Skip to main content

Academia (vs. Industry)

Changes in NIH Grant Policy?

There are some pretty big funding changes being proposed at the NIH that many people may not be aware of. A concerned NIH grantee has sent along some links and comments about these, and I think that they’re worth bringing up. Here’s a blog post by Mike Lauer of the NIH going into some details:

We will continue our work in monitoring, on a trans-agency level, the number and characteristics of the researchers we support with the idea that by doing so we can broaden and diversify the enterprise. We will take additional efforts to identify funding for even more early stage investigators who submit meritorious applications. When necessary, we will encourage use of bridge funding to offer additional stability and chances for obtaining an award.

To improve opportunities for early established mid-career investigators, we will take special steps to identify meritorious applicants who are only one grant away from losing all funding. Prioritizing these applicants for funding consideration may alleviate the squeeze being felt by mid-career investigators.

And we will monitor, on a trans-agency basis, investigators’ Grant Support Index, with the idea that over time and in close consultation with the extramural research community, we will phase in a resetting of expectation for total support provided to any one investigator. We plan to implement a Grant Support Index cap of 21 points, essentially the equivalent of 3 single-PI R01 grants. Over the next few weeks to months, we will meet with NIH Advisory Councils and other stakeholder groups to explore how best to phase in and implement this cap – so that formal assessment of grant support can be used to best inform, on a trans-NIH basis, our funding decisions.

It’s that last one especially that has some people cheering and some people booing (as you’ll be able to see from the extensive comments section that the post has already accumulated). The “Grant Support Index” is explained in more detail on this page, but what it comes down to (as in the quote above) is that the total support to any one PI is going to be basically three grants. And that will be that.

What has many people worried is that this seems to be a deliberate step away from any sort of merit-based funding. In fact, if the current granting system has any bias towards merit – as it is presumably supposed to – this is, in a way, an attempt to institute a bias against it. I completely understand that NIH grant review, as it is practiced in the real world, has problems. But it does at least attempt to fund what the agency and its committees believe are the best proposals.

Moreover, many grantees are objecting that the NIH has its thumb on the scale when it makes the case for this policy. Shown are two graphs – the one that the agency itself presents as evidence of diminishing returns from multiple grants, and an annotated version. The second one points out that the axes for these graphs are quite different – the Y axis is a straight log-10 scale, while the x-axis is more like a log-2 scale, which is an odd choice (as is the whole idea of plotting on two different log scales in general). It appears that in reality a lab with two R01 grants has roughly four times the impact of a lab with only one, and a lab with three has six times the impact. That straight dashed line, since it’s a straight line in a weirdo log world, seems to imply that a lab with two R01 grants would need to have about thirty times the impact of a single-grant lab in order not to be considered some sort of waste-of-money failure. What is bizarre.

I think a good airing-out of the criteria for R01 grants might end up doing everyone some good (although the opposite could happen, too). The NIH is completely within its rights (and in fact, its responsibilities) to keep an eye on how its grants are evaluated and awarded. But this proposal seems arbitrary and disingenuous, a good shot at getting the worst of both worlds. We’ll see how this plays out. . .

59 comments on “Changes in NIH Grant Policy?”

  1. NMH says:

    Regarding the 3-grant limit: needless to say, it helps R1 state schools, hurts Harvard/Stanford et al. I guess the question it may come down to is: who do you think possesses the most important worthy-of-funding ideas? If you think that a few unique individuals who have made it to the top (Harvard) have the best ideas, then the 3-grant limit will hurt progress.

    I know from some of the RO1 grant applications I have seen is that when productivity (publication) is considered, you can write down all the papers you published during that grants tenure, even if it was another grant supporting the project. IMO, this gives large labs with multiple grants seem more productive than they really are.

  2. curryworks says:

    What else do we not know- Why Everything We Know About Salt May Be Wrong

  3. JB says:

    I hate to use a dreaded management word (synergy) but I’ll just say, aside from the merit of research ideas, that infrastructure matters. Spreading grants out to a bunch of places and PIs can get quite wasteful because money is used replicating the same basic functionality instead of funding additional research once you have a foundation in place- and that’s not just equipment, but institutional knowledge and other trainees who help new ones learn things.
    What’s most interesting to me is how they chose to score the GSIs for the large network grants (U54- MLPCN, LINCS) which a lot of people have criticized over the years as consuming and concentrating too many resources. To be a PI on one of these you must be well-established which means at least one and probably two or more R01s. However the scoring system is very precisely set up such that it is impossible to be PI on a U54 and more than one R01- Even if you’re a co-PI on a network grant and on two R01s, that adds up to 22 points (10 + 6 + 6) which is exactly one more than the cutoff. Is this an effort to kill those types of grants by making it unattractive to PIs since they’d have to drop to one R01?

    1. NJBiologist says:

      If by “synergy” and “infrastructure” you mean “expensive assistant deans”, then yes, concentrating grants in a few institutions does facilitate those things. I’m not convinced, though.

      1. JB says:

        Uh, no, I mean things like instruments and databases and things that once you set them up for one grant are usable by every grant. Most instruments should be covered by a startup package or a separate grant, but there are a lot of useful things that make research more efficient that don’t make sense to do when you have 5 trainees but can be justified when you’re supporting 20.

  4. Wile E. Coyote, Genius says:

    Sounds like egalitarianism to me. Cap grants to the 1% and allow the 99% more chances to get them. I thought this idea would be broadly applauded.

    1. ParetoDivided says:

      To me this move seems like a step towards participation trophies rather than egalitarianism. ‘Egalitarian’ indicates that every application is equally deserving, which isn’t the case. The current system is supposed to reward grants based on the merit of the application (whether this is true in practice is not a topic I’m going to touch). Artificially limiting the number of grants a PI can be awarded could take away from more meritorious applications.

      1. AGMMGA says:

        When funding is scarce, a lot of excellent proposals are rejected not because they are undeserving, but simply because there are not enough grants to be awarded. In this case, selection of the awardees needs to be either random (in the best case scenario) or biased by other factors (e.g. PI name, sex, institution, etc.).
        If, as I suspect, the latter scenario applies, more grants will end up in fewer hands, not because those hands are more deserving, but simply because they are more known / are in big universities / etc… (Aka the Matthews effect)
        By limiting grant number to each lab, we should in theory give a shot at success to grants / labs that are excellent, but not so well known, and obtain far more bang for the buck (even considering the infrastructure problem that other commenters have highlighted)

  5. In Vivo Veritas says:

    3 RO1s total, or 3RO1s at one time? Back before my industry days I sat on a few study sections and was amazed to find that those people exists. 2-3 full grants at once. P.S. Being ion study section & seeing how the sausage was made as far as funding decisions was one of the main factors in me joining industry.

    1. JB says:

      These limits are for grants held at one time, but I think that’s in any given year so realistically it’s going to limit PIs to a max of 2 in some years, since you can’t align the renewals to always hold 3 and stay under the limit.

  6. JB says:

    Also doesn’t this greatly discourage co-investigator grants? You get 86% of the hit to your GSI total (6 vs 7) but 50% or less of the money depending on the number of PIs.

    1. Mzspectrum says:

      Yeah, there have been (and need to be more) comments regarding this. It will wreck collaborative proposals, and create an incentive to pursue absurdly inflated budgets with a point cap but no budget cap.

    2. CR says:

      I’m not sure why it would necessarily limit the budget. Any grant can go over the modular amount, so having 2 PI’s doesn’t necessarily limit the budget. As a co-PI you should ask for what your lab needs to do the work and not just take the normal R01 budget – which will be cut anyway.

      1. Paul Brookes says:

        This policy (6 points for multi-PI grants) will have exactly the effect described and kill collaboration. If you think multi-PI means non-modular budget automatically, tell that to NIGMS please! I have a multi-PI with one other investigator – on the A0 we went over-modular and were told by study section to go modular budget. On the A1 we got funded and they cut the budget by 40%. Our combined budget split between 2 labs is under 170 a year, so about 85k a year to my lab. So, on paper it looks like just another R01, and eats up 6 of my 21 point allowance, but the science I can do with that money is not even close to what someone with a 350k+ non-modular budget can do.

        Since 7 is a prime number, the sensible thing to do for multi-PI awards is divide the 7 points by the number of PIs and then round up. So for 2 PI’s it would be 3.5 = 4 points each. For 3 PIs it would be 2.33 = 3 points, etc.

  7. tnr says:

    Diversification of funding is a good intention but I don’t think this GSI is the way to go. Rather than set a cap, I would suggest a tiered approach to funding (think baseball: A, AA, AAA, major leagues) based on scientists’ current situation. NIH Institutes would decide how much of their money they want to spend on each tier and applications would be ranked within tiers. The early mid-career scientist mentioned in blog might be in AA tier.

    The current grant application system is like comparing a AA player to a major league player. Almost impossible for the AA player to win.

    1. ParetoDivided says:

      I think your idea is very creative and sensible. It addresses another commenter’s observation that having multiple grants allows for synergy (where a certain amount of funding is needed for the infrastructure of the lab, thus allowing additional funds to increase productivity much more efficiently) while still acknowledging that some budding researchers could do some very important work if they were only given a chance and not left scrounging for crumbs after the titans in the field have won all the grants. Your suggestion seems to be a good compromise that addresses the important arguments at both ends of the spectrum.

  8. Jumbo says:

    In the present system, if a hypothetical Albert Einstein (or Isaac Newton, or Marie Curie) of NIH grant writing existed, they could conceivably collect every grant, which surely isn’t what anyone wants. I think the change is right in that it diminishes the monopolistic role NIH has in biomedical funding. Let the market work. There are other sources. If your work is truly meritorious, you can get disease foundation or benefactor resources. We have clearly reached a time where simply asking for more funding cannot solve the research support crisis. Limiting RO1 equivalents is a sensible first step.

  9. Gildenstern says:

    “Shown are two graphs – the one that the agency itself presents as evidence of diminishing returns from multiple grants, and an annotated version. The second one points out that the axes for these graphs are quite different – the Y axis is a straight log-10 scale, while the x-axis is more like a log-2 scale, which is an odd choice (as is the whole idea of plotting on two different log scales in general). It appears that in reality a lab with two R01 grants has roughly four times the impact of a lab with only one, and a lab with three has six times the impact. That straight dashed line, since it’s a straight line in a weirdo log world, seems to imply that a lab with two R01 grants would need to have about thirty times the impact of a single-grant lab in order not to be considered some sort of waste-of-money failure. What is bizarre.”

    A statistical presentation worthy of the late and unlamented Trofim Lysenko, and perhaps what Disraeli or Mark Twain ( depending upon your view of provenance) meant by “Lies, damned lies and statistics”. An analysis unworthy of the serious and poorly understood impacts of an unconstrained Bureaucracy.

  10. banker says:

    It’s in the best interests for the research enterprise to run a diverse portfolio as with basic research (which many of these R01s are, btw) you never know what discoveries will actually matter in the long run. Big labs can still be big and productive with 2-3 R01s instead of 4-5 while small (and less established and easily funded, read: your lab is not already de-risked by being big and successful) labs can’t function with less than one. This proposal isn’t the big and sweeping attack on merit that some people are making it out to be, it’s simply making sure that we aren’t putting all our eggs in one basket that publishes lots of CNS papers now but will be insignificant in the long run.

  11. bacillus says:

    The biggest argument against this being an attack on merit is the fact that the current NIH extramural budget requires study section reviewersto make judgement calls on whether a grant scores in the 10th or 11th percentile. This can be the difference between funded or not. Time and again the case has been made that it is impossible to truly distinguish among the top 20-25% of grant applications. Therefore, asking someone to forego a fourth grant so that an equally meritorious application that is currently arbitrarily dismissed gets funded should not be detrimental to the NIH-backed scientific enterprise as a whole. Indeed, by allowing 1600 additional ideas a year to flourish it can only be beneficial.

  12. Ru barf says:

    The world will be sorry when the next incremental trivial photoredox nickel paper can’t get funded. We’ll forever have to question whether ortho ethyl would have been tolerated….. :'(

    1. Blue Lights Matter says:

      Won’t somebody please, please think of the LEDs?!

  13. mallam says:

    If NIH really want to change the way money is provided, the agency should provide a cap to the overhead costs added to each grant. Some institutions now charge 75% or more. This is robbery, simply to pay for services that would have to be paid by the university anyway, as trash, electricity, water, heating. Sure, their option would be to stop doing research, leaving the rooms and labs dark, dry, cold. But I don’t believe this will happen. So if NIH limits this number to something like 25% of the grant value, they will automatically create more funds to go around. Universities that won’t accept this percentage won’t get the money. Their top researchers will then go elsewhere,, where the university will accept the funds.

  14. Andy II says:

    If this argument is for distribution of research funds to more PI’s, let’s not forget the indirect cost associated with NIH grant. It is summarized very well in this Science article.

    “The portion of a research grant that universities can use to pay indirect costs is negotiated between individual universities and the government, although the administrative portion is capped by law. In general, universities that are awarded an NIH grant for the direct costs of a research project can receive from 10% to 100% of that amount to cover indirect costs. In fiscal year 2016, NIH paid out $6.4 billion for indirect costs in addition to the $16.9 billion in direct costs for research projects and other awards it funded.

    Indirect costs have long been controversial. Investigators often see the payments as a tax that reduces the size of their research grants. And some policy experts say that universities have too much leeway to use the money for things such as fancy new buildings. Meanwhile, many private foundations pay much lower overhead rates—for example, the Bill & Melinda Gates Foundation limits indirect costs to 10% for U.S. universities.

    A source close to the administration says the White House is concerned about indirect costs. The Trump budget’s $5.8 billion cut to NIH in 2018 could be offset partly by lowering indirect cost rates. (If the rate had been 10% in 2016, for instance, NIH’s indirect costs would have dropped by about $4.7 billion, to $1.7 billion.) But the source also acknowledged that any proposal to slash the rates would run into fierce opposition from universities, which argue that indirect costs are vital for maintaining high-quality research infrastructure and that they are already subsidizing the true cost of running a research lab.”

    1. MrRogers says:

      In fact current indirect policy encourages institutions to build new buildings. The (amortized) cost of those buildings is responsible for >50% of the overhead at some institutions.

      1. And old labs are so much better to work in.

  15. Cato says:

    To me this goes hand in hand in questions like, what is the true role of a PI (manager or truly an active scientist)? what is level of expectation of mentorship a “trainee” (graduate or postdoc) should expect from his PI in science? If my personal experience and observations in a top ten school hold broadly, too often these “big” professors fail to, in my opinion, be adequately involved in the research design and mentorship of their trainees. There are simply too many students for them to mentor effectively. I applaud efforts to reduce the amount of grants any one person can have, as this should reduce lab sizes, and provide better opportunities for scientists at non-premier institutions. It has always seemed like an unfair cycle to me: the “big” labs at premier institutions get the best students that are less likely to need attentive mentorship and teaching, which in turn allows that PI more time and resources to attract more funding. I can understand the argument that if a PI is at a premier institution he is likely to be at the top of his game, but then again the amount that insider politics and “favorite student” status goes into getting top academic positions probably screws over many bright upcoming scientists that must start at a lower tier institution.

  16. flyonthewall says:

    I concur with banker and bacillus. Anecdotally, the third and fourth R01’s to a single PI (no matter how brilliant the person) inevitably seem to fetch diminishing scientific returns. Maybe more papers are published but the cheese gets sliced thinner and thinner. We are better off getting the the brilliant ones to focus their energy on deciding on what are the 2 or 3 most interesting scientific questions to answer and focusing their efforts on answering them rather than building empires where their influence on day to day science gets dramatically diluted.

  17. gataguy says:

    It’s worth considering – at least within the context of more biologically-geared research – that the truly “impactful” science is now coming more and more from larger collaborative efforts (this is reflected in joint-grant applications). As long as the point-system accounts for split-grants (and it certainly should), setting a cap on the per-PI funding might actually foster more of these collaborative projects as a way to spread out funding in a more consistently available way.

    Large labs may generally be more productive given increased personnel & equipment, but I can’t help but approve of an effort to limit individual lab funding (and thus lab size). With the large number of erratums and retractions, fostering inter-lab and inter-institutional collaborations not only improves the probability of exciting and important findings, but enhances the value and (hopefully) the reproducibility of the data by forcing another look from collaborators.

  18. anon says:

    I agree, generally, with a cap on single PI R01 grants. Multi-PI grants need to be handled a lot differently though. These tend to be the practically translatable programs, with multiple components of chemistry, biochemistry, DMPK, pharmacology, etc. These ought to be encouraged at the expense of single PI programs.

  19. Bagger Vance says:

    Derek always seemed to say that having more pharma companies doing research from all angles was the most productive way to push (biomed) science forward. Isn’t this essentially the same approach, moving away from Prof Pfizer’s Supergroup with its multi-R01s to a larger number of smaller groups? Because going the other direction is the “Manhattan Project/Moonshot” that everyone thinks will solve problems, but never does.

  20. Concerned scientist says:

    Thanks Derek for showing how the NIH is presenting “fake” data to push the idea that increased grant support only slightly affects productivity. This is really unfortunate since the goal of the NIH and all scientists is to present data without manipulation.

    Overall, this new strategy of moving NIH from a merit-based system to a more socialist-style system for administering grants is going to be a disaster for US science. The guiding principle in the US is that you need to have great ideas that sway review committees, and the better your idea, the better your score (I have served on over 30 NIH study sections, and I have seen many great ideas get funded, and less good ideas let poorer scores). Under the new system, if you have a fantastic idea that could really be a game-changer for science or medicine, too bad. You have three R01s. NIH is saying: we don’t care about your great ideas, we care about giving money to people who don’t have as good ideas. That is sad.

    In response to previous comments, NIH already scrutinizes PIs who have funding greater than $700,000 to assess whether an additional grant application is redundant with previous work. Also, when a project is submitted for a five-year renewal, reviewers like myself are smart enough to figure out if the applicant has been productive on the previous grant cycle or if the applicant is using productivity from another grant to make their funding look more productive. New PIs are not disadvantaged in any study section I have served on. We routinely factor new PI status into the score. Frankly, everyone on the study section is rooting for the new PIs and PIs from resource-poor environments. I have seen many young PIs get great scores for great ideas, and now, many years later, they are thriving. The review system works well. I can’t speak to how well the NIH program officers root out overlapping funding. This may be something that needs looking at.

    NIH program officers have the authority to decide if there is evidence of “diminishing returns” and not award grants. Instead of some one-size-fits all system, NIH program officers should use their authority to prevent wasteful research awards. However, my own conversations with NIH program officers tells me that they feel they are doing a good job on this, and evidence for “diminishing returns” comes more from data manipulation rather than a real phenomenon.

    The reason why some PIs have four or sometimes five R01s is that they have great visionary ideas. Perhaps this is not politically correct, but there are some people who are better than others in terms of intelligence, creativity and science. We are better served by making sure these leaders have the resources to explore their best ideas. There is nothing nefarious about great scientists getting funding for their numerous great ideas. If someone only has one R01, it is either because they don’t have great resources to do the project or it’s because they aren’t able to conceptualize or present great ideas to the study sections. Until they can present a winning argument, they shouldn’t get government money.

    This plan is going to dissuade our great scientists from staying in science and academia since they will feel their work is not being rewarded and they lack resources to pursue their best ideas. We really need to look at the failure of communist systems where merit is not rewarded before we try to use these ideas in the US science system.

    1. anon says:

      Surely your post must be satire?

  21. me says:

    Could this lead to a system more like the Japanese tenure system, or maybe more like the Bercaw/Labinger setup? In which you have a head PI that (unofficially, maybe in this case) advises a number of junior PIs to submit R01 grants that, at least in spirit, the head PI would have submitted himself previously.

    Maybe not necessarily a good or bad thing, but it could be a way that the big-shot PIs find a way around the three grant limit.

    1. MrRogers says:

      Exactly this is what’s going to happen. The net result will be that the post-docs who currently write the applications will also be the PI, while the “big shot” who is co-Investigator will serve as guarantor of the work. This will necessitate some institutions (e.g. MIT) changing policies about who can serve as PI, but will otherwise allow the people doing the thinking to get the credit earlier (when they’re more creative).

      1. Curt F. says:

        This is the strongest argument in support of the reform that I have seen. I’m not sure it will happen exactly that way, but if it does, or if the reform can be tweaked to help insure that it does, then it seems promising. All PIs do not need be professors, and vice versa too. In fact, supporting non-professioral (e.g. post-doc) PIs at universities could be a nice way to get the best of both worlds: the most creative ideas might find some support, but at the university, these new PIs will still need to find permanent infrastructure in which to do their research. They aren’t all going to have their own custom-refurbished labs to work in. Thus, they may well decide to join an existing “lab”, but instead of being completely beholden to the head of that lab & associated infrastructure, they will have increased (not unlimited) flexibility to pursue their own ideas. It’s a bit like an NSF graduate fellowship, but for 32-year olds instead of 22-year olds.

  22. bacillus says:

    @ concerned scientist. Per my earlier comment, how the hell do you know that someone’s 4th or 5th brilliant idea is any better than those of the poor schmucks who arbitrarily missed the pay line on account of study section being required to decide which brilliant applications get funded and which do not. Whilst NIH budgets remain sclerotic many brilliant ideas never get the see the light of day.

    1. Concerned scientist says:

      First, study sections do not make funding decisions. They only give scores. It is the job of the program officers to decide who to fund. You are bringing up something that program officers routinely discuss. When you are a borderline score, they use numerous parameters to decide who gets funding, including the current financial status of the lab, programmatic directions of the NIH institute, overlap with other funded applications, etc. So, the review committee does not decide on anything. So, the system already takes this into account. Second, even among the top scoring applications, we routinely see a range of impact. I have seen many of these great projects come to fruition after funding, leading to huge advances in science. This was thanks to the great work of scientific reviewers and NIH program officers picking the best ideas, rather than less good ideas.

      The real solution is to increase funding, not to stifle scientists from proposing their ideas because they already go over the cap. At my institution, 3 RO1s funds a 8-person lab. Frankly, having four or five R01s for a great scientific visionary makes sense so they can get more done.

      1. SP says:

        Anecdotally, I heard of a grant that received a perfect score and was rejected by the program officer because the PI had too much funding.

        1. Concerned scientist says:

          Yes, this happens especially if the application is assigned to NIGMS. I have had a 3% grant that I was told was considered the best application in the entire study section but was not funded for this reason. NIGMS is already very strict on not allowing funding for a project to well-funded investigators regardless of how innovative the project is or how many lives could be saved. As scientific review committee members, we often know that a NIGMS-assigned grant from a well-funded investigator will not be funded, but we give our impact score regardless. Now NIH leadership wants to extend this stifling practice to all NIH institutes, even if the institute doesn’t want to follow this practice.

          US leadership in science is very precarious. For some reason NIH leadership views grants to be similar to children playing soccer where every child deserves a chance and every child gets a participation award. Francis Collins’s initiative is going to fundamentally hurt NIH by moving from a merit-based system which has allowed visionary scientists and thought leaders from the US to drive science. NIH will now be a “participation trophy” system where people with less merit get money. This will hurt NIH, visionary scientists, and US science.

  23. Toby says:

    I think log10(x)==log2(x)*log10(2) for all x. Hence changing an axis from log10 to log2 scale is nothing but a linear rescaling of that axis. Furthermore, if you plot log (either base 10 or base 2) transformed values but label the axis with the untransformed values (as in the example shown), then it makes absolutely no difference what basis you use. So it’s not such an odd idea after all…

    1. Curt F. says:

      Thank you, I was wondering how long it would take someone to say this.

    2. tangent says:

      The axes’ differing log bases is just a linear scaling, but that’s a big deal in a log-log plot, because of how it affects their reference line. Their dashed line looks innocent, but it’s a cubic function or so.

      Let me refresh myself on how this goes: in log-log, any linear proportion y = mx plots as a line with slope 1, no matter what the m, it just slides up and down as m varies (k + log x vs. log x). Their dashed line looks visually like a slope of 1, but it’s actually something like slope 3 in log10-log10 space (k log x vs. log x)– imagine compressing the horizontal axis so the [2, 20] interval takes up only the size of the vertical’s [1, 10].

      So their dashed line is the plot of a seriously superlinear polynomial. That’s tough, asking impact to go up as the third power of the dollars spent!

      Their data actually says there’s a pretty serious peak in grant efficiency, if you plot it as output/input versus input on an old-fashioned linear scale. If they believe in their data they should be aiming at that peak, not much higher or lower. I’m not buying very far into that.

  24. bacillus says:

    @ concerned scientist. Of course the crisis is due to the funding problems at NIH, especially the plummet in renewal applications which were almost certain of being funded on a continuing basis 30 years ago, hence ensuring a fairly stable career. However, we all know that the budget increases needed to return us to those heady days aren’t happening in our lifetime. From a purely statistical view, I very much doubt that any one person can beat out the next most fundable application 5 times in a row. The person with 3+ R01s most likely had the luxury of building a career and an empire starting from a single R01. Why deny the same opportunity to people with scores one percentile below the person who already has 3+ grants and is about to be awarded his 4th? As for the CNS imprimatur, these papers are far and away the most likely to be retracted because someone lied to obtain R01 number 3, 4, or 5, thereby preventing another’s career from ever taking off. I’m glad my career is no longer decided by the roll of the dice, and that I will soon be retiring. Biomedical science has turned into an anything goes dystopia during my 35 year career to the point that would place us in the “used car salesmen” category in the public eye if only they knew. Too many sociopaths willing to pull up the ladder on the hoi polloi with nary a wince of embarrassment or regret.

    1. Concerned scientist says:

      No one is saying that early career individuals should not get an R01. NIH has a fantastic program for “new investigators” and “early career investigators” which basically allows much poorer scores to be funded for exactly this reason. I see a lot of so-so applications getting funded under this program, and I think there is merit to this concept. Basically, many people get a chance to get their first R01. The challenge in academia is renewing the grant application or getting a second one. This is where we sometimes have to say that poor ideas in renewals or subsequent applications shouldn’t be prioritized over the amazing ideas from our visionaries and scientific geniuses.

      As I mentioned most program officers do not fund an applicant with 3 R01s who had an incrementally better score than someone with none. This is a long-standing NIH practice to take into account funding status. One one NIH institute funds solely based on score.

      Regarding lying in grants – there are severe penalties for this. NIH typically banishes them from getting more funding. Funding ways to further reduce this sounds like a good idea, but Francis Collins’s plan does not address this. His plan is about reducing merit-based awards, and preventing funding for the next CRISPR if you already have 3 R01s.

      1. CR says:

        @Concerned scientist… You said… “NIH has a fantastic program for “new investigators” and “early career investigators” which basically allows much poorer scores to be funded for exactly this reason.”.. and this is laughable at best and completely ignorant at worst, especially someone that has sat on so many review panels. Does the NIH have these designations? Yes. Do they help much? No. When I got my first R01 scored as a new investigator do you want to know how much “help” my score got for being a new investigator? 3 points! Can it help if the score is on the borderline? Of course. But it’s not going to help a much poorer score. It’s not a boost of 10-15 points.

      2. JimM says:

        His plan is about reducing merit-based awards, and preventing funding for the next CRISPR if you already have 3 R01s.

        Under this proposal, then, if I understand the limitations on receiving such grants properly, really brilliant academic scientists will find themselves unable to get adequate funding just as they have reached their prime.

        Where will such scientists turn? To industry?

        The proposal looks like an attempt to reduce universities and non-profit institutes to mere farm teams for the private sector.

  25. Anon says:

    What you’re seeing in their data is the law of dimishing returns.

    It simply makes sense to fund more ideas from a more diverse set of scientists, rather than give the one guy at the top the instrument he has to have because that will find the cure to everything.

    It’s only natural, more money for one idea will only tend toward a more expensive application of that idea. But the marginal value diminishes. A 1Ghz NMR costs 20x as much as an 600Mhz instrument ($16M vs. $800k no joke). So for approximately double the magnetitic field you pay 20 fold more. I would imagine a lot of science follows that sort of diminishing return.

  26. bacillus says:

    @concerned scientist. Thank you for a fair and frank exchange of views. I’m clearly more pessimistic than you about the future of the scientific enterprise which seems to be under severe strain in the West in general. I’d be delighted if your vision turns out to be correct. However, neither I nor most of the other scientists I’ve known during my career have children who followed in their footsteps. Mind you, I’ve always played in the second division!

  27. Seriously? says:

    What a stupid waste of time, and perfect example of “Scientism”. If you want to cap NIH R01s at three per PI, cap them at three. If you want to support young faculty, create a program to support young faculty. If not enough young faculty get grants, increase the budget and/or lower standards. If you want to measure citations per NIH dollar, go right ahead. Of course you’re going to observe diminishing returns.

    But don’t invent pseudo-scientific indexes plotted with “smoothing splines” on log scales and pretend you’re measuring something real.

  28. Eric says:

    I personally like the idea. I’ve never seen any actual data (we are scientists after all) to support the notion that the big labs produce more scientific breakthroughs per dollar spent. I’ve always felt that fame and ego play a much bigger role in this than most scientists are willing to admit. It’s a bit of hubris to believe that any idea I come up with is unique among the 7 billion inhabitants of this earth. The very few people that are exceptions can probably make do with 3 grants at a time.

    1. tangent says:

      This is also “explore versus exploit” in an n-armed bandit game. Do you keep pulling the levers that have worked best so far, or try some new ones to see if you hit what seems to be a better payoff? It depends how confident you are that you’ve explored the options. My gut feel is that we’re overconfident in what we think we know about research…

      (Harder than playing the slots though, because recognizing money is immediate, recognizing impact can take time…)

  29. Little buddy says:

    How about set up a grant system where one can have a professorship before age 60? I’m tired of the old guard milking the grant system into their 80s just to have a platform to sell their companies to pharma. I’m looking at you san diego…stay classy.

  30. Little buddy says:

    If you want to spend all your time in coMelanie’s, guess what big man, go into Pharma….no one buys “I am training STUDENTS”. We see right through that when we don’t see you for 4 months.

  31. Little buddy says:

    Bottom line is do not do a PhD in biomedicine. I worked for 5 years, published 2+ papers, one in Nature one in Nat Chem Biol. Professor treated me like s*** after my first author was done, did not write recommendation for further employment and gave my PhD project to pharma.

    As an experimental please understand that your ability to do quality science does not matter. Your hands matter, and that’s all your ” mentor” cares about. After your hands are done, you will be disposed of quicker than a pipette tip.

    1. anon says:

      Little buddy, you chose the wrong mentor.

      I am in regular contact with my PhD adviser for career guidance, more than 20 years since I left his lab. As a PI, I am trying to be like that.

      Sure, there are advisers who don’t want to see you after you walk out that door, and who make past students dream of the day that they can pi$$ on his grave.

      But I hope that students are better able to see that coming, these days, and avoid those labs. It is not worth it,

  32. BTDT says:

    too many of the ‘share the wealth’ grants go to schools/programs that should not be granting PhDs. Overhead is viewed as a cash cow by admin. With the shrinking/stagnant pharma/bio
    market more seek academic positions. A better limit would be on FTE allowed per grant. This goes to the classic Leroy Hood (industrial science) vs David Baltimore (cottage/village science)

  33. Andy II says:

    As “Concerned scientist” says: “…and I have seen many great ideas get funded, and less good ideas let poorer scores). Under the new system, if you have a fantastic idea that could really be a game-changer for science or medicine, too bad…”
    We all understand the review process is very subjective and “great” idea is no perfect definition. I have seen many times at the study sections, “great ideas” were generated by loud voice and a few “opinion leaders” in the field. New PI with “great idea” definition often comes from graduates/postdocs from “top tier” academic labs and do a continuation of their mentors’ work with a small tweak. Are we going to fund these “top tier labs” so that more students/postdocs get inspired by the PI and get better training? Those PI do not have time to come up with “great ideas,” and most of these ideas are generated by their students/postdocs, as we have been witnessing.

  34. Adam says:

    It’s hard to imagine there will ever be a correct system for awarding grants, because “merit” in science is almost impossible to measure until many years later when the impact and reproducibility of the work can be extensively vetted. Many of the greatest findings have been published in relatively low-impact journals.

    However, one thing which might be changed is to limit the number of Ph.D. students, perhaps by correlating this to the number of new tenure track positions that an institution opens. Currently the incentive for an institution/bigshot PI is to just bring on as many cheap hands (Ph.D students, later postdocs) as possible, rather than focus on actual training. This cheapens the degree and incentivizes sloppy work, fast publication and burying of negative results.

Comments are closed.