The NIH ESI policy was a generational screw job

August 23, 2012

As most of you know, when Science published the now-infamous graph showing that the average age of a PI when s/he received the first R01 NIH grant was 42, even the NIH realized there was a problem with the demographics of the extramural research workforce. This led to a number of initiatives, including the creation of the NIH’s first broadly available and genuine transition mechanism, the K99/R00. This was good because it incentivized University hiring committees to take a risk on a younger person. They would come with R level research grant money already in hand! It was also a mostly benign change because it was clear from the start that there were not going to be very many of these awards.

The NIH also realized (amazingly belatedly) that their “New Investigator” checkbox system was not having any effect on the age of first R01 award. Or, not a beneficial effect anyway. They realized what I had realized within two hours of the start of my first study section meeting, i.e. that the competitive NI applications were from highly experienced scientists who simply hadn’t sought funding from the NIH before. So they generated the “Early State Investigator” category of PI.

The ESI was distinguished by the time elapsed since the award of the PHD. Sounds okay, right? No more special perks for the previously NSF-, DOD- or CDC-funded established investigator. No more focus on the “New” investigator recently hired from a foreign country where she was highly experienced in extracting grant funding from that country’s NIH equivalent. Let us focus on the genuinely “new” investigator. Someone just starting their faculty appointment and needing help……uh-oh.

Problem was, there was this entire generation of scientists already in the pipeline. Waiting to transition.

The ESI program was an academic generation screw job which is why I say it is a blunder.

I’ve blogged about this before in the context of saying I don’t feel sorry for myself and people of my approximate generation who managed to make it over the transition hurdle. Those people of my approximate generation who managed to get their first grant before things really went in the toilet and are now complaining that they are stuck between ESI policies and the OldGuard really digging in their toes (facilitated by their “long term POs”). My sympathy for us is limited (but…grrrrr).

What I usually fail to talk about are those excellent scientists who, for one reason or another*, didn’t happen to get over the hurdle. Then, when they were all long past the ESI interval (and therefore asking for special extensions was kind of pointless) along comes the NIH “help” for younger investigators…but it explicitly jumped over them and said “too bad folks, we’re just going to ignore you and furthermore, we’re going to give hiring committees every reason in the world to screw you as well“.

Right? I mean would you hire someone who had the extra 5 yrs of postdoc’ing (with the productivity) over a younger someone with half-decent pubs but about 3 years of ESI time on the clock? You’d be doing your Department a serious disservice! Improved chances of your new hire getting their foot in the NIH grant door as early as possible is a major factor these days.

So, while I do like the NIH giving the Universities a reason to make faculty hires ever closer to the granting of the PhD…this method was a really brutal** way of choosing of winners and losers in the generational battle***.

____
*all too frequently women, all too frequently childbearing, all too frequently accommodating a slightly older academic spouse

**curious given the relative timidity of the NIH in making other dramatic changes, picking winners, etc.

**It will not surprise you in the least that I view this as yet more of the Boomers (those who run things at NIH) screwing the GenXers for whom they have zero affinity to benefit the GenY/Millenials who are their generational offspring.

40 Responses to “The NIH ESI policy was a generational screw job”

  1. Dave Says:

    I have never really grasped why the NIH is obsessed with the concept of “time elapsed since PhD”. When you think about, what fucking difference does it make?

    Like

  2. whimple Says:

    It’s clear. What the NIH really wants is age discrimination: younger than X gets a bonus. That’s illegal. “time elapsed since terminal degree” is the closest proxy that stays within the law.

    Like

  3. Annika Says:

    Though, it’s not completely cut and dried for asterisk 1, because you can ask for ESI extensions to account for time away from the lab. As far as I can tell from a quick read, you can get up to 120 months of extension for “family care responsibilities, extended periods of clinical training, extended periods of additional didactic instruction, disability, illness, active duty military service, loan repayment, natural disasters or comparable disruptive factors.” Which would allow women who had a kid or two to still be ESI perhaps 11-15 years after their PhD.

    Like

  4. brooksphd Says:

    This Gen-Xer weeps for lost opportunities as deadlines rush towards him.

    Like

  5. Dr Becca Says:

    Is ESI meant to reward people who find faculty positions the most efficiently? Since average post-doc time is 5+ years (and growing), ESI should be time from start of faculty position, not time since PhD. Shorten it to 5 years out, and I’d say they’ll get the right population.

    Like


  6. It will not surprise you in the least that I view this as yet more of the Boomers (those who run things at NIH) screwing the GenXers for whom they have zero affinity to benefit the GenY/Millenials who are their generational offspring.

    Interesting idea as a Gen Xer eager to blame Boomers myself. But “generational offspring” in what sense? In the literal sense that the parents of GenY are often Boomers (but some of their parents are the older Gen Xer’s as well) or in the cultural sense that GenY movements like Occupy resemble the movements of the Boomers in their hippie days?

    Like

  7. Dave Says:

    ESI should be time from start of faculty position, not time since PhD.

    This is much more sensible……

    Like

  8. drugmonkey Says:

    I have never really grasped why the NIH is obsessed with the concept of “time elapsed since PhD”.
    What the NIH really wants is age discrimination: younger than X gets a bonus.
    Is ESI meant to reward people who find faculty positions the most efficiently?

    My opinion is that this is all 100% motivated by the PR nightmare of the “42 years old at first major NIH award” and their plan to encourage Universities to hire Assistant Professors with fewer postdoctoral years.

    and let us be fair. once the immediate pain is over for the lost GenXers, then we’re back to an even playing field with the *potential* that Universities will indeed start finding ppl only 1-3 years out from their PhD attractive to hire.

    Like

  9. drugmonkey Says:

    Badgie-

    Not literally their *own* children, of course. More that they affiliate more with the plight of that generation because it is that of their own children.

    Like


  10. Dude, this post is incomprehsnesible. Can’t you write a goddamn fucken sentence in English without more parentheticals than there are actual words in the motherfucken sentence?

    Like

  11. Pinko Punko Says:

    Thanks for putting the big post up on this- it just makes no sense that it is not timed by first faculty award OR maybe with other considerations- is NIH that worried about a huge wave of experienced industry people coming into academia and “abusing” the policy?

    Like

  12. Virgil Says:

    What always intrigued me about these data, is 30 years ago the number was already 36. That’s pretty high!

    I would hazard a guess that at least some of the extra 6 years that have crept in over the past 3 decades, were due to people getting their PhDs later. This in turn is partly due to longer undergrad courses (now more commonly 4 instead of 3 years), taking gap years after high school, or working to pay off loans in between bachelor’s degree and grad school. Thus, we’re really looking at a 3-4 year delay being due to the “real” reason NIH wants to address… i.e., difficulty in landing a faculty position and being independent/competitive. Applying this correction, the question becomes… is a 1 year per decade actual rise in age-at-first-grant really worth all the fuss? If NIH had done nothing, in 2030 it would still be in the mid 40s when normalized for these other factors. Is that necessarily a bad thing given that everything is done later nowadays, and retirement age and life expectancy are also higher? (discounting the looming epidemic of “the sugars”)

    Like

  13. drugmonkey Says:

    Yes, it is bad.

    Next question?

    Like

  14. Ass(isstant) Prof Says:

    Ahh. Yeah. I had put as much together to explain that feeling in my backside.

    I had a gap few years between undergrad and PhD, hurried through grad school (5.5 years!), then spent 5.5 years as a postdoc (2004-2009). Breaking into the old boys club of R01 funding has been seemingly impossible.

    It’s not that early stage has not been considered by other agencies. NIH seems to have been willfully ignorant of the situation. NSF seems to know what early stage means: 5 years post-first faculty or equivalent position. Of course, NSF has but a pittance in budget compared to NIH, and it’s expected to cover a much broader subject area.

    Like

  15. Crystaldoc Says:

    Are there present examples of ESIs being treated differently than NIs in grant review or awarding? The study sections with which I am familiar (NCI) still cluster ESI and NI discussions and have the same formal policies for pick-ups. Subject to change at any moment, of course.

    Re: Annika, extension of ESI status is for time literally not employed or engaged in research for above reasons as you can formally document ( leaves of absence on FMLA or whatever) not just slower progress. I had two babies and succeeded in extending ESI status by 6 months, whoopee-fucking-do.

    Like

  16. Crystaldoc Says:

    … and unfortunately, that single extra review cycle did not result in my receipt of R01 funding within the ESI period. I could have used a few more cycles, if they were actually giving special consideration based on ESI status, which I am not at all convinced of.

    Like

  17. arrzey Says:

    While much of what is said here is correct, you are still looking at a censored (in the epidemological sense) sample. There are, in fact, lots of boomers who didn’t make it. When I got my PhD (tail end of boomers), my entering group was about 20, now the same size dept takes 5-8. Of that 20, approximately 4-5 are still in academic research. The bluehairs/greybeards (who are early-boomers) have gone through their own selection gradient. Whether they are still as good as they were when they got their first big bucks is open to debate.

    Like


  18. The study sections with which I am familiar (NCI) still cluster ESI and NI discussions and have the same formal policies for pick-ups.

    You are very confused about a number of things:

    (1) Investigator-initiated R01s (i.e., not in response to RFAs)–which are the only grants subject to NI/ESI policies–are reviewed by standing study sections at CSR, not in any Institutes, such as NCI.

    (2) Study sections do not have any policies whatsoever for “pick-ups”, as they do not make funding decisions. All they do is score grants.

    (3) NCI explicitly differentiates between NI and ESI in its official funding policy:

    For traditional R01 selections beyond the 7thpercentile, the SPL will give special consideration for new investigators, particularly those who are in an early stage of their careers. The current definition of new R01 investigators can be found on the NIH Office of Extramural Research website at: http://grants1.nih.gov/grant /glossary.htm#N4. The definition of Early Stage Investigators can be found at: http://grants.nih.gov/grant /glossary.htm#E.

    Like

  19. Jeremy Berg Says:

    I feel I should provide some background/history of the early stage investigator policy. NIH had long tracked “new investigators” (defined as those who had not previously received substantial NIH funding) and this is the basis for the “age of first NIH grant” plots. After the budget doubling ended. the NIH budget flattened out and the number of applications grew (due to events that began during the doubling). This led to substantial decreases in the success rates for applications, the “cueing up” of A1s and A2s, etc. Dr. Zerhouni and others were very concerned about the impact of these effects on individuals starting their independent careers. This led to the National Academy report “Bridges to Independence” and to the initiation of some analysis within NIH. I had already been tracking some of this within NIGMS which led to the presentation of times spent at various training stages that you recently posted (http://scientopia.org/blogs/drugmonkey/2012/02/14/nigms-data-on-the-timeline-from-ba-to-phd-to-asst-prof-to-r01/ ). NIH looked systematically at the pool of “new investigators” and discovered what many of us already knew, namely, that this was a heterogeneous pool of individuals with those early in their careers but also established investigators who were new to NIH. An analysis of data from (I believe) 5 institutes revealed that approximately half of the new investigators were not really in the early stages of their careers. NIH then examined the time since these individuals had received their doctoral degrees to find a value that would best identify the early career group and came up with the 10 year definition for “early stage investigators”. Other possibilities such as the time since initiation of the first faculty position were considered but these are harder to define and easier to manipulate than the time of the doctoral degree (which is relatively although not completely well-defined). NIH did want to encourage institutions to hired individuals earlier in their lives. The K99/R00 award was developed in parallel, largely driven by “Bridges to Independence” report.

    I do not think that there was any attempt to favor one generation over another. The only push-back (from some) was that peer review measures scientific merit and that this should be only basis for funding decisions. Dr. Zerhouni would have none of this, repeatly pointing out that the playing field is not level and that NIH and the scientific enterprise have a substantial interest in supporting the next generation of researchers. We also gathered some data that peer review scores became worse (in aggregate) for new investigators when some institutes starting using different paylines for them. This led to the policy of having the same success rate of new investigators as for established investigators submitting new (rather than competing renewal) applications.

    If there are flaws with the current policy, I do not believe they are based on inappropriate intent. The analysis may have been imperfect or based on data that did not keep up with ongoing changes. I frequently remind myself of an observation (often attributed to Napoleon): Never ascribe to malice that which is adequately explained by incompetence.

    Like

  20. whimple Says:

    NIH did want to encourage institutions to hired individuals earlier in their lives.

    Another way to do this would be to clear out the post-doctoral holding pen by putting a lifetime cap on post-doctoral funding, eg: 300% effort – 3 years at 100% funding, or 4 years at 75% funding etc from all NIH sources.

    Like

  21. drugmonkey Says:

    Given the data posted at RockTalk showing the huge increase in postdocs supported on research grant funds, that would have interesting effects whimple.

    I can see the rioting now….

    Like

  22. Jeremy Berg Says:

    I do not believe that NIH has the legal authority to cap time supported on research grants. Individuals on research grants are hired to do research; training is not acknowledged.

    Like

  23. whimple Says:

    Given the data posted at RockTalk showing the huge increase in postdocs supported on research grant funds, that [capping lifetime postdoctoral support] would have interesting effects whimple.

    Sure, but this is the actual problem. Getting a first major NIH award at 42 is just a downstream consequence of the huge increase in supported postdocs.

    Like

  24. Dave Says:

    Getting a first major NIH award at 42 is just a downstream consequence of the huge increase in supported postdocs.

    I’m not sure about that.

    Like

  25. Hermitage Says:

    ” Individuals on research grants are hired to do research; training is not acknowledged.”

    I know that’s the reality of the situation, but that makes me cry a little bit on the insides.

    Like

  26. crystaldoc Says:

    Re CPP,
    You are right about the first 2 points; I misspoke. I should have specified not “NCI study sections” but rather CSR standing study sections of the Oncology I and Oncology II integrated review groups, the vast majority of applications of which seem to be assigned to NCI. For a number of study sections there that I’ve heard about, ESI and NI grants are handled together, and those in a position to give scores may or may not pay much attention to the ESI vs NI designation. I’d be curious as to how the designations are handled in practice at other study sections– is it fairly uniform, or at the discretion of SROs?

    And yes, I know that study sections do not fund grants.

    The wording of the NCI policy on pick-ups is interesting, but a little vague. In practice, I am unconvinced that they are treating NIs that are 11 or 12 years out any differently from the official ESIs.

    Like

  27. drugmonkey Says:

    The wording of the NCI policy on pick-ups is interesting, but a little vague.

    HAHHAHAAHHAHAHAAHAAAAAAHAHAHHA!!!!!

    Like

  28. physioprof Says:

    For a number of study sections there that I’ve heard about, ESI and NI grants are handled together, and those in a position to give scores may or may not pay much attention to the ESI vs NI designation.

    This is by CSR mandate and is followed by all study sections. In my experience, assigned reviewers do point out during panel discussion whether an applicant is NI-only or NI/ESI, and I have hear discussion like, “Oh, that fucker has had NSF funding for decades”, or whatever. Whether that influences scoring behavior is never explicitly addressed.

    Like

  29. drugmonkey Says:

    In the days of the checkbox I was careful to delineate which kind of “NI” it was.

    Like

  30. qaz Says:

    Jeremy Berg writes: “I do not believe that NIH has the legal authority to cap time supported on research grants. Individuals on research grants are hired to do research; training is not acknowledged.”

    Perhaps NIH needs to start thinking a little more about training as NSF does. Perhaps the NIH system would be better if it were more organized as a means of BOTH doing great research AND training up a scientifically-literate American workforce (as NSF does). Thus, for example, we would get credit for training people who go on to teach at undergraduate institutions, and we would get credit for training people who go on to industry or policy or even other roles in society. Especially if they’ve done important research along their way.

    Perhaps NIH needs to have a better policy on postdoctoral training and a more realistic perspective on careers.

    Also, whether the generational screw job was “intentional” or not, the fact that there was a “lost generation” was definitely brought up as soon as the ESI and K99/R00 policies were introduced. There were a host of complaints brought up to CSR and to the institutes at the time. I remember this vividly because I and my colleagues were desperately trying to get that renewal or second R01. We were getting F%!@ed in comparison to the huge labs with years of experience, and suddenly the kids behind us were getting special treatment. I look around and I see only a very few who survived the drought.

    I also know that at least Volkow (NIDA) and Landis (NINDS) both knew that neither of these (hiring initial faculty, getting a first R01) were the problems because I spoke to them after the infamous SFN discussion meeting. They both said that they had done studies and that the real “losses” were happening at tenure – when people were unable to get the renewal or the second R01. They said that at the time (2006 I think it was), universities were hiring, but people were not getting tenure. I wonder if this is going to happen to the new ESI-helped generation, pushing the starvation cycle one more step farther (get them the first R01, but starve them out at tenure).

    I have said all along that what NIH needs is to manage research like a stock portfolio – you need blue chip stocks, high-risk stocks, low-risk bonds, etc. NIH needs a portfolio of empires, of small labs, of large labs, and of new labs. It should be planning this ratio explicitly, NOT just trusting to the vagaries of who is “proposing the best science”. Everyone knows that chasing the latest stock boom is a recipe for disaster.

    Like

  31. Jeremy Berg Says:

    qaz: I was speaking about legal authorities not intent. NIH and NSF are covered by different laws and the laws that gives the authority to support research does not include training as a component. Training activities are governed by a different law. I agree completely that this is unfortunate, but getting it fixed is not simple. However, I think many at NIH would be pleased if it could be modified.

    I think NIH leadership (through recent data analyses and reports) is coming to understand better what has been and is going on in the extramural world. It remains to be seen how they and the extramural community will respond.

    I know that many at NIH were (and are) aware of the importance of getting R01s renewed or getting second R01s in order to get tenure. The question of what to do about it was the topic of considerable discussion when I worked at NIGMS. NIGMS program staff certainly did discuss this aspect of proposals during paylist meetings. However, the argument that was frequently raised against giving formal special consideration to applicants from individuals up for tenure was that, at many institutions, tenure was based on the presumption that such individuals would be able to maintain funding into the future and that skewing the playing field would be potentially encouraging institutions to tenure faculty who would not be competitive for funding in the long run. The ESI policy was shaped at a time that it was believed that the post-doubling period (2004-2008?) would be a dip and then the NIH budget would improve. The financial crisis and other factors have put this is doubt. The ESIs who have been supported are coming into a very tough period for renewing their R01s. How institutions and NIH will react to this is hard to say.

    I agree with you completely that NIH should manage its investments like a stock portfolio with a relatively explicit mix of different sorts of investigators, etc. NIGMS certainly does this to some extent.

    Like

  32. GAATTC Says:

    Well. They can’t fire everybody if nobody has funding. Faculty need to make sure that they do a good/great job in other missions of the school (service on committees, teaching, etc.). If you are a jerk that does not care about these other two legs of the tenure stool, and expect to get promoted based on 2 R01s, then these current times may become ugly for you. However, if you are a team player who helps justify your base salary by doing excellent teaching and service work, you will be kept around longer than the guy who only cares about research and has no funding.

    BTW, thanks Jeremy Berg and DM (and others) for providing insight into these concerns.

    Like

  33. whimple Says:

    if you are a team player who helps justify your base salary by doing excellent teaching and service work, you will be kept around longer than the guy who only cares about research and has no funding.

    At institutions where I have been, the “three legged stool” of research, teaching and service has always been a polite fiction. The real weighting would come down something like:
    50% cash
    30% research (papers)
    18% teaching (still counts for something)
    2% service (mainly don’t piss off the dean)

    Regarding Dr. Berg’s “skewing the playing field” concerns, I gently point out that the NIH owns the playing field. If the NIH “skews” it, that is the playing field. The cognitive dissonance this creates in NIH administrators is that there is an inconsistency between how they’d like the playing field to look (qaz’s portfolio model) and the tyranny of the priority score. Once the NIH gets back to realizing that priority scores are advisory only they will be able to implement the portfolio according to the best long-term interests of their research base.

    Like

  34. Jeremy Berg Says:

    whimple: I agree with you that NIH overall tends to be too shy about considering factors other than percentile scores in making funding decisions. NIGMS is better than most ICs in this regard and I tried, through data analysis, to make it clear that funding an application at the 15th percentile was not different than funding an application at the 12th percentile (or even 10th percentile) in terms of any objective measure of expected future output. Nonetheless, we were often accused on “not supporting the best science” or “not believing in meritocracy” by some whenever we were explicit about taking other factors into account in making funding decisions.

    Like

  35. Confounding Says:

    Rather than suspect this is because of some odd “generational offspring” favoring conspiracy, I think its much more likely they had to draw the line *somewhere* and recognize that moving it later was going to be a problem.

    Whenever you do that, you screw someone. Making it too far forward dilutes “ESI” to the point that it’s meaningless or favors Gen Xers with established careers over newly minted Gen Y profs, make it too far back and Gen X gets left holding the bag.

    Unfortunately, the second scenario corrects itself over time, where the first just wrecks the intent for the life of the program.

    Like

  36. DrugMonkey Says:

    Since they’ve only managed to flatline the trend at 42 years I’d say it didn’t really work anyway.

    Like

  37. DrugMonkey Says:

    JB-

    It continues to amaze me how POs can have spent so much time listening to study section discussions and still think that a 14%ile score is meaningfully different from a 18%ile score.

    Like

  38. Jeremy Berg Says:

    DM: At least at NIGMS, I do not believe they do. But acting on this fact requires an institutional (IC) culture and processes that permit it and appropriate checks and balances that prevents program officers from just playing favorites. Furthermore, for ICs with strong factions (particularly disease-related) opening up the issue of not simply following percentiles can be an issue, I gather. Nonetheless, as you know, my approach was to try to empower program staff to make these calls and then be as transparent as possible about what they did.

    Like

  39. miko Says:

    we were often accused on “not supporting the best science” or “not believing in meritocracy” by some whenever we were explicit about taking other factors into account

    Jebus. Maybe a crash stats course should be required for participation in this process?

    Like


  40. […] The push to equalize success rates of ESI applicants with those of established investigators (generational screw-job that it is) started back in 2007 with prior NIH Director Elias Zerhouni. The mechanism to accomplish this goal […]

    Like


Leave a comment