K99 eligibility limited to 4 years and you better not be out on the job market

Somehow, the changes announced in this Notice are supposed to help postdocs make a "more timely" transition to independence.

I think the main problem will be that reviewers are already largely biased with the mindset that a certain publication profile is necessary to make someone competitive. Meaning a certain number of papers in a certain "level" of journal. In order for the timing of this process to work, postdocs will need to start applying for the K99 in about year 2, by which time it is very unlikely that anyone but the ultra super productive (which is often aided by the good fortune of having a PI who lets you choose quick-to-paper projects over those that they might WANT you to be working on), or those in big-name labs, will have such a profile. Since reviewers won't have as lengthy of a publication record to go by, they will almost surely fall into the habit they do for every other kind of NIH grant and go instead on other aspects of the CV (and the CV of the mentor), further pushing this towards a "glamour" award. The "rich get richer" situation will be exacerbated, in contrast to what I have always seen as an advantage of the K99 award (that even the not as "fancy"--like myself in a lot of ways, lol--can have a fighting chance).

Also, year 2 is when a lot of people become most competitive for the F32, so it might become a choice between F32 or K99, which seems kind of stupid. Is F32 going to end up as some kind of consolation prize for not being fancy enough for a K99?

Not only that, but this will disproportionately disadvantage women (and men who are primary or co-primary caregivers) who have children during their postdoc years. So far, the longest extension on K99 eligibility that I have heard of anyone getting for family leave is the actual number of weeks/months they were out on parental leave. Anyone who has had a kid knows that the effect on your productivity goes FAR beyond those few weeks/months. If they want to avoid this kind of bias, they will need to get real about extension times--people should get at least a year per kid, just like in the tenure clock stoppage situation.

Lastly, I think it is paternalistic and invasive for Program to be making judgements about someone's need for a K99 award and readiness for the tenure track based on their job application timelines. A large proportion of postdocs go out on the job market before they are truly ready because their PI won't or can't pay them anymore. Making that the postdoc's FAULT by now also telling them they are no longer eligible for one of the best options to win their independence from that PI is just gross and unhelpful. This treats postdocs like they are little kids who say they want dessert now even though they didn't eat their vegetables: "Well then, you must not need any dessert because you must not be hungry." Well, guess what: your reviewers are scoring highest the people who already look ready for an independent position. So you're going to have to figure out how to get them to change their mindsets--something that so far, nobody has figured out how to do very effectively. How is that going to be reconciled with the new rule that:

"Individuals who are close to achieving an independent faculty position, and cannot make a strong case for needing a minimum of 12 months of additional mentored training, are not ideal candidates for this award"?

I wonder if they have even thought about this. I also wonder if these changes are based on any actual metrics about applicants and awardees, or just some vague, poorly thought-out knee jerk idea to make postdoc-hood shorter. If you can show me Jeremy Berg-style data demonstrating that there will be some benefit to candidates, fine--but it just doesn't look like this is going to be a good thing.

25 responses so far

  • DJMH says:

    So the question is, what would be a better solution? I agree that I don't like the solution, because I think it ends up emphasizing grad productivity, hurting women with kids, and diminishing the chances of anyone who's taken a bigger leap in their postdoctoral training rather than going somewhere safe.

    But I do think they're trying to address a legitimate problem: that a bunch of K99 awardees are going on the market in the first year of K99-dom, indicating that they didn't need the award they were applying for, in essence. I even know one guy who had his K funded *after* he'd been offered a position at a top-10 university.

    The other problem is that the stated goal of the K99 is to push postdocs out of the postdoc earlier...but if the cutoff for application is 5 years, takes another year for the \( to start rolling, and then you're on the K for 2 years, that's an 8 year postdoc--not *anyone's* idea of getting pushed out of the nest!

    I wish they could shorten time between application and \), which is the real problem. Realistically that won't happen. Alternatively, study sections could stop giving good score to people with scads of papers who are clearly competitive for the job market. Somehow I doubt that's going to happen either. "Hey, let's turn down Joe Schmoe with his Nature and Neuron papers for this other person who has one J Neurosci paper and a lot of PROMISE!" That'll be the day.

    Maybe it's just an ill-conceived award.

  • chemicalbilology says:

    One option might be to merge the F32 and K99 programs. Have F32 be the "trainee" phase and have it automatically come with the option to submit an R00 phase application after two years that gets evaluated on a competitive basis. Competition for the R00 phase would be based on TT job readiness. I think that would do a better job of keeping the K99 part training-specific, but deal with the problem of most currently well-reviewed candidates being pretty much TT ready already.

  • DJMH says:

    Wow, that's kind of brilliant, actually. It would make life hard on people for whom the F32 was never going to be a good fit (me, for example, b/c of switching labs partway through the postdoc) but that's probably a pretty small pool of people.

    The only thing I'm not sure about--isn't the F32 only open to Americans, but the K99 open to any nationality? Haven't got time to check right now.

  • Socal_dendrite says:

    Yes, F32 is only for citizens and legal permanent residents (green card holders), whereas K99 is open to anyone.

    This lowering of the time limit definitely sucks for anyone attempting to broaden their skills by learning something new during their postdoc (which imho is one of the things a postdoc position is for), and also for anyone doing expts that are particularly time-consuming. Eg primate work, where it often takes a year to just train the animal, let alone start collecting/analysing/publishing data.

  • chemicalbilology says:

    I think it should be possible to still have the K99 training phase exist so that non-citizens were still able to apply for a training phase. I just think they should also allow the F32 to be considered a training phase towards the R00, as well.

  • bashir says:

    I dunno. The way the mechanism seems to be going is just another example of the distance between what NIH thinks is going on in the science workforce, and what actually is. The NIH decision makers and the grant reviewers are one big "left hand doesn't know what the right hand is doing"

  • qaz says:

    The K99 was designed to fix a problem that never existed - that universities weren't hiring. Universities were already over-hiring. The place that people were "falling out of the pipeline" was (and still is) actually at the tenure decision when they didn't get the grant renewal/second grant. (Because all that "NI/ESI" stuff went away.) All the K99 did was raise the bar for university hires, making universities even more dependent on NIH's internal decision-making rather than than on their own quality assessments.

    Look, the people who were talented enough to get K99s were ALREADY the cream of the crop going to get hired by universities. The K99 changed nothing for them or for anyone else.

    The K99 was stupid. It should be replaced with more F32s (for postdoc training) and better grant policies for faculty.

  • Well, guess what: your reviewers are scoring highest the people who already look ready for an independent position. So you're going to have to figure out how to get them to change their mindsets--something that so far, nobody has figured out how to do very effectively.

    When I served on a K99 review panel, I asked the IC's program staff who were observing the panel and had explained the K99 review to us, "Should we be rewarding the most accomplished post-docs or helping pull up on the bootstraps of the post-docs who aren't already as accomplished but who demonstrate potential and could most benefit from the award. Program staff flat-out refused to answer the question and said only, "Each reviewer needs to consult their own conscience in light of the stated review criteria".

    My personal opinion is that there is a large faction of hatred for the K99/R00 program within NIH, and they are doing everything they can internally to sabotage it. Sort of like republicans and the government.

  • Pharm Sci Grad says:

    As Socal_dendrite said, woe for any of us who were actually stupid enough to think that a post-doc really was a time for training and did change focus or, *gasp* even fields semi-significantly from our PhD work. Yes, in four years we can certainly learn a new area and push out some top tier papers AND apply for these awards, of course. Forget if any of these new areas involve time consuming experiments that need to be replicated for said amazing publications. *sigh* I didn't expect a Care Bears Tea Party but this shit is getting pretty ridiculous, assuming one is interested in training good scientists rather than creating pedigreed clones.

  • jakester says:

    I would not have gotten a job prior to having my K99. I was not from a famous lab, but had decent publications and put together an awesome K99/R00 application. The K phase gave me time and more independence within my postdoc lab so that I could get one more paper out and generate preliminary data for my job talk. The job I ended up getting actually required some form of funding in the job ad so the K made me eligible.

    I'm also not understanding why anyone would HAVE TO apply in year 2. One could just as easily apply in year 3.9. You should have publications by that point. You might not be able to resubmit, but you can certainly give it your best shot at the point in your postdoc when you're most competitive. In my opinion the change from 5 to 4 years doesn't really change things much.

    • arlenna says:

      Sure, you *could* apply in year 3.9 for the first time, but that is a pretty risky game if you really need or want to get the award. Even applying at the beginning of year 3 will be risky based on how long NIH takes to make decisions--it's pretty typical to wait more than 9 months before you know whether you really need to resubmit or whether you'll get picked up.

      This is the gamble that people have to make already, and the deadline change just makes it more of a gamble since as you note, the earlier you go the fewer publications you are likely to have. And I reiterate that while this might not make much of a difference for postdocs who are not affected by family timelines or who stay in their same research area/field, a year shorter eligibility period will significantly affect those who do start families or move into new research areas (unless NIH does something about it).

    • Bashir says:

      How many people do you think are getting this on one submission vs. on revision? I know one guy who got it the first time, just barley slipping in there. That was considered remarkable, and that was a few years ago when the K99 payline looked almost robust. Now the pay line is creeping down. It would be foolish to bet much one one shot.

      • jakester says:

        In my award year at the NIAID, 4/6 were given on the first try.

        • arlenna says:

          It's much, much worse at NCI (which gives out the most K99s overall). Between 2008 and 2013, there have been 148 awards made. 68 of those were on A1 (or, when it was still possible, A2) submissions. That's 46%.

        • arlenna says:

          But whoa, dang--NIGMS is way better (they give the 2nd biggest # IIRC) than NCI, at only about 12% A1 or A2 since 2008.

          I don't know how to interpret this without knowing how many applications each institute gets, though. Does this mean that nobody needs to resubmit at institutes other than NCI because the applicant pool is smaller and more people really do succeed right off the bat, or does it mean that other institutes just never fund people's resubmissions (or discourage resubmissions)?

        • arlenna says:

          Further analyses--combining NINDS and NIA, there were 108 awards since 2008 and 45 were on revision. So again, % is up there--42%.

  • me says:

    This shit is gettin scary w me bein a noob on the postdoc market 4 the first time...esp since I know my only option is to shift my focus drastically toward sexier technologies n unexplored neural turf....but having now seen the top hustlers up close n personal...I'm even more committed to sticking to my moral convictions n tryin to help the young succeed to drop that real talk n hopefully survive...despite the fact that I'm not gonna be highly competitive 4 an F32 if I'm workin on some1s startup package instead of a BSD lab...with this new K rule...it sounds like ill prob run out of time n get kicked down hard...oh well, $*&% it...get rich or die tryin!!...on that note, back to the lab to crush that daters...

  • DJMH says:

    Bashir, I don't think it's as bad as all that, unless I am dreadfully misreading Reporter. I put in 1K99NS%, figuring NINDS is the major player in the neuro community, got 11 hits, and only 2 of those were A1. So that means 9/11 = 82% of those who are getting K99s get it on the first round.

    • arlenna says:

      I commented above, too, but I just checked NCI and since 2008, 46% of K99s have been given to revised proposals. It's 12% at NIGMS, and if you look at the NIAID numbers from Jakester above, 2/6 = 33% (albeit small sample size, only one award year).

  • [...] The discovery of a microscopic world shook the foundations of theology and created modern demons K99 eligibility limited to 4 years and you better not be out on the job market Why I’m Worried About Ethical Shenanigans in the “Citizen Science” Movement (excellent) [...]

  • I suspect that this will also hurt those of us doing translational research in which there is no way that a study is done in two years.

  • Dave says:

    My personal opinion is that there is a large faction of hatred for the K99/R00 program within NIH, and they are doing everything they can internally to sabotage it. Sort of like republicans and the government.

    This is very true. Best example is NIDDK. The payline has plummeted remarkably in recent years and a few years ago you could land one with a score in mid-20s or low 30s. Now a score of 10 does not guarantee you an award. I'm convinced many ICs will drop many Ks in the next couple of years as they try and keep R paylines steady.

  • geranium says:

    I'm pretty pissed about this notice. Not just because I completely agree with everything Arlenna said---but because I made a very calculated decision to forgo applying for the K99 last month (Feb 2013) in order to maximize my application for the next cycle (June 2013). With either of these deadlines I had enough time to resubmit under the 5-year mark with my revision, under the old rules, which was the whole point. Now with the 4-year rule implemented for Feb 2014, I have no option for a revision. I don't like this grantsmanship game but I've made some very specific research and writing plans to play along, and now that the rules have changed it actually affects me pretty significantly. Definitely makes me hate the game even more.

  • Elsa says:

    So far, the longest extension on K99 eligibility that I have heard of anyone getting for family leave is the actual number of weeks/months they were out on parental leave.

    From what I'm hearing as I prepare to request EIS extension, they will not even allow me the actual number of months I was on parental leave because it was longer than the typical 3 months. I'm told this constitutes a "life choice" and cannot be considered maternal leave. Isn't that a crazy example of this paternalistic and invasive attitude? Aren't any choices about whether and when to have children or take family/eldercare leave actually "life choices"?

    How on earth is the NIH qualified to judge which choices are valid and appropriate?

  • Andrea says:

    The double whammy of shortened eligiblity and sequester falling simultaneously is an additional blow for the K99's.
    Not to bitch from a personal standpoint, buuuuut...to bitch from a personal standpoint, my first K99 application just got returned with a score in the low 20s, which my funding agency would have funded in the several preceding years. This year, budget anxiety means they lowered their payline into the teens. This would still have been ok, since all three reviewers had the same primary complaint, now easily addressable with preliminary data. So I would normally resubmit. But since I'm a 4th-about-to-enter-5th-year postdoc, this turns out to have been my last cycle of eligibility. Ouch, and also ouch.
    I echo the sentiment that this means there will be more reward for those who are handed quick-to-publish projects immediately, rather than developing something novel, and that there will be even less of a distinction between competitive job candidates and K99 candidates. I've published two mid-level first author papers in my postdoc and a handful of contributing authorships; I'm simply not ready to be a competitive job candidate without something Cell or Nature-ish. The K99 would have been the perfect bridge to get my high-risk project far enough out to publish before I looked for jobs, but now there may not be $$ to support me in the lab, and I'll end up looking into industry positions instead.

Leave a Reply