much as 50 other professions? Whatever. It doesn't matter because it's demonstrably false. There are many absolutely brilliant people in science, from students to emeritus profs. Why do they put up with the difficult career path? Because they know they have a unique aptitude for it and they have a burning passion for science.

    Yes, a lot of people (like, apparently, you) get discouraged and even angry that you have to fight so hard for what seems like so little, but that's unfortunately just the process of weeding you out. For the rest of us, we stay in the game because what we get out of it is actually not so little after all. We value our work more than money. If you don't, then maybe you should have opted for a career as a surgeon or a banker instead.

    Finally, don't take any of this as defending the system. I'd LOVE for it to be easier for everyone. But that would take more money for more academic research positions, and even that might ultimately not fundamentally change the system.

  • FU says:

    @ Grumble: My apologies. I might have had one too many glasses of wine last night. You are correct; there are a lot of brilliant people with academic careers. That totally misses the point. Of course a lot of brilliant people IN THE PAST would want to go into academic science. We are talking about the present/future. Again, the cull is happening people. It is blatantly obvious for people "on the quad" vs those "off the quad" where I am. Of course this will discourage a high percentage of capable and bright people from committing several years of the prime of their life to what is essentially a fools errand. And yes, there will always be some who don't care about the realities. They are a small minority, who are not the cream of the crop. That is the future. A race to the bottom.

  • FU says:

    Oh, and it is easy to be committed when 350000$/year (and academic freedom) is at the end of the tunnel.

  • jmz4 says:

    "And are you not concerned that hypercompetition may select for unwanted traits such as data faking?"
    -It will, but increasing the number of PI's just exacerbates the problem, since they go and train more postdocs. The fix, therefore, is to change the postdoc, so there's not so much drive to escape it by fudging data.
    Whether you think its justified or not, most of the postdocs are bitter about the low pay and lack of autonomy/career control, and that pressures drives a large part of their desire to become a PI* (to escape this situation). Since we can't fix the bottleneck, the other source of the pressure (since it makes escape harder), we have to make the internal pressure of the holding tank more bearable.

    And these issues can be alleviated without increasing the number of PI-ships, or even altering funding very much. Make postdocs a 2 or 3 year training grant-like stint, at which point an adviser has to either nominate them as a research assistant prof (at the higher pay scale and employee benefits), fire them, or hire them as lab staff, like super-tech position (salary at the high end of the current pay scale). Forcing people to choose a route, by giving them viable career options that are laid out by university policy (not by rumor and hearsay), might help alleviate the "PI or bust" mentality of many postdocs who want to stay in academia but don't necessarily see other options for them. This will cool the flow of underqualified or uninterested applicants to PI jobs, as well as ease the perma-postdoc problem by settling them into more permanent academic jobs. And if these jobs exist, then qualified undergrads won't be quite so scared to go into grad school.
    Getting the universities to take a little more responsibility for their postdoc's training and career options wouldn't hurt either.

    "Another way might be if a flurry of postdoctoral solicitations in your subfield appeared."
    -What did you mean by this. PI's advertising postdoc openings? Why would this indicate a reduction in trainee quality?

    *Which, I think, might be why you see so many poor applicants these days. They wandered into the postdoc out of grad school to feel it out, and then felt locked (sunk research costs, I know its not really valid reasoning) into pursuing a PI-ship, even if they didn't really think they were suited for it. Its the only explanation I can think of for the postdocs I talk to that have no interest in one or more the key PI responsibilities. "Oh, I'm a postdoc, but I hate mentoring, grant-writing and giving talks."

  • poke says:

    ... most of the postdocs are bitter about the low pay and lack of autonomy/career control, and that pressures drives a large part of their desire to become a PI* (to escape this situation).

    Where do assertions like this come from? I don't want to claim that my experience is universal, but I do not see this. At all.

  • drugmonkey says:

    We absolutely need to develop nonPI career scientist positions. I agree entirely that this is a huge problem with the workforce.

  • Yeah -- career staff scientists do exist in NIH intramural research and at the National Labs -- while there isn't really anything like tenure for them, they seem to be more stable than the typical soft-money position.

  • dr24hours says:

    There are lots of those positions in industry. I think the title of the post suggests the myopia: maybe trainees aren't "bailing" on anything. Maybe they're pursuing something you don't care about or pay attention to.

  • rs says:

    "There are lots of those positions in industry"

    and most of these jobs do not require a PhD. A master degree and on the job training by the company is good enough to start. Why waste 5-7 years?

  • BugDoc says:

    @becca: "*nobody* is actually collecting data on grad student/postdoc career outcomes, like the kind of data that could actually empirically answer this question properly."

    We are and have been for the last 15 years. But here's the problem: career outcomes are a lagging indicator. The people who graduated 10 years ago mostly moved on to do postdocs (whether they wanted an academic position or not) and only in the last few years have they moved on to more permanent positions of different sorts. To look at career outcomes from even a few years ago may be to underestimate the magnitude of the over-training problem. Also, in tracking career outcomes, as postdocs pass the 5 yr point and can no longer be officially called postdocs, they often transition into other transient soft money positions like "research investigator" or "senior research associate", etc. In career outcomes tracking, our graduate administrative is still calling these "academic research", even if they are not stable career tracks. That makes it look like quite a high % of our trainees are going into "academic research" as opposed to the 15-20% that were able to achieve more stable positions in academia.

    @dm "We absolutely need to develop nonPI career scientist positions" - since these are all soft money positions, until funding stabilizes how do we make this happen?

  • Lady Scientist says:

    I agree with dr24hours (sorry - couldn't figure out a way to abbreviate that), and jmz4's comment about universities needing to take more responsibility for postdoc training/career is *exactly* what I was about to write. Don't leave that up to individual PI's, exclusively, because then you have ripe conditions for abuse and temptation for data fakery.

    Also, BugDoc's point about soft money positions is a good one.

    Basically, what it boils down to is the need for massive change at the level of universities and medical colleges, but they are all prioritizing, more and more, profit over anything else. If they can take money for tuition, from the NIH, from wealthy benefactors, from sports shenanigans, etc., they will. If you can't bring in money, then sayonara, dude.

  • There is a bifurcation occurring in biomed science, just as there is throughout society. Those brilliant young scientists who are plugged in to the BSD coastal network and do their PhDs and post-docs in such labs have a very reasonable chance of becoming PIs themselves. The cull is not even, and is occurring at the next tier down (with the exception of certain well-known highly productive labs at second-tier institutions). It is no longer realistic to think that doing a post-doc at random single-R01 lab in Ohio is a route to scientific independence. But the elite is doing fine.

  • drugmonkey says:

    since these are all soft money positions, until funding stabilizes how do we make this happen?

    My proposal is to use a renewable K mechanism to do it. The scientist has to have sponsoring R-mech to tie up to but this can be flexible and easily transferable. So if one PI runs out of funding they can look around on campus for a good fit. Even take it with them to a new city when their spouse's job transfers them or whatever.

    competitive review every 5 years but it is based on being productive, not on some sort of wow factor or expectation of "independence" or whatall.

  • drugmonkey says:

    what it boils down to is the need for massive change at the level of universities and medical colleges

    There is a reasonable alternative to this. Which is the NIH stepping up and taking greater responsibility for what is in fact *their* workforce, no matter the thin fig leaf of "extramural" definitions.

  • Lady Scientist says:

    Yeah, I agree with you, DM. But, like that's gonna happen.... Not that the institutions changing is likely to happen, either.

  • jmz4 says:

    "Where do assertions like this come from? I don't want to claim that my experience is universal, but I do not see this. At all."
    -I recently went to the Boston Area Future of Research symposium which was a gathering of postdocs and faculty to draft a white paper discussing how to fix the postdoc (http://futureofresearch.org/). The Q&A session quickly became a rant-fest of disgruntlement. Also, I've talked with the organizers of the conference, who had a section for open comments from registrants, and they said, quote, "wading through those comments was like going through the Swamp of Sadness in the Neverending Story."
    Then there's the recent NPR story (http://www.npr.org/blogs/health/2014/09/16/343539024/too-few-university-jobs-for-americas-young-scientists), and my own anecdotal evidence.

    My note that many postdocs don't seem to actually want to do the things PIs do is much more anecdotal.

    "and most of these jobs do not require a PhD. A master degree and on the job training by the company is good enough to start. Why waste 5-7 years?"
    -The fact that your hypothetical person would look at doing a PhD as a "waste" of 5-7 years is kinda of my point. If the idea of doing a substantial research project doesn't excite you, you really shouldn't be in the PI pipeline to begin with, so I don't think we're "losing" those people.
    Also, if you go into industry with only a masters, my impression from my brief stints is that you have to be exceptional to be promoted above the people with PhDs and MDs. You'll make a good salary, but you'll probably be doing someone else's benchwork your entire career (which is totally cool if that's what you want to do).

    "My proposal is to use a renewable K mechanism to do it. The scientist has to have sponsoring R-mech to tie up to but this can be flexible and easily transferable. So if one PI runs out of funding they can look around on campus for a good fit. Even take it with them to a new city when their spouse's job transfers them or whatever. "
    -The paylines would have to be very forgiving. One of the benefits of having permanent research support (super-techs), is that their insensitive to sensationalize results is much lower, since they're not out to personally gain from the impact factor of the paper. I'd also like to see a system where the NIH puts some of these skilled experimentalists on retainer for experimental duplication of published results. They're the only ones likely to be able to do it, since they don't need to churn out papers for their career, and they have the technical skills to make a negative result convincing. Not quite sure how that would best work, though.

    Lady Scientist, do you think the PIs are the one's likely to initiate data tomfoolery? The postdocs have a lot more reward for comparatively little risk (they'll probably be gone by the time someone figures out their work is flawed).

  • Cynric says:

    But the elite is doing fine.

    I agree with this sentiment, but when the elite becomes so incestuous and isolated from a wider pool of diverse talent, there's a big risk that you'll no longer be selecting on the basis of scientific talent. Patronage is rarely synonymous with meritocracy.

  • Cynric says:

    The cull is not even, and is occurring at the next tier down

    To expand on this: I think the elite need to worry about this cull too. Although there isn't an immediate threat to their livelihood or research programmes, renewal of the elite depends on a large pool of talent spread over the lower tiers. If the other tiers are undermined, there is no prospect of new highly-productive labs emerging in second tier institutions, no prospect for unconventional career trajectories, and no safe space for individuals that dislike hypercompetitive environments. No hybrid vigour.

  • becca says:

    BugDoc- I am aware individual grad programs (and fellowships) sometimes do track this info- which is great.
    I am also aware the info is difficult to parse on occasion, I'm ok with that. I think, fundamentally, we're not at a stage yet where I *care* about how individual grad programs or fellowships perform- I want to be able to get a gestalt feel for the entire enterprise (this is the best example of the kind of analysis I think makes sense right now http://www.ascb.org/ascbpost/index.php/compass-points/item/285-where-will-a-biology-phd-take-you- but, of course, that figure raises almost as many questions as it answers).

    It's a bit like the situation for teacher evaluations in the K-12 arena- I don't particularly like the politics (read: evil intentions) of those who are trying to ram nonsense politically motivated VAT formulas through state legislatures to bust unions. At the same time, the data they are collecting on this generation of students is *astonishing* in it's breadth- it's like the equivalent of genomes suddenly available in biology. It turns out that analyzing/annotating the genome info is the bottleneck, and I suspect the same will be or already is true for the study of education and research as processes. We're idiots if we can't develop integrated ways to evaluate outcomes of processes from this much data, but I don't think we're quite at the level of resolution people would really like yet (e.g. "is prof A better at producing good scientists than prof B?" or even "is grad program A better at that than grad program B?"). I just want the data to answer this type of question: "It seems like a disturbing amount of the people that wash out of this type of program are women and minorities... is that objectively more than what we'd expect and if so are there programs that look like exceptions and can we learn from them?"
    If CPP is right (and I hate to say it, but I suspect he is, though he may underestimate how elite you have to be to be "fine"), it's even *more* important to be able to define what is functionally read as "elite" by this system.

  • drugmonkey says:

    The "elite" can sod off.

  • MorganPhD says:

    Point #1: The current postdoctoral system is setup to obfuscate postdoctoral outcome tracking, even within a single institution.

    My university has 3 different names for postdocs, depending on the funding source. We have postdoctoral researchers (PI-paid), postdoctoral trainees (T32-trained), and postdoctoral fellows (any other personal grant paid, ex. NRSA F32). Then, after 5 years of "training" you must become a "Research Scientist", whilst usually performing the same job with no independence. Are research scientists still postdocs if they are in the same laboratory? I would argue yes.

    Point #2: The current postdoctoral "training" system lacks consistency, structure, and formality.

    Postdoctoral training could (and should) be setup more like PhD-training OR more like an apprentice-ship system. In the PhD training-like model, postdocs would have formal committees. The threshold for continued training would be significantly higher than what we do for PhD students. Postdocs would be required to take courses in ethics, publishing, the grant system, statistics, budgeting/accounting, and laboratory management, demonstrating their proficiency in the tasks that PI's ACTUALLY perform. Bench science should be the main focus, but perhaps no more than 75% of total "effort".

    In the skilled trades (and throughout history), a single artisan or craftsman might have a single apprentice that he/she provides with in-depth, hands-on training. The craftsman might also have a set of workers performing the other tasks and running day-to-day operations and working on less complex tasks, but he wouldn't consider them all apprentices.

    In the current system a single PI might have 10+ postdoctoral trainees, which leads to...

    Point #3: Postdocs are exploited employees in a labor system that values cheap labor over training.

    All you have to do is look to the ratio of grad student/postdocs:staff scientists/technicians to understand that the current system views "relatively" low paying postdocs as employees and not trainees.

    This, to me, is similar to the current problems with internships/co-op experiences in other fields: interns, like postdocs, are performing labor which should be reserved for paid employees, and companies exploit these trainees work cheaply.

    (Pretty controversial) Point #4: The nearly unlimited supply of foreign postdoctoral trainees shifts the American academic system out of equilibrium, makes outcomes tracking more difficult, and leads to radically different incentives for postdoctoral training.

    It probably sounds insensitive, xenophobic, and (hopefully not) racist, but the unlimited supply of foreign postdoctoral trainees is a real issue in US system. The rationale that US academic science is a special flower that needs to be allowed unchecked J-1 Visa so it can "help human health" and attract the best and the brightest into STEM is patently untrue, and is a lie propagated to keep Point #3 rolling. Why academic science and not car manufacturers or banks? Why do we not want to recruit the best and brightest into Goldman Sachs or Ford?

    The large majority of scientists, perfectly happy with the postdoctoral system, would collectively flip out if CitiBank, Google, or Haliburton started a graduate postdoctoral training program, with below market pay and unlimited J-1 and H1-B visas. Or if McDonald's lobbied congress to start a training program which allowed for 1000's of J-1 Visa holders to "train" here in the US.

    Ultimately, I think this is realistically only a small problem with postdoctoral research, but if we agree the current system is broken, difficult to track, and needs reform, we have to look at the entire system, how it functions, and what could theoretically be changed to make it better.

    Of course, that leads me to another potential post, titled "The current academic system is filled with middle managers, and I'm not talking about the boogeyman administrators, or Why I think principal investigators do too little science and too much management."

  • louis says:

    Morgan,

    I like your assessment of the postdoctoral system and agree mostly with it. I am not too sure I understood your view on the "unlimited supply of foreign postdoctoral trainees". For many years, the US had been considered the 'mecca" for scientific research and the aspiration of many young PhDs was to come to the US for a postdoctoral training, particularly in the biomedical field. Some chose England or Germany. I think that the mobility and intellectual exchange that comes with it is great and have served very well the US, as well as other host countries. I think that now the oversupply has become global and, perhaps, solutions ought to be thought and designed globally.

    In recent days I've heard that some big companies, one of them Apple, are lobbying Congress to change the H-1 visa rules as to allow qualified workers to join their workforce as to favor and increase US competitiveness.

  • MorganPhD says:

    Louis,

    I was scared to come off as xenophobic, protectionist, and just plain too harsh when writing on foreign nationals in the US academic system. I think the key is not to look at this as a "they are here to take our jobs" problem, rather one of "what is the right balance of attracting top talent while also being stewards of US taxpayer dollars and giving US students appropriate opportunities".

    This is a national immigration and economic issue, but one that is being focused on quite a bit in the STEM fields. Just look to the ideas like "staple a green card to a PhD" and "we need to increase US participation in STEM".

    Do we really have a shortage of STEM workers and PhD-level scientists, thus we need to change our immigration policies? When businesses say "we can't be competitive in a global marketplace" or "US workers aren't qualified and we need foreign workers", are these things true or is it THAT WE REALLY DON'T KNOW WHO IS PARTICIPATING IN THE MARKET and WHO MIGHT BE QUALIFIED because WE DON'T TRACK ANYTHING.

    The original question posed by DrugMonkey is hyper-relevant and important and is something we (as a field) should strive to be better at.

  • drugmonkey says:

    It is certainly difficult to discuss worker protectionism (which I favor) in the biomedical sciences without appearing to direct one's attention at scientists/trainees who arrive from a specific set of countries.

    I agree that this is a discussion that needs to be had, nevertheless.

    All the blah-blah about free exchange of ideas is nice but it does serve in large part as cover for the labor exploitation that is inherent in this, and any, immigrant-labor situation.

    Personally I would welcome CongressCritter attention placed on this issue.

  • @MorganPHD
    "Why do we not want to recruit the best and brightest into Goldman Sachs or Ford?"

    Don't know about Ford, but Goldman Sachs already does employ plenty of non-Americans. I recently read Michael Lewis' "Flash Boys" about the problems and dangers of computational stock trading and noticed how many of the "flash boys" themselves were from Russia or India...

  • jmz4 says:

    " The nearly unlimited supply of foreign postdoctoral trainees shifts the American academic system out of equilibrium, makes outcomes tracking more difficult, and leads to radically different incentives for postdoctoral training."
    -Could you elaborate on this more? I don't see the clear line between increased foreign visas and poor applicant quality to PI jobs. Or even necessarily how it drives postdoctoral exploitation, since the only way it would do that is if it were masking an actual scarcity of US born postdocs. We have the postdocs we need to do the science the NIH has funded. If we cut back on foreign postdocs, who will do that work?

    Totally agree with you on points 1-3. A more structured postdoctoral program with clearer expectations, benchmarks and off ramps would make things a lot easier and more efficient. And labs need to be restructured to remove at least some of the experimental burden from the postdoc and retain in-house bench expertise.

  • Established PI says:

    It is important to approach the issue of foreign postdocs cautiously given the lack of hard data. It is undoubtedly the case that some foreign postdocs go on to take academic or industry jobs that would otherwise go to U.S. citizens. But science is an international endeavor and I can't imagine anyone is advocating drastic protectionism or closing our scientific borders. We benefit as a nation from attracting the best and the brightest from around the world, and indeed we owe a great deal to immigrant scientists who have come here. Were it not for the large number of scientists who came in the period surrounding WW II, American science would look very different today.

    My perception, which is by definition anecdotal, is that faculty and industry positions go overwhelmingly to U.S. citizens and permanent residents, and that foreign postdocs are in fact at a significant disadvantage in competing for these jobs. If anything, I view at least a subset of foreign postdocs as exploited labor who often float from one position to another just to hold onto their visa status, in the often futile hope that they can find a job here. With no NIH rules limiting how long someone can still be considered a postdoctoral fellow, foreign postdocs can still be used as cheap labor. My own institution, which prides itself on having been a leader in improving the lot of postdoctoral fellows, still has all sorts of loopholes that enable PIs to pay someone on their second or third postdoc the base NIH salary if you can claim that they are being trained in a new skill.

    As suggested by several commenters, the NIH needs to have a clear policy that forbids paying anyone as a postdoc after a certain number of years post-Ph.D. and that requires something akin to a (post-) thesis committee to oversee postdoctoral training. This won't solve the problem of the lack of jobs for independent investigators, be it at the university, research institute or industry level, but it will go part way towards addressing concerns regarding postdoctoral training as well as low morale.

    One of the big problems here is the lack of data. The NIH does collect data on students, particularly those on training grants, but lacks data on students supported by other grant mechanisms. As far as I know, there is very little data on the vast majority of postdoctoral fellows who are supported by grants or institutional funds. The NIH can't do anything about the latter but it could start to track the former by requiring every single postdoctoral fellow that is paid by NIH money to have an ERA Commons ID and to track their employment and salary annually. This would at least make it possible to figure out how long postdoctoral fellows stay in the NIH-funded system (or enter and exit). This data still won't be complete as it will miss postdocs with private fellowships or paid out of start-up or institutional funds. The NIH could try to add some of them, too, if they required anyone working on an NIH grant, even with zero salary, to be entered into the system as well. Again, the data won't be complete but it would at least be a start. It wouldn't be a bad idea to partner with the NSF so that postdocs they support could similarly be counted, perhaps on research.gov.

    Without more data, and better quality data, it is going to be hard to come up with good policies to address the very serious workforce issues we are facing.

  • MorganPhD says:

    @ Jonathan Badger
    I wholeheartedly agree with immigration policy that allows US companies to attract top talent to further their economic interests. While the H1-B visa program has numerous faults and is frequently abused, there are safeguards and regulations (at least on paper) that help protect wage deflation and American jobs.

    This program is set-up as "we need the talent that doesn't exist in the US and the people we are bringing to the US specifically to work are already trained and capable of performing these tasks". In contrast, the postdoctoral system is specifically recruiting workers under the guise of "training", as in these workers do not yet have the skills to do the job and we are going to use US taxpayer dollars (mainly through the NIH) to do that.

    This is intertwined with the comment above from @jmz24. " We have the postdocs we need to do the science the NIH has funded. If we cut back on foreign postdocs, who will do that work?"

    This is the problem as I see it, and is not necessarily related to the foreign trainee question. The solution to this problem is that you hire employees. The postdoctoral workforce should not scale with the number of grants in a lab. The conversation should not be "well, I have a new grant, time to hire a postdoc to do the work". We should be thinking "do I have the time, resources, experience, and opportunities to train a postdoc, because I can hire a staff scientist with demonstrated 5 years of experience in this field to do the job instead"

    I'm in favor of limiting the postdoctoral trainee #/lab to 3 total, regardless of the number of grants/funding. This would at least (on paper) increase the time a PI might have to "train" a postdoc. This is a pipe dream of mine, obviously.

    @ EstablishedPI
    We absolutely benefit from recruiting the best and the brightest to the United States, and indeed, I think that this needs to continue. I, in no way, am trying to minimize the contribution made by foreign scientists who immigrate to the US. I do not believe, though, that it has to continue using the current method in which we do it.

    I personally favor a strong reduction in the number of total postdoctoral trainees, with cuts to both foreign and domestic trainees, with the actual "best-of-the-best" going on to postdoctoral training. If we can't identify great scientists by the end of their PhD or help PhD students find employment after 6-7 years, then we have to admit we are doing a poor job of training them to be successful.

    As a political and economic issue, the current and pervasive STEM meme of "we need to have a competitive workforce and we need to recruit more people into the STEM fields" is not backed up by actual data. I think that having data about trainee outcomes is central to confirming or rejecting this hypothesis and solution.

    The amount of time needed to efficiently track postdoctoral and graduate trainees would be approximately 15 minutes/year for an individual PI. Each year, a university that uses NIH funds would be required to submit a de-identified list of current PhD students, postdoctoral trainees, and FTE employees, along with their current immigration status. An individual PI would send forward that list to the Office of Research Services (or equivalent) and they would compile that list to forward to the NIH, who would then add a new tab to the Reporter page with this info. The penalty for non-compliance would be a reduction in NIH indirect costs, not direct. This would make the University care.

    Someone might say "well, the NIH can't actually put strings on the granted money." The key, then, is to make it a subsection of a grant application. If you do not supply that information, like an Aims page or Vertebrate Animal section, you do not get a grant.

  • Joe says:

    @MorganPhD "We should be thinking "do I have the time, resources, experience, and opportunities to train a postdoc, because I can hire a staff scientist with demonstrated 5 years of experience in this field to do the job instead" "

    Staff scientists and post-docs are generally not similar in multiple ways, and I don't find that you can substitute one for the other on a project. Furthermore, post-docs are expecting to stay for a few years and then move on, whereas staff scientists are expecting to stay for many years. Also, rules at my university make it hard to let an employee (staff scientist) go unless you completely run out of money.

  • wasapostdoc says:

    As a former post-doc who bailed after 3 years for a number of the reasons ya'll have discussed, I find the fact these conversations are happening encouraging for biomedical research. While I realize that any real change is likely many years in the making, kudos to you for bringing this issues out for discussion!

  • drugmonkey says:

    How are they not similar?

  • MorganPhD says:

    @joe
    When you say you can't substitute them on a project, do you mean you can't "financially" substitute them, or that the postdoc has a skill set that the staff member does not have?

    Maybe I'm a commie, but I think the job security is a good thing for the staff scientist.

  • E rook says:

    Many industries hire long term contractors to complete specified projects. I don't see why a contracting system for the staff scientist can't be implemented. Five years is plenty of time for them find new employment by the time the contract is up. If the project is renewed or a new grant funded, the PI could choose to renew the contract for the scientist. Unionization at some institutions might prevent this. But as long as the NIH extramural system is "project based," I don't see how PIs could reasonably hire a large permanent staff. It is even tacitly acknowledged in some labs that the lab tech is typically a recent BS recipient in a holding pattern till med school...designed this way to keep the salary low (in order the keep a reasonable budget). In other labs, a permanent scientist position might be there, but I have seen their "effort" reduced to fit the budget, despite an annually increasing salary, it's kinda pathetic.

  • former staff scientist says:

    I thought a staff scientist was just a post-doc whose plan of being out in a few years did not work out. That is what happened to me in the sixth year of my postdoc. I got "promoted" to staff scientist and did not even know it until a few months later. Terms like "staff scientist" is part of the NIH strategy to hide the problem of the increasing number of post-docs turning into perma-docs. If you change the name, the problem is solved.

    I have seen a lot of good suggestions for solving the post-doc glut problem, though the real solution would be to choke off the excess earlier in the process, i.e. reduce the number of graduate students. Regardless, I don't think any of these solutions will be implemented, because the NIH wants to maintain the current system. The Ph.D. glut in the biomedical sciences has been going on for at least 20 years, and little if anything has changed. ~15 years ago the National Academy of Sciences recommended that we begin reducing the number of PhDs we train. The NIH response was to say in essence "you never know when you might need some extra PhDs."

  • Jonathan says:

    @MorganPHD, odd that you bring up McDonalds because that's exactly what they did: http://www.npr.org/2013/03/18/174410945/u-s-probes-abuse-allegations-under-worker-visa-program

    As a former foreign postdoc who was very involved in the whole broken pipeline issue (until I realized it wasn't ever going to get fixed and therefore decided to stop caring about it so much because it just makes me angry) I don't see how things will improve until the numbers are limited (and universities are also allowed to bring in H1-B workers without being subject to the cap, fwiw).

    Becca - tracking outcomes is extremely difficult. For starters, how do you do it? You'll want to track people for at least 10 years, but how do you stay in contact with them? Particularly the ones that come from abroad and then leave again? Or the ones who leave the field (80% of us and counting) and who think "eh, I've got nothing to do with academic science now why should I bother filling in this survey?"?

  • becca says:

    Resolved: All postdocs will cost grants at least as much as an entry level tech. Universities will be legally required to put the difference in fringe rates into an insurance fund for postdocs, that pays for short term salary support such that international postdocs do not have to worry about their visa expiring as soon as a grant runs out. All postdocs (domestic and international) who prefer to take a lump-sum for moving expenses instead of the unemployment benefit may do so.
    Winners: foreign postdocs who no longer work for jerkface PIs, domestic postdocs who can take jobs in flyoverlandia without worrying about how they will escape once their job is no longer funded, non-jerkface PIs in flyoverlandia
    Losers: jerkface PIs who can only lure people into working for them by being located in a desirable locale

    The only non-xenophobic way to address the problem is to view it as a social justice travesty that foreign postdocs can be forced to work for despicable PIs simply to avoid being kicked out of the country. The fact that this *also* depresses all postdoc wages is important to me, but it's not why there is a collective moral responsibility to act.

    #VoteForSciliz2016

    Jonathan- Google knows what I want for breakfast before I do. We have the technology! In all seriousness, I know this kind of work is challenging. But I'm not looking for all the data ever, it'd be a huge improvement to just get all grad programs to do what training grant funded programs already do. The real problem is not that it's difficult to track outcomes (I mean, it's a legitimate research topic in its own right, and there are nuances of statistical analysis that are very interesting and all that, but again, I'm not looking to reinvent the wheel here), the real problem is that it costs money.

  • MorganPhD says:

    @becca
    Entry level technicians at my (I guess) elite East Coast university cost grants less than $30,000/year at our allowable fringe rates. So postdocs make a bit more than that off the bat. I'm not necessarily worried about current postdoc pay, as I think that number goes up naturally by enacting other reforms to the system. I'm actually one of the few postdocs who think we make more than enough money, even in Big Coastal City USA.

    I don't think targeting foreign trainees as a source of potential cuts in a revised, more lean postdoctoral training system is xenophobic. We already restrict NIH F31 and F32 awards to US citizens. I don't think it's xenophobic to say that academic science is not a special flower and shouldn't be able to do things no other job field is able to do (J-1 visa, Professor and Research Scholar class, 31,868 sponsored members in 2012). Everyone is a protectionist when it comes to their own job or future, regardless of their political leanings.

    My idea is that both domestic and foreign trainees should have to demonstrate they are the best and the brightest before proceeding to postdoctoral training, as opposed to just being warm bodies to perform work for a cash-strapped PI. How many of us have settled on "good enough" for a postdoc because we need someone to do the work we have money to do?

    I do think that we need to revise the system to be more like PhD training whereby an individual PI does not make renewal/firing decisions. It almost takes an executive order to kick a graduate student out of their PhD training, and I think the same should be true for postdoctoral training. Having your postdoctoral training renewed and monitored by a committee should alleviate some of the fear of losing ones job and subsequent deportation, while also, as I suggested earlier, making the training experience better by having multiple PI's to interact with throughout your early career.

    Sometimes I agree with DrugMonkey about Congress Critter; maybe they should get involved more with our field. Scientists (and professors in general) love citing that academic freedom and self-governance are important issues and make us special, and then completely botch the self-governance part (broken pipeline and poor outcomes for trainees, peer review problems, the great Impact Factor publication chase).

  • Jonathan says:

    Becca - grad students who were trained in the US are tracked like that, the data go into NSF's Science and Engineering Indicators. The issue is tracking postdocs, half of whom are foreign (many only here temporarily), classified with myriad different job titles (fellow, trainee, scholar, whatever). It was something we wanted to do at the NPA back when I was involved with the organization but the sheer difficulty of it was always overwhelming.

  • E rook says:

    "My proposal is to use a renewable K mechanism to do it. The scientist has to have sponsoring R-mech to tie up to but this can be flexible and easily transferable. So if one PI runs out of funding they can look around on campus for a good fit. Even take it with them to a new city when their spouse's job transfers them or whatever. "

    So like a renewable K02? Or tweek the K02 to make it more flexible. Or expand its deployment. I would support that.

  • drugmonkey says:

    It will be a very different type of approach so there is no need to say "like a K02" or "like a K05". Sure, it's a K and it's renewable. But let's not be constrained in thinking what is "possible" by the way K's are used now.

  • jmz4 says:

    "I thought a staff scientist was just a post-doc whose plan of being out in a few years did not work out. That is what happened to me in the sixth year of my postdoc. I got "promoted" to staff scientist and did not even know it until a few months later. Terms like "staff scientist" is part of the NIH strategy to hide the problem of the increasing number of post-docs turning into perma-docs. If you change the name, the problem is solved. "
    -There's nothing wrong with the staff-scientist position, but yes, the term does end up mostly encompassing the type of person you describe. But that's mainly because our current system is so inflexible, that if someone with a PhD wants to do benchwork in an academic lab, becoming a postdoc is their only real option, whether they want to be a PI or not. I know plenty of people that want to do a 2-year stint to round out their skill set or move closer to a medically relevant discipline and then go to industry. Do these people need to be postdocs?
    Additionally postdocs should be getting more training and experience in grant and personnel management. Since we still need the same amount of research done, the only way to help shift this burden is to recruit skilled technicians, which would conveniently provide the personnel management experience as well as diversify the career options for PhDs.

    In short, if you want better applicants to PI jobs, start training them while they're postdocs. If you're a PI and you want them to become good PI's, make sure they understand how your grants work, and give them some people to work under them.

  • DJMH says:

    If you do not supply that information, like an Aims page or Vertebrate Animal section...

    I think the info on trainees should be included IN the Vertebrate Animals section.

Leave a Reply