Tenure expectations and PI dropout from NIH funding

Oct 31 2014 Published by under Careerism, NIH, NIH Careerism, NIH funding

Rockey's post on PI survival also had a graph on the persistence of PIs in submitting applications.

IMG_1135.JPG

She noted that the 1989 first-R01-equivalent folks dropped off in their grant submitting persistence around years 3-5 more than the younger cohorts.

A comment by qaz on the prior post of mine triggered a thought.

What about people who otherwise didn't really want or need a NIH R01 grant but it was a requirement for (or strongly supportive of) a tenure case?

If department expectations/preferences (for tenure or in who they hired in the first place) have changed since the late 80s, this could explain the difference in early drop-out, one-and-done rates across cohort.

19 responses so far

  • Dr24 says:

    And what about those who get an R01 and then don't get tenure, so never submit again?

  • dr24hours says:

    And what of those who get their R01 but never submit again because they don't get tenure or decide they dislike professorship?

  • DrugMonkey says:

    Are you proposing this varies by cohort?

  • datahound says:

    Another related factor is the balance between medical school (or research institute)-based investigators versus school of arts and sciences-based investigators. Tenure can mean different things at different types of institutions. As of 2010, tenure implied no financial guarantee at half of medical schools (https://www.aamc.org/download/125190/data/aibvol9_no6.pdf ).

  • Joseph says:

    Without data, I would suggest that alternatives to academia might play a role. I would be interested to see how robust alternative employment opportunities were. After all, disliking a professorship and quitting is really a continuum and it probably matters a lot what else you could end up doing.

  • dr24hours says:

    @DM I see no particular reason to believe it couldn't. But I also don't have a dependent variable in mind.

  • lurker says:

    All this discussion is missing another key point of the problem. NIH has finite resources to dole out to the most productive researchers, and it would behoove NIH to spread those resources to as many investigators in the most equitable and meritorious fashion as possible. Taxpayers should be getting their best research bang for tax buck. Concentrating disproportionate funds in the cadre of the "well-connected" graybeards/bluehairs while disenfranchising the rest of the productive but unlucky not to score high enough for payline are tremendous inefficiencies that will stifle the system and future innovation.

    What these survival studies are missing is that how productive overall are the survivors (pubs/grant capita/year or citaton index/grant capita/year) versus the exiters. What would the distribution look like? I wager it would look like a bell curve that centers right at the 50% survival rate in the early 1989 year , and this distribution is shifting more and more down in the later cohorts. Only recently have I heard even elite colleagues are feeling the bite of grant triage where previously it was a guaranteed thing.

    Getting re-funded I think is becoming more and more a crapshoot of how well you grease the PO or other potential SS members at a major conference (like what Drugmoneky is doing at SfN), how well you can blow smoke into a fancy shape for your 12page research strategy, and how aware everyone else is of the size of your big swinging apparatus. Less and less is it a meritocracy based on research productivity. I personally know of several PIs with 5 or more corresponding author papers from their first R01, and all their renewals are getting triaged.

    And of course, Rockey would never ever admit to the real problem, not if it looks like the extramural grant review process is more a lottery or club game than true meritocracy.

  • lurker says:

    And to add, those PIs I know are all getting or have gotten tenure, despite their renewals getting triaged. Based on their research productivity, they deserve every bit of that tenure credit, but the achievement is completely bittersweet if they have to close down their research and teach more and more or do more administration that really is more appropriate for the 65+. We can prattle on on how fukked up the extramural program is, but unless there is a real shakeup in how NIH handles extramural finding decisions, these curves are going to look like a blood bath when plotted another 5-10 years from now.

  • poke says:

    I think it's clear from this discussion (and many others like on this blog and elsewhere) that the NIH granting system needs a radical overhaul. As a complete break with previously established methods of choosing who gets funded, I propose an adaptive system, wherein the probability of grant success is titrated according to a PI's track record.

    The current system rewards productivity with more money and punishes lack of productivity with a reduction in funding. This is backwards! Instead, CNS-only labs need to have their funding CUT until their productivity tapers off. In this way, NIH can determine the minimum amount of money a particular lab needs to maintain an acceptable level of scientific progress, as measured by publications in high-impact, peer-reviewed outlets.

    Similarly, a lack of productivity indicates that a lab needs more resources to pursue cutting edge technologies and poach experienced personnel from formerly productive labs that are suddenly unable to cope with the reduced cash inflow that their success begat. Thus, foundering labs boost their output and science moves forward.

    In such a scheme, the successful PI is the one that utilizes resources most efficiently. The true scientist (in the mold of McKnight and others) will surely out-compete all these poorly trained young pups by virtue of their penetrating insight and general brilliance alone. And the riff raff get a chance to contribute too, by having extra grant money compensate for their lack of acumen and scientific ability. Best of all, once the system equilibrates, NIH ensures that tax payer dollars are transmuted into Science at a maximal possible rate.

  • ecologist says:

    This has been an interesting series of posts using these survivorship curves. But some of the comments try to pick out changes in the *rate* of loss from the pool of awardees. It would be much easier to do that if these graphs were plotted on a logarithmic y-axis; in that case the rate of loss at any time is the slope of the graph (with the sign reversed). Just a suggestion to make it easier to find interesting patterns.

  • zb says:

    I'd look at soft money positions and the growing expectations of increased PI salary support via grants (even among officially tenured positions), and the difference in them for the 1989, 1997, 2003 cohorts.

  • E rook says:

    Along the same theme as others, I would wonder whether there's been an increase (since 1989) in research institutes (think Salk Institute etc) where PIs are not traditional Professors at Big State U. More of their job description requires successful NIH grants and this is probably reflected in their activities and support/infrastructure. Given what I've seen in reading the history of the enterprise, I'm not at all surprised by this. I confused as to what the "news" is with this analysis. Some attrition rate is to be expected. That it decreased during certain conditions and increased in certain other conditions in directions that coincide with what we know about history of congressional allocation of funds to NIH and the market response to that around the country should not all be surprising. Is Rocky proposing a change in policy to improve the type 2 success rate or the success rate of established PIs would be warranted given these data? That ... "ermagerd, people don't apply for / always renewals." If I had my tin foil hat on, I'd say we're being prepped for a policy change change that favors established PIs. Lacking this accoutrement, I'd say they were trying to apply survival analysis to NIH data without any real goal or policy issue in mind, and (we maybe?) hyped up a totally expected observation.

  • qaz says:

    We should also remember that there was a very different economy in 1993-1995 (when the 1989 group crashes) than there was in 2001-2003 (when the 1997 group crashes) or in 2007-2009 (when the 2003 group crashes). In the mid 1990s there was a booming economy, particularly in the dot-com and big-pharma worlds. So there were a lot of other options. In contrast, there was a big economic crash in 2001 and then again in 2008. So there aren't a lot of options, particularly for a 40 year old scientist with only grad school, postdocs, and a research spot or professorship on the resume.

    This graph suggests that the 1989 group is very different from the 1997 or 2003 group. The 1989 group seem to be getting out of the NIH system, whether because they have hard money positions and can go teach or whether they have non-academia options. In contrast, the 1997 and 2003 groups seem to be struggling because they want to stay in the NIH system and it's hard to get grants.

  • DrugMonkey says:

    Erook- we talked about Rep Harris' jihad and he's hardly the first to recognize the age-of-first-award issue. NIH is highly motivated to find "reasons" that can't be blamed on themselves and evidence of the benefit of the doubling interval or harm of the absence of robust budget increases. C'mon, I know you've been reading our discussions on these topics....

  • Dave says:

    As of 2010, tenure implied no financial guarantee at half of medical schools (https://www.aamc.org/download/125190/data/aibvol9_no6.pdf ).

    Indeed. We haven't experienced the Tenure of Title thing at my place yet, but I bet there are some committees debating it now.

  • anonymous postdoc (shrewshrew) says:

    I think that Grumble's grand scheme, which uses survival data generated under the current NIH system to predict the point at which an investigator is a 'solid bet', is likely to change the survival curve. There isn't any inherent law which dictates that 9 years of earned funding is necessary to separate the wheat from the chaff; it is simply a function of grant durations and tenure clocks under the current NIH system.

    I have a visceral response to all of these ideas of how to fix the system. Its not really because I would like to be a junior investigator and I see these ideas making it harder for me; after all, I can engage in magical thinking to believe that I would somehow be one of the lucky few to beat the odds. No, it's that they are so damned complicated.

    One complication in particular, though there are many: One of the strengths of the current format is that by judging a project, you can jump in at any time. A new member of study section can evaluate a set of experiments and hypotheses with their knowledge base of the field and techniques, and that is fine. But if evaluation were to become "enough of a track record" "history of excellence" rabble rabble etc, it will make evaluation by younger members of these study sections almost impossible. Because rest assured, these assessments will rely either on "my recollection and impression" (the biased approval of your friends) which is based on a shared history which is inaccessible to younger scientists (other than internalizing the biases and kill list of whoever trained you), or on "metrics" (bullshit like JIF) which are equally accessible to everyone, and which everyone agrees are probably destroying science because we chase them even though they are incapable of actually assessing "true merit" into posterity.

    Besides that, its just too fucking complicated. We would lose money that could go into R01s just to pay the additional admin staff at the NIH that would be necessary to institute these kinds of arcane systems.

  • Skeptic says:

    I have seen MDs and MD/Ph.D.s dropping out without renewal because they have other option, clinical practice, available and they don't want to go through this crazy grant writing. I thought it would be interesting to analyze if there is any difference between Ph.D. and MD.(including MD/Ph.D.) investigators in the dropping out rate.

  • drugmonkey says:

    The MD is only running for her dinner, the PHD is running for her life.

  • jmz4 says:

    " But if evaluation were to become "enough of a track record" "history of excellence" rabble rabble etc, it will make evaluation by younger members of these study sections almost impossible."
    -I agree. Impact factor is a problematic metric, and citation numbers need to be normalized for size of field, relevance of topic to medicine, difficulty of topic and a host of other factors that would necessarily go into computing an accurate track record. Not to mention when you put that much stock in self-promotion, that becomes a requisite career skill, which is something we'd probably like to avoid.

    So is the success of an R01 application in the 3-6 year period a prerequisite for tenure at most places?

    @Poke:
    That post was sarcasm, right? Sort of needling the communist vibe of egalitarianism we sometimes get in these funding discussions?

Leave a Reply